Amazon Elastic Compute Cloud

User Guide for Linux Instances

 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Amazon Elastic Compute Cloud: User Guide for Linux Instances

Copyright © 2021 Amazon Web Services, Inc. and/or its affiliates. All rights reserved.

Amazon's trademarks and trade dress may not be used in connection with any product or service that is not
Amazon's, in any manner that is likely to cause confusion among customers, or in any manner that disparages or
discredits Amazon. All other trademarks not owned by Amazon are the property of their respective owners, who may
or may not be affiliated with, connected to, or sponsored by Amazon.
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Table of Contents
What is Amazon EC2? ......................................................................................................................... 1
Features of Amazon EC2 ............................................................................................................. 1
How to get started with Amazon EC2 ........................................................................................... 2
Related services ......................................................................................................................... 2
Access Amazon EC2 .................................................................................................................... 3
Pricing for Amazon EC2 .............................................................................................................. 4
PCI DSS compliance .................................................................................................................... 4
Instances and AMIs ..................................................................................................................... 4
Instances ........................................................................................................................... 5
AMIs ................................................................................................................................. 6
Regions and Zones ..................................................................................................................... 7
Regions ............................................................................................................................. 7
Availability Zones ............................................................................................................. 13
Local Zones ..................................................................................................................... 15
Wavelength Zones ............................................................................................................ 18
AWS Outposts .................................................................................................................. 20
Root device volume .................................................................................................................. 21
Root device storage concepts ............................................................................................. 21
Choose an AMI by root device type .................................................................................... 23
Determine the root device type of your instance .................................................................. 23
Change the root volume to persist ..................................................................................... 24
Change the initial size of the root volume ........................................................................... 26
Set up ............................................................................................................................................. 28
Sign up for AWS ...................................................................................................................... 28
Create a key pair ...................................................................................................................... 28
Create a security group ............................................................................................................. 29
Get started tutorial .......................................................................................................................... 32
Overview ................................................................................................................................. 32
Prerequisites ............................................................................................................................ 33
Step 1: Launch an instance ........................................................................................................ 33
Step 2: Connect to your instance ............................................................................................... 34
Step 3: Clean up your instance .................................................................................................. 34
Next steps ............................................................................................................................... 35
Best practices .................................................................................................................................. 36
Tutorials .......................................................................................................................................... 38
Install a LAMP server (Amazon Linux 2) ...................................................................................... 38
Step 1: Prepare the LAMP server ........................................................................................ 39
Step 2: Test your LAMP server ........................................................................................... 42
Step 3: Secure the database server ..................................................................................... 43
Step 4: (Optional) Install phpMyAdmin ................................................................................ 44
Troubleshoot ................................................................................................................... 47
Related topics .................................................................................................................. 47
Install a LAMP server (Amazon Linux AMI) ................................................................................... 48
Step 1: Prepare the LAMP server ........................................................................................ 48
Step 2: Test your Lamp server ........................................................................................... 52
Step 3: Secure the database server ..................................................................................... 53
Step 4: (Optional) Install phpMyAdmin ................................................................................ 54
Troubleshoot ................................................................................................................... 57
Related topics .................................................................................................................. 58
Tutorial: Host a WordPress blog ................................................................................................. 58
Prerequisites .................................................................................................................... 59
Install WordPress .............................................................................................................. 59
Next steps ....................................................................................................................... 66
Help! My public DNS name changed and now my blog is broken ............................................. 66

iii
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Tutorial: Configure SSL/TLS on Amazon Linux 2 ........................................................................... 67

Prerequisites .................................................................................................................... 68
Step 1: Enable TLS on the server ....................................................................................... 68
Step 2: Obtain a CA-signed certificate ................................................................................. 70
Step 3: Test and harden the security configuration ............................................................... 75
Troubleshoot ................................................................................................................... 77
Certificate automation: Let's Encrypt with Certbot on Amazon Linux 2 .................................... 78
Tutorial: Configure SSL/TLS on Amazon Linux .............................................................................. 82
Prerequisites .................................................................................................................... 82
Step 1: Enable TLS on the server ....................................................................................... 83
Step 2: Obtain a CA-signed certificate ................................................................................. 84
Step 3: Test and harden the security configuration ............................................................... 89
Troubleshoot ................................................................................................................... 91
Certificate automation: Let's Encrypt with Certbot on Amazon Linux ....................................... 91
Tutorial: Increase the availability of your application ..................................................................... 95
Prerequisites .................................................................................................................... 96
Scale and load balance your application .............................................................................. 96
Test your load balancer ..................................................................................................... 98
Amazon Machine Images ................................................................................................................... 99
Use an AMI .............................................................................................................................. 99
Create your own AMI .............................................................................................................. 100
Buy, share, and sell AMIs ......................................................................................................... 100
Deregister your AMI ................................................................................................................ 101
Amazon Linux 2 and Amazon Linux AMI .................................................................................... 101
AMI types .............................................................................................................................. 101
Launch permissions ......................................................................................................... 101
Storage for the root device .............................................................................................. 102
Virtualization types ................................................................................................................. 104
Find a Linux AMI .................................................................................................................... 106
Find a Linux AMI using the Amazon EC2 console ................................................................ 106
Find an AMI using the AWS CLI ........................................................................................ 107
Find the latest Amazon Linux AMI using Systems Manager ................................................... 107
Use a Systems Manager parameter to find an AMI .............................................................. 108
Find a Quick Start AMI .................................................................................................... 110
Shared AMIs ........................................................................................................................... 111
Find shared AMIs ............................................................................................................ 112
Make an AMI public ........................................................................................................ 114
Share an AMI with specific AWS accounts .......................................................................... 115
Use bookmarks .............................................................................................................. 117
Guidelines for shared Linux AMIs ...................................................................................... 117
Paid AMIs .............................................................................................................................. 121
Sell your AMI ................................................................................................................. 122
Find a paid AMI .............................................................................................................. 122
Purchase a paid AMI ....................................................................................................... 123
Get the product code for your instance ............................................................................. 124
Use paid support ............................................................................................................ 124
Bills for paid and supported AMIs ..................................................................................... 125
Manage your AWS Marketplace subscriptions ..................................................................... 125
Create an Amazon EBS-backed Linux AMI .................................................................................. 125
Overview of creating Amazon EBS-backed AMIs ................................................................. 126
Create a Linux AMI from an instance ................................................................................. 126
Create a Linux AMI from a snapshot ................................................................................. 129
Launch an instance from an AMI you created ..................................................................... 130
Automate the EBS-backed AMI lifecycle .................................................................................... 130
Create an instance store-backed Linux AMI ................................................................................ 130
Overview of the creation process for instance store-backed AMIs .......................................... 131
Prerequisites .................................................................................................................. 131

iv
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Set up the AMI tools ....................................................................................................... 132

Create an AMI from an instance store-backed instance ........................................................ 135
Convert to an Amazon EBS-Backed AMI ............................................................................ 142
AMI tools reference ......................................................................................................... 145
Use encryption with EBS-backed AMIs ....................................................................................... 161
Instance-launching scenarios ............................................................................................ 161
Image-copying scenarios .................................................................................................. 164
Copy an AMI .......................................................................................................................... 166
Permissions for copying an instance store-backed AMI ......................................................... 167
Cross-Region copying ...................................................................................................... 168
Cross-account copying ..................................................................................................... 169
Encryption and copying ................................................................................................... 169
Copy an AMI .................................................................................................................. 170
Stop a pending AMI copy operation .................................................................................. 171
Understanding AMI billing ....................................................................................................... 172
AMI billing fields ............................................................................................................ 172
Find AMI billing information ............................................................................................ 173
Verify AMI charges on your bill ........................................................................................ 175
Deregister your Linux AMI ....................................................................................................... 175
Clean up your Amazon EBS-backed AMI ............................................................................ 175
Clean up your instance store-backed AMI .......................................................................... 177
Amazon Linux ........................................................................................................................ 178
Amazon Linux availability ................................................................................................ 178
Connect to an Amazon Linux instance ............................................................................... 179
Identify Amazon Linux images ......................................................................................... 179
AWS command line tools ................................................................................................. 180
Package repository ......................................................................................................... 181
Extras library (Amazon Linux 2) ........................................................................................ 183
Access source packages for reference ................................................................................ 184
cloud-init ....................................................................................................................... 184
Subscribe to Amazon Linux notifications ........................................................................... 186
Run Amazon Linux 2 on premises ..................................................................................... 187
Kernel Live Patching ....................................................................................................... 191
User provided kernels ............................................................................................................. 196
HVM AMIs (GRUB) .......................................................................................................... 196
Paravirtual AMIs (PV-GRUB) ............................................................................................. 196
Use the MATE desktop environment ......................................................................................... 201
Instances ....................................................................................................................................... 203
Instance types ........................................................................................................................ 203
Available instance types .................................................................................................. 204
Hardware specifications ................................................................................................... 207
AMI virtualization types ................................................................................................... 208
Instances built on the Nitro System .................................................................................. 209
Networking and storage features ...................................................................................... 209
Instance limits ................................................................................................................ 213
General purpose ............................................................................................................. 213
Compute optimized ........................................................................................................ 264
Memory optimized .......................................................................................................... 272
Storage optimized .......................................................................................................... 283
Accelerated computing .................................................................................................... 290
Find an instance type ...................................................................................................... 310
Change the instance type ................................................................................................ 311
Get recommendations ..................................................................................................... 317
Instance purchasing options ..................................................................................................... 320
Determine the instance lifecycle ....................................................................................... 320
On-Demand Instances ..................................................................................................... 322
Reserved Instances ......................................................................................................... 325

v
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Scheduled Instances ........................................................................................................ 363

Spot Instances ................................................................................................................ 364
Dedicated Hosts ............................................................................................................. 465
Dedicated Instances ........................................................................................................ 496
On-Demand Capacity Reservations ................................................................................... 501
Instance lifecycle .................................................................................................................... 520
Instance launch .............................................................................................................. 521
Instance stop and start (Amazon EBS-backed instances only) ................................................ 522
Instance hibernate (Amazon EBS-backed instances only) ...................................................... 522
Instance reboot .............................................................................................................. 522
Instance retirement ......................................................................................................... 523
Instance termination ....................................................................................................... 523
Differences between reboot, stop, hibernate, and terminate ................................................. 523
Launch .......................................................................................................................... 525
Connect ......................................................................................................................... 598
Stop and start ................................................................................................................ 623
Hibernate ...................................................................................................................... 627
Reboot .......................................................................................................................... 639
Retire ............................................................................................................................ 640
Terminate ...................................................................................................................... 643
Recover ......................................................................................................................... 648
Configure instances ................................................................................................................. 649
Common configuration scenarios ...................................................................................... 650
Manage software ............................................................................................................ 650
Manage users ................................................................................................................. 656
Processor state control .................................................................................................... 658
Set the time .................................................................................................................. 664
Optimize CPU options ..................................................................................................... 668
Change the hostname ..................................................................................................... 694
Set up dynamic DNS ....................................................................................................... 697
Run commands at launch ................................................................................................ 698
Instance metadata and user data ...................................................................................... 704
Elastic Inference ..................................................................................................................... 738
Identify instances ................................................................................................................... 738
Inspect the instance identity document ............................................................................. 738
Inspect the system UUID ................................................................................................. 738
Monitor ......................................................................................................................................... 740
Automated and manual monitoring .......................................................................................... 741
Automated monitoring tools ............................................................................................ 741
Manual monitoring tools ................................................................................................. 742
Best practices for monitoring ................................................................................................... 742
Monitor the status of your instances ......................................................................................... 743
Instance status checks ..................................................................................................... 743
Scheduled events ........................................................................................................... 750
Monitor your instances using CloudWatch .................................................................................. 761
Enable detailed monitoring .............................................................................................. 761
List available metrics ...................................................................................................... 763
Get statistics for metrics .................................................................................................. 774
Graph metrics ................................................................................................................ 782
Create an alarm ............................................................................................................. 782
Create alarms that stop, terminate, reboot, or recover an instance ........................................ 784
Automate Amazon EC2 with EventBridge .................................................................................. 797
Monitor memory and disk metrics ............................................................................................ 797
Collect metrics using the CloudWatch agent ...................................................................... 798
Deprecated: Collect metrics using the CloudWatch monitoring scripts .................................... 798
Logging API calls with AWS CloudTrail ...................................................................................... 805
Amazon EC2 and Amazon EBS information in CloudTrail ...................................................... 806

vi
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Understand Amazon EC2 and Amazon EBS log file entries ................................................... 806
Auditing users that connect via EC2 Instance Connect ......................................................... 807
Networking .................................................................................................................................... 809
Instance IP addressing ............................................................................................................. 809
Private IPv4 addresses and internal DNS hostnames ............................................................ 809
Public IPv4 addresses and external DNS hostnames ............................................................ 810
Elastic IP addresses (IPv4) ................................................................................................ 811
Amazon DNS server ........................................................................................................ 811
IPv6 addresses ............................................................................................................... 811
Work with the IPv4 addresses for your instances ................................................................ 812
Work with the IPv6 addresses for your instances ................................................................ 815
Multiple IP addresses ...................................................................................................... 817
Bring your own IP addresses .................................................................................................... 825
Requirements ................................................................................................................. 825
Prepare to bring your address range to your AWS account ................................................... 826
Provision the address range for use with AWS .................................................................... 828
Advertise the address range through AWS ......................................................................... 829
Work with your address range .......................................................................................... 829
Deprovision the address range ......................................................................................... 830
Elastic IP addresses ................................................................................................................. 831
Elastic IP address basics .................................................................................................. 831
Work with Elastic IP addresses ......................................................................................... 832
Use reverse DNS for email applications ............................................................................. 838
Elastic IP address limit .................................................................................................... 838
Network interfaces ................................................................................................................. 839
Network interface basics ................................................................................................. 839
Network cards ................................................................................................................ 841
IP addresses per network interface per instance type .......................................................... 841
Work with network interfaces .......................................................................................... 854
Scenarios for network interfaces ...................................................................................... 860
Best practices for configuring network interfaces ................................................................ 862
Requester-managed network interfaces ............................................................................. 863
Enhanced networking .............................................................................................................. 864
Enhanced networking support .......................................................................................... 865
Enable enhanced networking on your instance ................................................................... 865
Enhanced networking: ENA .............................................................................................. 865
Enhanced networking: Intel 82599 VF ............................................................................... 875
Operating system optimizations ....................................................................................... 881
Monitor network performance .......................................................................................... 884
Troubleshoot ENA ........................................................................................................... 887
Elastic Fabric Adapter .............................................................................................................. 895
EFA basics ...................................................................................................................... 896
Supported interfaces and libraries .................................................................................... 897
Supported instance types ................................................................................................ 897
Supported AMIs ............................................................................................................. 898
EFA limitations ............................................................................................................... 898
Get started with EFA and MPI .......................................................................................... 898
Get started with EFA and NCCL ........................................................................................ 906
Work with EFA ............................................................................................................... 923
Monitor an EFA .............................................................................................................. 925
Verify the EFA installer using a checksum .......................................................................... 926
Placement groups ................................................................................................................... 927
Cluster placement groups ................................................................................................ 928
Partition placement groups .............................................................................................. 929
Spread placement groups ................................................................................................ 930
Placement group rules and limitations .............................................................................. 930
Create a placement group ............................................................................................... 931

vii
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Tag a placement group ................................................................................................... 932

Launch instances in a placement group ............................................................................. 935
Describe instances in a placement group ........................................................................... 936
Change the placement group for an instance ..................................................................... 937
Delete a placement group ............................................................................................... 938
Network MTU ......................................................................................................................... 939
Jumbo frames (9001 MTU) .............................................................................................. 940
Path MTU Discovery ........................................................................................................ 940
Check the path MTU between two hosts ........................................................................... 940
Check and set the MTU on your Linux instance ................................................................... 941
Troubleshoot .................................................................................................................. 942
Virtual private clouds .............................................................................................................. 942
Amazon VPC documentation ............................................................................................ 942
EC2-Classic ............................................................................................................................ 943
Detect supported platforms ............................................................................................. 943
Instance types available in EC2-Classic .............................................................................. 944
Differences between instances in EC2-Classic and a VPC ...................................................... 945
Share and access resources between EC2-Classic and a VPC ................................................. 949
ClassicLink ..................................................................................................................... 950
Migrate from EC2-Classic to a VPC ................................................................................... 962
Security ......................................................................................................................................... 972
Infrastructure security ............................................................................................................. 972
Network isolation ........................................................................................................... 973
Isolation on physical hosts ............................................................................................... 973
Controlling network traffic ............................................................................................... 973
Interface VPC endpoints .......................................................................................................... 974
Create an interface VPC endpoint ..................................................................................... 974
Create an interface VPC endpoint policy ............................................................................ 974
Resilience .............................................................................................................................. 975
Data protection ...................................................................................................................... 976
Encryption at rest ........................................................................................................... 976
Encryption in transit ....................................................................................................... 977
Identity and access management .............................................................................................. 977
Network access to your instance ....................................................................................... 977
Amazon EC2 permission attributes .................................................................................... 977
IAM and Amazon EC2 ..................................................................................................... 978
IAM policies ................................................................................................................... 979
IAM roles ..................................................................................................................... 1032
Network access ............................................................................................................. 1041
Key pairs ............................................................................................................................. 1044
Create or import a key pair ............................................................................................ 1045
Tag a key pair .............................................................................................................. 1048
Retrieve the public key for your key pair .......................................................................... 1050
Retrieve the public key for your key pair through instance metadata .................................... 1050
Locate the public key on an instance ............................................................................... 1051
Identify the key pair that was specified at launch ............................................................. 1051
(Optional) Verify your key pair's fingerprint ...................................................................... 1052
Add or replace a key pair for your instance ...................................................................... 1052
Connect to your Linux instance if you lose your private key ................................................ 1053
Delete your key pair ...................................................................................................... 1057
Security groups .................................................................................................................... 1058
Security group rules ...................................................................................................... 1060
Connection tracking ...................................................................................................... 1061
Default and custom security groups ................................................................................ 1062
Work with security groups ............................................................................................. 1063
Security group rules for different use cases ...................................................................... 1072
Update management ............................................................................................................ 1077

viii
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Compliance validation ........................................................................................................... 1078

Storage ....................................................................................................................................... 1079
Amazon EBS ........................................................................................................................ 1080
Features of Amazon EBS ................................................................................................ 1081
EBS volumes ................................................................................................................ 1082
EBS snapshots .............................................................................................................. 1126
EBS data services .......................................................................................................... 1163
EBS volumes and NVMe ................................................................................................ 1214
EBS optimization .......................................................................................................... 1217
EBS performance .......................................................................................................... 1237
EBS CloudWatch metrics ................................................................................................ 1252
EBS CloudWatch events ................................................................................................. 1258
EBS quotas .................................................................................................................. 1268
Instance store ....................................................................................................................... 1268
Instance store lifetime ................................................................................................... 1269
Instance store volumes .................................................................................................. 1270
Add instance store volumes ........................................................................................... 1277
SSD instance store volumes ........................................................................................... 1280
Instance store swap volumes .......................................................................................... 1282
Optimize disk performance ............................................................................................ 1284
File storage .......................................................................................................................... 1285
Amazon S3 .................................................................................................................. 1285
Amazon EFS ................................................................................................................. 1287
Instance volume limits ........................................................................................................... 1291
Nitro System volume limits ............................................................................................ 1291
Linux-specific volume limits ........................................................................................... 1292
Bandwidth versus capacity ............................................................................................. 1292
Name devices ....................................................................................................................... 1292
Available device names .................................................................................................. 1292
Device name considerations ........................................................................................... 1293
Block device mapping ........................................................................................................... 1294
Block device mapping concepts ...................................................................................... 1294
AMI block device mapping ............................................................................................. 1297
Instance block device mapping ....................................................................................... 1299
Resources and tags ....................................................................................................................... 1304
Resource locations ................................................................................................................ 1304
Resource IDs ........................................................................................................................ 1305
List and filter your resources .................................................................................................. 1306
List and filter resources using the console ........................................................................ 1306
List and filter using the CLI and API ................................................................................ 1309
Tag your resources ................................................................................................................ 1311
Tag basics .................................................................................................................... 1312
Tag your resources ........................................................................................................ 1313
Tag restrictions ............................................................................................................. 1315
Tag your resources for billing ......................................................................................... 1316
Work with tags using the console ................................................................................... 1316
Work with tags using the command line .......................................................................... 1319
Add tags to a resource using CloudFormation ................................................................... 1322
Service quotas ...................................................................................................................... 1323
View your current limits ................................................................................................ 1323
Request an increase ...................................................................................................... 1325
Limits on email sent using port 25 ................................................................................. 1325
Usage reports ....................................................................................................................... 1325
Troubleshooting ............................................................................................................................ 1326
Troubleshoot launch issues .................................................................................................... 1326
Instance limit exceeded ................................................................................................. 1326
Insufficient instance capacity .......................................................................................... 1327

ix
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

The requested configuration is currently not supported. Please check the documentation for
supported configurations. .............................................................................................. 1327
Instance terminates immediately .................................................................................... 1328
Connect to your instance ....................................................................................................... 1329
Common causes for connection issues ............................................................................. 1329
Error connecting to your instance: Connection timed out ................................................... 1330
Error: unable to load key ... Expecting: ANY PRIVATE KEY ................................................... 1333
Error: User key not recognized by server .......................................................................... 1333
Error: Permission denied or connection closed by [instance] port 22 .................................... 1334
Error: Unprotected private key file .................................................................................. 1336
Error: Private key must begin with "-----BEGIN RSA PRIVATE KEY-----" and end with "-----END
RSA PRIVATE KEY-----" .................................................................................................. 1337
Error: Server refused our key or No supported authentication methods available ................... 1337
Cannot ping instance .................................................................................................... 1337
Error: Server unexpectedly closed network connection ....................................................... 1338
Error: Host key validation failed for EC2 Instance Connect .................................................. 1338
Stop your instance ................................................................................................................ 1339
Force stop the instance ................................................................................................. 1339
Create a replacement instance ........................................................................................ 1340
Terminate your instance ........................................................................................................ 1342
Delayed instance termination ......................................................................................... 1342
Terminated instance still displayed .................................................................................. 1342
Instances automatically launched or terminated ............................................................... 1342
Failed status checks .............................................................................................................. 1342
Review status check information ..................................................................................... 1343
Retrieve the system logs ................................................................................................ 1344
Troubleshoot system log errors for Linux-based instances .................................................. 1344
Out of memory: kill process ........................................................................................... 1345
ERROR: mmu_update failed (Memory management update failed) ...................................... 1346
I/O error (block device failure) ....................................................................................... 1346
I/O ERROR: neither local nor remote disk (Broken distributed block device) .......................... 1347
request_module: runaway loop modprobe (Looping legacy kernel modprobe on older Linux
versions) ...................................................................................................................... 1348
"FATAL: kernel too old" and "fsck: No such file or directory while trying to open /dev" (Kernel
and AMI mismatch) ....................................................................................................... 1349
"FATAL: Could not load /lib/modules" or "BusyBox" (Missing kernel modules) ........................ 1350
ERROR Invalid kernel (EC2 incompatible kernel) ................................................................ 1351
fsck: No such file or directory while trying to open... (File system not found) ......................... 1352
General error mounting filesystems (failed mount) ............................................................ 1353
VFS: Unable to mount root fs on unknown-block (Root filesystem mismatch) ........................ 1355
Error: Unable to determine major/minor number of root device... (Root file system/device
mismatch) .................................................................................................................... 1356
XENBUS: Device with no driver... ..................................................................................... 1357
... days without being checked, check forced (File system check required) ............................. 1358
fsck died with exit status... (Missing device) ...................................................................... 1358
GRUB prompt (grubdom>) ............................................................................................. 1359
Bringing up interface eth0: Device eth0 has different MAC address than expected, ignoring.
(Hard-coded MAC address) ............................................................................................. 1361
Unable to load SELinux Policy. Machine is in enforcing mode. Halting now. (SELinux
misconfiguration) .......................................................................................................... 1362
XENBUS: Timeout connecting to devices (Xenbus timeout) ................................................. 1363
Troubleshoot an unreachable instance ..................................................................................... 1363
Instance reboot ............................................................................................................ 1364
Instance console output ................................................................................................. 1364
Capture a screenshot of an unreachable instance .............................................................. 1365
Instance recovery when a host computer fails .................................................................. 1366
Boot from the wrong volume ................................................................................................. 1366

x
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

EC2Rescue for Linux .............................................................................................................. 1367

Install EC2Rescue for Linux ............................................................................................ 1368
(Optional) Verify the signature of EC2Rescue for Linux ...................................................... 1369
Work with EC2Rescue for Linux ...................................................................................... 1371
Develop EC2Rescue modules .......................................................................................... 1373
Send a diagnostic interrupt .................................................................................................... 1377
Supported instance types .............................................................................................. 1378
Prerequisites ................................................................................................................ 1378
Send a diagnostic interrupt ............................................................................................ 1380
Document history ......................................................................................................................... 1381
History for previous years ...................................................................................................... 1387

xi
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Features of Amazon EC2

What is Amazon EC2?

Amazon Elastic Compute Cloud (Amazon EC2) provides scalable computing capacity in the Amazon Web
Services (AWS) Cloud. Using Amazon EC2 eliminates your need to invest in hardware up front, so you
can develop and deploy applications faster. You can use Amazon EC2 to launch as many or as few virtual
servers as you need, configure security and networking, and manage storage. Amazon EC2 enables you
to scale up or down to handle changes in requirements or spikes in popularity, reducing your need to
forecast traffic.

For more information about cloud computing, see What is Cloud Computing?

Topics on this page:

• Features of Amazon EC2 (p. 1)
• How to get started with Amazon EC2 (p. 2)
• Related services (p. 2)
• Access Amazon EC2 (p. 3)
• Pricing for Amazon EC2 (p. 4)
• PCI DSS compliance (p. 4)
• Instances and AMIs (p. 4)
• Regions and Zones (p. 7)
• Amazon EC2 root device volume (p. 21)

Features of Amazon EC2

Amazon EC2 provides the following features:

• Virtual computing environments, known as instances

• Preconfigured templates for your instances, known as Amazon Machine Images (AMIs), that package
the bits you need for your server (including the operating system and additional software)
• Various configurations of CPU, memory, storage, and networking capacity for your instances, known as
instance types
• Secure login information for your instances using key pairs (AWS stores the public key, and you store
the private key in a secure place)
• Storage volumes for temporary data that's deleted when you stop, hibernate, or terminate your
instance, known as instance store volumes
• Persistent storage volumes for your data using Amazon Elastic Block Store (Amazon EBS), known as
Amazon EBS volumes
• Multiple physical locations for your resources, such as instances and Amazon EBS volumes, known as
Regions and Availability Zones
• A firewall that enables you to specify the protocols, ports, and source IP ranges that can reach your
instances using security groups
• Static IPv4 addresses for dynamic cloud computing, known as Elastic IP addresses
• Metadata, known as tags, that you can create and assign to your Amazon EC2 resources
• Virtual networks you can create that are logically isolated from the rest of the AWS Cloud, and that
you can optionally connect to your own network, known as virtual private clouds (VPCs)

1
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
How to get started with Amazon EC2

For more information about the features of Amazon EC2, see the Amazon EC2 product page.

For more information about running your website on AWS, see Web Hosting.

How to get started with Amazon EC2

First, you need to get set up to use Amazon EC2. After you are set up, you are ready to complete the
Getting Started tutorial for Amazon EC2. Whenever you need more information about an Amazon EC2
feature, you can read the technical documentation.

Get up and running

• Set up to use Amazon EC2 (p. 28)

• Tutorial: Get started with Amazon EC2 Linux instances (p. 32)

Basics

• Instances and AMIs (p. 4)

• Regions and Zones (p. 7)
• Instance types (p. 203)
• Tags (p. 1311)

Networking and security

• Key pairs (p. 1044)

• Security groups (p. 1058)
• Elastic IP addresses (p. 831)
• Virtual private clouds (p. 942)

Storage

• Amazon EBS (p. 1080)

• Instance store (p. 1268)

Working with Linux instances

• AWS Systems Manager Run Command in the AWS Systems Manager User Guide
• Tutorial: Install a LAMP web server on Amazon Linux 2 (p. 38)
• Tutorial: Configure SSL/TLS on Amazon Linux 2 (p. 67)

If you have questions about whether AWS is right for you, contact AWS Sales. If you have technical
questions about Amazon EC2, use the Amazon EC2 forum.

Related services
You can provision Amazon EC2 resources, such as instances and volumes, directly using Amazon EC2.
You can also provision Amazon EC2 resources using other services in AWS. For more information, see the
following documentation:

2
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Access Amazon EC2

• Amazon EC2 Auto Scaling User Guide

• AWS CloudFormation User Guide
• AWS Elastic Beanstalk Developer Guide
• AWS OpsWorks User Guide

To automatically distribute incoming application traffic across multiple instances, use Elastic Load
Balancing. For more information, see the Elastic Load Balancing User Guide.

To get a managed relational database in the cloud, use Amazon Relational Database Service (Amazon
RDS) to launch a database instance. Although you can set up a database on an EC2 instance, Amazon
RDS offers the advantage of handling your database management tasks, such as patching the software,
backing up, and storing the backups. For more information, see the Amazon Relational Database Service
Developer Guide.

To make it easier to manage Docker containers on a cluster of EC2 instances, use Amazon Elastic
Container Service (Amazon ECS). For more information, see the Amazon Elastic Container Service
Developer Guide or the Amazon Elastic Container Service User Guide for AWS Fargate.

To monitor basic statistics for your instances and Amazon EBS volumes, use Amazon CloudWatch. For
more information, see the Amazon CloudWatch User Guide.

To detect potentially authorized or malicious use of your EC2 instances, use Amazon GuardDuty. For
more information see the Amazon GuardDuty User Guide.

Access Amazon EC2

Amazon EC2 provides a web-based user interface, the Amazon EC2 console. If you've signed up for an
AWS account, you can access the Amazon EC2 console by signing into the AWS Management Console and
selecting EC2 from the console home page.

If you prefer to use a command line interface, you have the following options:

AWS Command Line Interface (CLI)

Provides commands for a broad set of AWS products, and is supported on Windows, Mac, and Linux.
To get started, see AWS Command Line Interface User Guide. For more information about the
commands for Amazon EC2, see ec2 in the AWS CLI Command Reference.
AWS Tools for Windows PowerShell

Provides commands for a broad set of AWS products for those who script in the PowerShell
environment. To get started, see the AWS Tools for Windows PowerShell User Guide. For more
information about the cmdlets for Amazon EC2, see the AWS Tools for PowerShell Cmdlet
Reference.

Amazon EC2 supports creating resources using AWS CloudFormation. You create a template, in JSON
or YAML, that describes your AWS resources, and AWS CloudFormation provisions and configures
those resources for you. You can reuse your CloudFormation templates to provision the same resources
multiple times, whether in the same Region and account or in multiple Regions and accounts. For more
information about the resource types and properties for Amazon EC2, see EC2 resource type reference in
the AWS CloudFormation User Guide.

Amazon EC2 provides a Query API. These requests are HTTP or HTTPS requests that use the HTTP verbs
GET or POST and a Query parameter named Action. For more information about the API actions for
Amazon EC2, see Actions in the Amazon EC2 API Reference.

3
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Pricing for Amazon EC2

If you prefer to build applications using language-specific APIs instead of submitting a request over
HTTP or HTTPS, AWS provides libraries, sample code, tutorials, and other resources for software
developers. These libraries provide basic functions that automate tasks such as cryptographically signing
your requests, retrying requests, and handling error responses, making it is easier for you to get started.
For more information, see Tools to Build on AWS.

Pricing for Amazon EC2

When you sign up for AWS, you can get started with Amazon EC2 for free using the AWS Free Tier.

Amazon EC2 provides the following purchasing options for instances:

On-Demand Instances

Pay for the instances that you use by the second, with no long-term commitments or upfront
payments.
Savings Plans

You can reduce your Amazon EC2 costs by making a commitment to a consistent amount of usage,
in USD per hour, for a term of 1 or 3 years.
Reserved Instances

You can reduce your Amazon EC2 costs by making a commitment to a specific instance
configuration, including instance type and Region, for a term of 1 or 3 years.
Spot Instances

Request unused EC2 instances, which can reduce your Amazon EC2 costs significantly.

For a complete list of charges and prices for Amazon EC2, see Amazon EC2 pricing.

To calculate the cost of a sample provisioned environment, see Cloud Economics Center.

To see your bill, go to the Billing and Cost Management Dashboard in the AWS Billing and Cost
Management console. Your bill contains links to usage reports that provide details about your bill. To
learn more about AWS account billing, see AWS Billing and Cost Management User Guide.

If you have questions concerning AWS billing, accounts, and events, contact AWS Support.

For an overview of Trusted Advisor, a service that helps you optimize the costs, security, and performance
of your AWS environment, see AWS Trusted Advisor.

PCI DSS compliance

Amazon EC2 supports the processing, storage, and transmission of credit card data by a merchant or
service provider, and has been validated as being compliant with Payment Card Industry (PCI) Data
Security Standard (DSS). For more information about PCI DSS, including how to request a copy of the
AWS PCI Compliance Package, see PCI DSS Level 1.

Instances and AMIs

An Amazon Machine Image (AMI) is a template that contains a software configuration (for example, an
operating system, an application server, and applications). From an AMI, you launch an instance, which is

4
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instances

a copy of the AMI running as a virtual server in the cloud. You can launch multiple instances of an AMI, as
shown in the following figure.

Your instances keep running until you stop, hibernate, or terminate them, or until they fail. If an instance
fails, you can launch a new one from the AMI.

Instances
An instance is a virtual server in the cloud. Its configuration at launch is a copy of the AMI that you
specified when you launched the instance.

You can launch different types of instances from a single AMI. An instance type essentially determines
the hardware of the host computer used for your instance. Each instance type offers different compute
and memory capabilities. Select an instance type based on the amount of memory and computing power
that you need for the application or software that you plan to run on the instance. For more information
about the hardware specifications for each Amazon EC2 instance type, see Amazon EC2 Instance Types.

After you launch an instance, it looks like a traditional host, and you can interact with it as you would any
computer. You have complete control of your instances; you can use sudo to run commands that require
root privileges.

Your AWS account has a limit on the number of instances that you can have running. For more
information about this limit, and how to request an increase, see How many instances can I run in
Amazon EC2 in the Amazon EC2 General FAQ.

Storage for your instance

The root device for your instance contains the image used to boot the instance. For more information,
see Amazon EC2 root device volume (p. 21).

Your instance may include local storage volumes, known as instance store volumes, which you
can configure at launch time with block device mapping. For more information, see Block device
mapping (p. 1294). After these volumes have been added to and mapped on your instance, they are
available for you to mount and use. If your instance fails, or if your instance is stopped or terminated,
the data on these volumes is lost; therefore, these volumes are best used for temporary data. To keep
important data safe, you should use a replication strategy across multiple instances, or store your
persistent data in Amazon S3 or Amazon EBS volumes. For more information, see Storage (p. 1079).

5
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMIs

Security best practices

• Use AWS Identity and Access Management (IAM) to control access to your AWS resources, including
your instances. You can create IAM users and groups under your AWS account, assign security
credentials to each, and control the access that each has to resources and services in AWS. For more
information, see Identity and access management for Amazon EC2 (p. 977).
• Restrict access by only allowing trusted hosts or networks to access ports on your instance. For
example, you can restrict SSH access by restricting incoming traffic on port 22. For more information,
see Amazon EC2 security groups for Linux instances (p. 1058).
• Review the rules in your security groups regularly, and ensure that you apply the principle of least
privilege—only open up permissions that you require. You can also create different security groups
to deal with instances that have different security requirements. Consider creating a bastion security
group that allows external logins, and keep the remainder of your instances in a group that does not
allow external logins.
• Disable password-based logins for instances launched from your AMI. Passwords can be found or
cracked, and are a security risk. For more information, see Disable password-based remote logins for
root (p. 118). For more information about sharing AMIs safely, see Shared AMIs (p. 111).

Stop and terminate instances

You can stop or terminate a running instance at any time.

Stop an instance

When an instance is stopped, the instance performs a normal shutdown, and then transitions to a
stopped state. All of its Amazon EBS volumes remain attached, and you can start the instance again at a
later time.

You are not charged for additional instance usage while the instance is in a stopped state. A minimum of
one minute is charged for every transition from a stopped state to a running state. If the instance type
was changed while the instance was stopped, you will be charged the rate for the new instance type
after the instance is started. All of the associated Amazon EBS usage of your instance, including root
device usage, is billed using typical Amazon EBS prices.

When an instance is in a stopped state, you can attach or detach Amazon EBS volumes. You can also
create an AMI from the instance, and you can change the kernel, RAM disk, and instance type.

Terminate an instance

When an instance is terminated, the instance performs a normal shutdown. The root device volume is
deleted by default, but any attached Amazon EBS volumes are preserved by default, determined by each
volume's deleteOnTermination attribute setting. The instance itself is also deleted, and you can't
start the instance again at a later time.

To prevent accidental termination, you can disable instance termination. If you do so, ensure that
the disableApiTermination attribute is set to true for the instance. To control the behavior
of an instance shutdown, such as shutdown -h in Linux or shutdown in Windows, set the
instanceInitiatedShutdownBehavior instance attribute to stop or terminate as desired.
Instances with Amazon EBS volumes for the root device default to stop, and instances with instance-
store root devices are always terminated as the result of an instance shutdown.

For more information, see Instance lifecycle (p. 520).

AMIs
Amazon Web Services (AWS) publishes many Amazon Machine Images (AMIs) that contain common
software configurations for public use. In addition, members of the AWS developer community have

6
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Regions and Zones

published their own custom AMIs. You can also create your own custom AMI or AMIs; doing so enables
you to quickly and easily start new instances that have everything you need. For example, if your
application is a website or a web service, your AMI could include a web server, the associated static
content, and the code for the dynamic pages. As a result, after you launch an instance from this AMI,
your web server starts, and your application is ready to accept requests.

All AMIs are categorized as either backed by Amazon EBS, which means that the root device for an
instance launched from the AMI is an Amazon EBS volume, or backed by instance store, which means
that the root device for an instance launched from the AMI is an instance store volume created from a
template stored in Amazon S3.

The description of an AMI indicates the type of root device (either ebs or instance store). This is
important because there are significant differences in what you can do with each type of AMI. For more
information about these differences, see Storage for the root device (p. 102).

You can deregister an AMI when you have finished using it. After you deregister an AMI, you can't use it
to launch new instances. Existing instances launched from the AMI are not affected. Therefore, if you are
also finished with the instances launched from these AMIs, you should terminate them.

Regions and Zones

Amazon EC2 is hosted in multiple locations world-wide. These locations are composed of Regions,
Availability Zones, Local Zones, AWS Outposts, and Wavelength Zones. Each Region is a separate
geographic area.

• Availability Zones are multiple, isolated locations within each Region.

• Local Zones provide you the ability to place resources, such as compute and storage, in multiple
locations closer to your end users.
• AWS Outposts brings native AWS services, infrastructure, and operating models to virtually any data
center, co-location space, or on-premises facility.
• Wavelength Zones allow developers to build applications that deliver ultra-low latencies to 5G devices
and end users. Wavelength deploys standard AWS compute and storage services to the edge of
telecommunication carriers' 5G networks.

AWS operates state-of-the-art, highly available data centers. Although rare, failures can occur that affect
the availability of instances that are in the same location. If you host all of your instances in a single
location that is affected by a failure, none of your instances would be available.

To help you determine which deployment is best for you, see AWS Wavelength FAQs.

Contents
• Regions (p. 7)
• Availability Zones (p. 13)
• Local Zones (p. 15)
• Wavelength Zones (p. 18)
• AWS Outposts (p. 20)

Regions
Each Amazon EC2 Region is designed to be isolated from the other Amazon EC2 Regions. This achieves
the greatest possible fault tolerance and stability.

7
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Regions

The following diagram illustrates multiple AWS Regions.

When you view your resources, you see only the resources that are tied to the Region that you specified.
This is because Regions are isolated from each other, and we don't automatically replicate resources
across Regions.

When you launch an instance, you must select an AMI that's in the same Region. If the AMI is in
another Region, you can copy the AMI to the Region you're using. For more information, see Copy an
AMI (p. 166).

Note that there is a charge for data transfer between Regions. For more information, see Amazon EC2
Pricing - Data Transfer.

Contents
• Available Regions (p. 8)
• Regions and endpoints (p. 10)
• Describe your Regions (p. 10)
• Get the Region name (p. 12)
• Specify the Region for a resource (p. 12)

Available Regions
Your account determines the Regions that are available to you.

• An AWS account provides multiple Regions so that you can launch Amazon EC2 instances in locations
that meet your requirements. For example, you might want to launch instances in Europe to be closer
to your European customers or to meet legal requirements.
• An AWS GovCloud (US-West) account provides access to the AWS GovCloud (US-West) Region and the
AWS GovCloud (US-East) Region. For more information, see AWS GovCloud (US).
• An Amazon AWS (China) account provides access to the Beijing and Ningxia Regions only. For more
information, see AWS in China.

The following table lists the Regions provided by an AWS account. You can't describe or access additional
Regions from an AWS account, such as AWS GovCloud (US) Region or the China Regions. To use a Region
introduced after March 20, 2019, you must enable the Region. For more information, see Managing AWS
Regions in the AWS General Reference.

For information about available Wavelength Zones, see Available Wavelength Zones in the AWS
Wavelength Developer Guide.

Code Name Opt-in Status Local Zone

us-east-2 US East (Ohio) Not required Not available

us-east-1 US East (N. Virginia) Not required The following Local

Zones are in preview:

• Boston (us-east-1-
bos-1a)

8
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Regions

Code Name Opt-in Status Local Zone

• Houston (us-
east-1-iah-1a)
• Miami (us-east-1-
mia-1a)

us-west-1 US West (N. California) Not required Not available

us-west-2 US West (Oregon) Not required The following Local

Zones are in Los
Angeles:

• us-west-2-lax-1a
• us-west-2-lax-1b

af-south-1 Africa (Cape Town) Required Not available

ap-east-1 Asia Pacific (Hong Required Not available

Kong)

ap-south-1 Asia Pacific (Mumbai) Not required Not available

ap-northeast-3 Asia Pacific (Osaka- Not required Not available

Local)

ap-northeast-2 Asia Pacific (Seoul) Not required Not available

ap-southeast-1 Asia Pacific (Singapore) Not required Not available

ap-southeast-2 Asia Pacific (Sydney) Not required Not available

ap-northeast-1 Asia Pacific (Tokyo) Not required Not available

ca-central-1 Canada (Central) Not required Not available

eu-central-1 Europe (Frankfurt) Not required Not available

eu-west-1 Europe (Ireland) Not required Not available

eu-west-2 Europe (London) Not required Not available

eu-south-1 Europe (Milan) Required Not available

eu-west-3 Europe (Paris) Not required Not available

eu-north-1 Europe (Stockholm) Not required Not available

me-south-1 Middle East (Bahrain) Required Not available

sa-east-1 South America (São Not required Not available

Paulo)

For more information, see AWS Global Infrastructure.

The number and mapping of Availability Zones per Region may vary between AWS accounts. To get a list
of the Availability Zones that are available to your account, you can use the Amazon EC2 console or the
command line interface. For more information, see Describe your Regions (p. 10).

9
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Regions

Regions and endpoints

When you work with an instance using the command line interface or API actions, you must specify its
Regional endpoint. For more information about the Regions and endpoints for Amazon EC2, see Amazon
EC2 endpoints and quotas in the Amazon Web Services General Reference.

For more information about endpoints and protocols in AWS GovCloud (US-West), see AWS GovCloud
(US-West) Endpoints in the AWS GovCloud (US) User Guide.

Describe your Regions

You can use the Amazon EC2 console or the command line interface to determine which Regions are
available for your account. For more information about these command line interfaces, see Access
Amazon EC2 (p. 3).

To find your Regions using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, view the options in the Region selector.

10
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Regions

3. Your EC2 resources for this Region are displayed on the EC2 Dashboard in the Resources section.

11
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Regions

To find your Regions using the AWS CLI

• Use the describe-regions command as follows to describe the Regions that are enabled for your
account.

aws ec2 describe-regions

To describe all Regions, including any Regions that are disabled for your account, add the --all-
regions option as follows.

aws ec2 describe-regions --all-regions

To find your Regions using the AWS Tools for Windows PowerShell

• Use the Get-EC2Region command as follows to describe the Regions for your account.

PS C:\> Get-EC2Region

Get the Region name

You can use the Amazon Lightsail API to view the name of a Region.

To view the Region name using the AWS CLI

• Use the get-regions command as follows to describe the name of the specified Region.

aws lightsail get-regions --query "regions[?name=='region-name'].displayName" --output

text

The following example returns the name of the us-east-2 Region.

aws lightsail get-regions --query "regions[?name=='us-east-2'].displayName" --output

text

The following is the output:

Ohio

Specify the Region for a resource

Every time you create an Amazon EC2 resource, you can specify the Region for the resource. You can
specify the Region for a resource using the AWS Management Console or the command line.

Considerations

Some AWS resources might not be available in all Regions. Ensure that you can create the resources that
you need in the desired Regions before you launch an instance.

To specify the Region for a resource using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Use the Region selector in the navigation bar.

12
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Availability Zones

To specify the default Region using the command line

You can set the value of an environment variable to the desired Regional endpoint (for example,
https://ec2.us-east-2.amazonaws.com):

• AWS_DEFAULT_REGION (AWS CLI)

• Set-AWSDefaultRegion (AWS Tools for Windows PowerShell)

Alternatively, you can use the --region (AWS CLI) or -Region (AWS Tools for Windows PowerShell)
command line option with each individual command. For example, --region us-east-2.

For more information about the endpoints for Amazon EC2, see Amazon Elastic Compute Cloud
Endpoints.

Availability Zones
Each Region has multiple, isolated locations known as Availability Zones. When you launch an instance,
you can select an Availability Zone or let us choose one for you. If you distribute your instances across
multiple Availability Zones and one instance fails, you can design your application so that an instance in
another Availability Zone can handle requests.

The following diagram illustrates multiple Availability Zones in an AWS Region.

You can also use Elastic IP addresses to mask the failure of an instance in one Availability Zone by rapidly
remapping the address to an instance in another Availability Zone. For more information, see Elastic IP
addresses (p. 831).

An Availability Zone is represented by a Region code followed by a letter identifier; for example,
us-east-1a. To ensure that resources are distributed across the Availability Zones for a Region, we
independently map Availability Zones to names for each AWS account. For example, the Availability Zone
us-east-1a for your AWS account might not be the same location as us-east-1a for another AWS
account.

To coordinate Availability Zones across accounts, you must use the AZ ID, which is a unique and
consistent identifier for an Availability Zone. For example, use1-az1 is an AZ ID for the us-east-1
Region and it has the same location in every AWS account.

You can view AZ IDs to determine the location of resources in one account relative to the resources in
another account. For example, if you share a subnet in the Availability Zone with the AZ ID use-az2 with
another account, this subnet is available to that account in the Availability Zone whose AZ ID is also use-
az2. The AZ ID for each VPC and subnet is displayed in the Amazon VPC console. For more information,
see Working with Shared VPCs in the Amazon VPC User Guide.

As Availability Zones grow over time, our ability to expand them can become constrained. If this
happens, we might restrict you from launching an instance in a constrained Availability Zone unless you
already have an instance in that Availability Zone. Eventually, we might also remove the constrained
Availability Zone from the list of Availability Zones for new accounts. Therefore, your account might have
a different number of available Availability Zones in a Region than another account.

Contents
• Describe your Availability Zones (p. 14)

13
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Availability Zones

• Launch instances in an Availability Zone (p. 14)

• Migrate an instance to another Availability Zone (p. 14)

Describe your Availability Zones

You can use the Amazon EC2 console or the command line interface to determine which Availability
Zones are available for your account. For more information about these command line interfaces, see
Access Amazon EC2 (p. 3).

To find your Availability Zones using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, view the options in the Region selector.
3. On the navigation pane, choose EC2 Dashboard.
4. The Availability Zones are listed under Service health, Zone status.

To find your Availability Zones using the AWS CLI

1. Use the describe-availability-zones command as follows to describe the Availability Zones within the
specified Region.

aws ec2 describe-availability-zones --region region-name

2. Use the describe-availability-zones command as follows to describe the Availability Zones regardless
of the opt-in status.

aws ec2 describe-availability-zones --all-availability-zones

To find your Availability Zones using the AWS Tools for Windows PowerShell

Use the Get-EC2AvailabilityZone command as follows to describe the Availability Zones within the
specified Region.

PS C:\> Get-EC2AvailabilityZone -Region region-name

Launch instances in an Availability Zone

When you launch an instance, select a Region that puts your instances closer to specific customers, or
meets the legal or other requirements that you have. By launching your instances in separate Availability
Zones, you can protect your applications from the failure of a single location.

When you launch an instance, you can optionally specify an Availability Zone in the Region that you are
using. If you do not specify an Availability Zone, we select an Availability Zone for you. When you launch
your initial instances, we recommend that you accept the default Availability Zone, because this allows us
to select the best Availability Zone for you based on system health and available capacity. If you launch
additional instances, specify a Zone only if your new instances must be close to, or separated from, your
running instances.

Migrate an instance to another Availability Zone

If necessary, you can migrate an instance from one Availability Zone to another. For example, let's say
you are trying to modify the instance type of your instance and we can't launch an instance of the new

14
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Local Zones

instance type in the current Availability Zone. In this case, you can migrate the instance to an Availability
Zone where we are able to launch an instance of that instance type.

The migration process involves:

• Creating an AMI from the original instance

• Launching an instance in the new Availability Zone
• Updating the configuration of the new instance, as shown in the following procedure

To migrate an instance to another Availability Zone

1. Create an AMI from the instance. The procedure depends on your operating system and the type of
root device volume for the instance. For more information, see the documentation that corresponds
to your operating system and root device volume:

• Create an Amazon EBS-backed Linux AMI (p. 125)

• Create an instance store-backed Linux AMI (p. 130)
• Creating a custom Windows AMI
2. If you need to preserve the private IPv4 address of the instance, you must delete the subnet in the
current Availability Zone and then create a subnet in the new Availability Zone with the same IPv4
address range as the original subnet. Note that you must terminate all instances in a subnet before
you can delete it. Therefore, you should create AMIs from all of the instances in your subnet so that
you can move all instances from the current subnet to the new subnet.
3. Launch an instance from the AMI that you just created, specifying the new Availability Zone or
subnet. You can use the same instance type as the original instance, or select a new instance type.
For more information, see Launch instances in an Availability Zone (p. 14).
4. If the original instance has an associated Elastic IP address, associate it with the new instance. For
more information, see Disassociate an Elastic IP address (p. 836).
5. If the original instance is a Reserved Instance, change the Availability Zone for your reservation. (If
you also changed the instance type, you can also change the instance type for your reservation.) For
more information, see Submit modification requests (p. 357).
6. (Optional) Terminate the original instance. For more information, see Terminate an
instance (p. 644).

Local Zones
A Local Zone is an extension of an AWS Region in geographic proximity to your users. Local Zones have
their own connections to the internet and support AWS Direct Connect, so that resources created in a
Local Zone can serve local users with low-latency communications. For more information, see AWS Local
Zones.

15
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Local Zones

A Local Zone is represented by a Region code followed by an identifier that indicates the location, for
example, us-west-2-lax-1a.

To use a Local Zone, you must first enable it. For more information, see the section called “Opt in to
Local Zones” (p. 17). Next, create a subnet in the Local Zone. Finally, launch any of the following
resources in the Local Zone subnet, so that your applications are close to your end users:

• Amazon EC2 instances

• Amazon EBS volumes
• Amazon ECS
• Amazon EKS

In addition to the above list, the following resources are available in the Los Angeles Local Zones.

• Amazon FSx file servers

• Elastic Load Balancing
• Amazon EMR
• Amazon ElastiCache
• Amazon Relational Database Service
• Dedicated Hosts

For information about the available Local Zones, see the section called “Available Regions” (p. 8).

Contents
• Describe your Local Zones (p. 16)
• Opt in to Local Zones (p. 17)
• Launch instances in a Local Zone (p. 17)

Describe your Local Zones

You can use the Amazon EC2 console or the command line interface to determine which Local Zones
are available for your account. For more information about these command line interfaces, see Access
Amazon EC2 (p. 3).

To find your Local Zones using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, view the options in the Region selector.
3. On the navigation pane, choose EC2 Dashboard.
4. The Local Zones are listed under Service health, Zone status.

To find your Local Zones using the AWS CLI

1. Use the describe-availability-zones command as follows to describe the Local Zones in the specified
Region.

aws ec2 describe-availability-zones --region region-name

2. Use the describe-availability-zones command as follows to describe the Local Zones regardless of
whether they are enabled.

16
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Local Zones

aws ec2 describe-availability-zones --all-availability-zones

To find your Local Zones using the AWS Tools for Windows PowerShell

Use the Get-EC2AvailabilityZone command as follows to describe the Local Zones in the specified
Region.

PS C:\> Get-EC2AvailabilityZone -Region region-name

Opt in to Local Zones

Before you can specify a Local Zone for a resource or service, you must opt in to Local Zones.

If you need to opt in to a Local Zone that is in preview (Boston, Houston, or Miami), request access by
filling in the AWS Local Zone Signup Form.

Consideration

Some AWS resources might not be available in all Regions. Make sure that you can create the resources
that you need in the desired Regions or Local Zones before launching an instance in a specific Local Zone.

To opt in to Local Zones using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the upper-left corner of the page, select New EC2 Experience. You cannot complete this task
using the old console experience.
3. From the Region selector in the navigation bar, select the Region for the Local Zone.
4. On the navigation pane, choose EC2 Dashboard.
5. In the upper-right corner of the page, choose Account attributes, Zones.
6. Choose Manage.
7. For Zone group, choose Enabled.
8. Choose Update zone group.

To opt in to Local Zones using the AWS CLI

• Use the modify-availability-zone-group command.

Launch instances in a Local Zone

When you launch an instance, you can specify a subnet which is in a Local Zone. You also allocate an
IP address from a network border group, which is a unique set of Availability Zones, Local Zones, or
Wavelength Zones from which AWS advertises IP addresses, for example, us-west-2-lax-1a.

You can allocate the following IP addresses from a network border group:

• Elastic IPv4 addresses that Amazon provides

• IPv6 Amazon-provided VPC addresses

To launch an instance in a Local Zone:

1. Enable Local Zones. For more information, see Opt in to Local Zones (p. 17).

17
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Wavelength Zones

2. Create a VPC in a Region that supports the Local Zone. For more information, see Creating a VPC in
the Amazon VPC User Guide.
3. Create a subnet. Select the Local Zone when you create the subnet. For more information, see
Creating a subnet in your VPC in the Amazon VPC User Guide.
4. Launch an instance, and select the subnet you created in the Local Zone. For more information, see
Launch your instance (p. 525).

Wavelength Zones
AWS Wavelength enables developers to build applications that deliver ultra-low latencies to mobile
devices and end users. Wavelength deploys standard AWS compute and storage services to the edge of
telecommunication carriers' 5G networks. Developers can extend a virtual private cloud (VPC) to one or
more Wavelength Zones, and then use AWS resources like Amazon EC2 instances to run applications that
require ultra-low latency and a connection to AWS services in the Region.

A Wavelength Zone is an isolated zone in the carrier location where the Wavelength infrastructure is
deployed. Wavelength Zones are tied to a Region. A Wavelength Zone is a logical extension of a Region,
and is managed by the control plane in the Region.

A Wavelength Zone is represented by a Region code followed by an identifier that indicates the
Wavelength Zone, for example, us-east-1-wl1-bos-wlz-1.

To use a Wavelength Zone, you must first opt in to the Zone. For more information, see the section called
“Enable Wavelength Zones” (p. 19). Next, create a subnet in the Wavelength Zone. Finally, launch your
resources in the Wavelength Zones subnet, so that your applications are closer to your end users.

Wavelength Zones are not available in every Region. For information about the Regions that support
Wavelength Zones, see Available Wavelength Zones in the AWS Wavelength Developer Guide.

Contents
• Describe your Wavelength Zones (p. 18)
• Enable Wavelength Zones (p. 19)
• Launch instances in a Wavelength Zone (p. 20)

Describe your Wavelength Zones

You can use the Amazon EC2 console or the command line interface to determine which Wavelength
Zones are available for your account. For more information about these command line interfaces, see
Access Amazon EC2 (p. 3).

To find your Wavelength Zones using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

18
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Wavelength Zones

2. From the navigation bar, view the options in the Region selector.
3. On the navigation pane, choose EC2 Dashboard.
4. The Wavelength Zones are listed under Service health, Zone status.

To find your Wavelength Zones using the AWS CLI

1. Use the describe-availability-zones command as follows to describe the Wavelength Zones within
the specified Region.

aws ec2 describe-availability-zones --region region-name

2. Use the describe-availability-zones command as follows to describe the Wavelength Zones

regardless of the opt-in status.

aws ec2 describe-availability-zones --all-availability-zones

To find your Wavelength Zone using the AWS Tools for Windows PowerShell

Use the Get-EC2AvailabilityZone command as follows to describe the Wavelength Zone within the
specified Region.

PS C:\> Get-EC2AvailabilityZone -Region region-name

Enable Wavelength Zones

Before you specify a Wavelength Zone for a resource or service, you must enable Wavelength Zones.

Considerations

• You must request access in order to use Wavelength Zones. For information about how to request
Wavelength Zone access, see AWS Wavelength.
• Some AWS resources are not available in all Regions. Make sure that you can create the resources
that you need in the desired Region or Wavelength Zone before launching an instance in a specific
Wavelength Zone.

To opt in to Wavelength Zone using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the upper-left corner of the page, select New EC2 Experience. You cannot complete this task
using the old console experience.
3. From the Region selector in the navigation bar, select the Region for the Wavelength Zone.
4. On the navigation pane, choose EC2 Dashboard.
5. In the upper-right corner of the page, choose Account attributes, Zones.
6. Under Wavelength Zones, turn on each Wavelength Zone.
7. Enable the Wavelength Zone.

To enable Wavelength Zones using the AWS CLI

Use the modify-availability-zone-group command.

19
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AWS Outposts

Launch instances in a Wavelength Zone

When you launch an instance, you can specify a subnet which is in a Wavelength Zone. You also allocate
a carrier IP address from a network border group, which is a unique set of Availability Zones, Local Zones,
or Wavelength Zones from which AWS advertises IP addresses, for example, us-east-1-wl1-bos-
wlz-1.

For information about how to launch an instance in a Wavelength Zone, see Get started with AWS
Wavelength in the AWS Wavelength Developer Guide.

AWS Outposts
AWS Outposts is a fully managed service that extends AWS infrastructure, services, APIs, and tools to
customer premises. By providing local access to AWS managed infrastructure, AWS Outposts enables
customers to build and run applications on premises using the same programming interfaces as in AWS
Regions, while using local compute and storage resources for lower latency and local data processing
needs.

An Outpost is a pool of AWS compute and storage capacity deployed at a customer site. AWS operates,
monitors, and manages this capacity as part of an AWS Region. You can create subnets on your Outpost
and specify them when you create AWS resources such as EC2 instances, EBS volumes, ECS clusters, and
RDS instances. Instances in Outpost subnets communicate with other instances in the AWS Region using
private IP addresses, all within the same VPC.

To begin using AWS Outposts, you must create an Outpost and order Outpost capacity. For more
information about Outposts configurations, see our catalog. After your Outpost equipment is installed,
the compute and storage capacity is available for you when you launch Amazon EC2 instances and create
Amazon EBS volumes on your Outpost.

Launch instances on an Outpost

You can launch EC2 instances in the Outpost subnet that you created. Security groups control inbound
and outbound traffic for instances in an Outpost subnet, as they do for instances in an Availability Zone
subnet. To connect to an EC2 instance in an Outpost subnet, you can specify a key pair when you launch
the instance, as you do for instances in an Availability Zone subnet.

The root volume must be 30 GB or smaller. You can specify data volumes in the block device mapping of
the AMI or the instance to provide additional storage. To trim unused blocks from the boot volume, see
How to Build Sparse EBS Volumes in the AWS Partner Network Blog.

We recommend that you increase the NVMe timeout for the root volume. For more information, see I/O
operation timeout (p. 1217).

For information about how to create an Outpost, see Get started with AWS Outposts in the AWS
Outposts User Guide.

Create a volume on an Outpost

You can create EBS volumes in the Outpost subnet that you created. When you create the volume,
specify the Amazon Resource Name (ARN) of the Outpost.

The following create-volume command creates an empty 50 GB volume on the specified Outpost.

aws ec2 create-volume --availability-zone us-east-2a --outpost-arn arn:aws:outposts:us-

east-2:123456789012:outpost/op-03e6fecad652a6138 --size 50

20
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Root device volume

You must detach an Outpost volume before you can modify it. For more information about detaching
volumes, see Detach an Amazon EBS volume from a Linux instance (p. 1123).

Amazon EC2 root device volume

When you launch an instance, the root device volume contains the image used to boot the instance. When
we introduced Amazon EC2, all AMIs were backed by Amazon EC2 instance store, which means the root
device for an instance launched from the AMI is an instance store volume created from a template stored
in Amazon S3. After we introduced Amazon EBS, we introduced AMIs that are backed by Amazon EBS.
This means that the root device for an instance launched from the AMI is an Amazon EBS volume created
from an Amazon EBS snapshot.

You can choose between AMIs backed by Amazon EC2 instance store and AMIs backed by Amazon EBS.
We recommend that you use AMIs backed by Amazon EBS, because they launch faster and use persistent
storage.
Important
Only the following instance types support an instance store volume as the root device: C3, D2,
G2, I2, M3, and R3.

For more information about the device names Amazon EC2 uses for your root volumes, see Name devices
on Linux instances (p. 1292).

Contents
• Root device storage concepts (p. 21)
• Choose an AMI by root device type (p. 23)
• Determine the root device type of your instance (p. 23)
• Change the root volume to persist (p. 24)
• Change the initial size of the root volume (p. 26)

Root device storage concepts

You can launch an instance from either an instance store-backed AMI or an Amazon EBS-backed AMI.
The description of an AMI includes which type of AMI it is; you'll see the root device referred to in some
places as either ebs (for Amazon EBS-backed) or instance store (for instance store-backed). This is
important because there are significant differences between what you can do with each type of AMI. For
more information about these differences, see Storage for the root device (p. 102).

Instance store-backed instances

Instances that use instance stores for the root device automatically have one or more instance store
volumes available, with one volume serving as the root device volume. When an instance is launched, the
image that is used to boot the instance is copied to the root volume. Note that you can optionally use
additional instance store volumes, depending on the instance type.

Any data on the instance store volumes persists as long as the instance is running, but this data is
deleted when the instance is terminated (instance store-backed instances do not support the Stop
action) or if it fails (such as if an underlying drive has issues).

21
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Root device storage concepts

After an instance store-backed instance fails or terminates, it cannot be restored. If you plan to use
Amazon EC2 instance store-backed instances, we highly recommend that you distribute the data on
your instance stores across multiple Availability Zones. You should also back up critical data from your
instance store volumes to persistent storage on a regular basis.

For more information, see Amazon EC2 instance store (p. 1268).

Amazon EBS-backed instances

Instances that use Amazon EBS for the root device automatically have an Amazon EBS volume attached.
When you launch an Amazon EBS-backed instance, we create an Amazon EBS volume for each Amazon
EBS snapshot referenced by the AMI you use. You can optionally use other Amazon EBS volumes or
instance store volumes, depending on the instance type.

An Amazon EBS-backed instance can be stopped and later restarted without affecting data stored in the
attached volumes. There are various instance– and volume-related tasks you can do when an Amazon
EBS-backed instance is in a stopped state. For example, you can modify the properties of the instance,
change its size, or update the kernel it is using, or you can attach your root volume to a different running
instance for debugging or any other purpose.

If an Amazon EBS-backed instance fails, you can restore your session by following one of these methods:

• Stop and then start again (try this method first).

• Automatically snapshot all relevant volumes and create a new AMI. For more information, see Create
an Amazon EBS-backed Linux AMI (p. 125).

22
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Choose an AMI by root device type

• Attach the volume to the new instance by following these steps:

1. Create a snapshot of the root volume.
2. Register a new AMI using the snapshot.
3. Launch a new instance from the new AMI.
4. Detach the remaining Amazon EBS volumes from the old instance.
5. Reattach the Amazon EBS volumes to the new instance.

For more information, see Amazon EBS volumes (p. 1082).

Choose an AMI by root device type

The AMI that you specify when you launch your instance determines the type of root device volume that
your instance has.

To choose an Amazon EBS-backed AMI using the console

1. Open the Amazon EC2 console.

2. In the navigation pane, choose AMIs.
3. From the filter lists, select the image type (such as Public images). In the search bar choose
Platform to select the operating system (such as Amazon Linux), and Root Device Type to select
EBS images.
4. (Optional) To get additional information to help you make your choice, choose the Show/Hide
Columns icon, update the columns to display, and choose Close.
5. Choose an AMI and write down its AMI ID.

To choose an instance store-backed AMI using the console

1. Open the Amazon EC2 console.

2. In the navigation pane, choose AMIs.
3. From the filter lists, select the image type (such as Public images). In the search bar, choose
Platform to select the operating system (such as Amazon Linux), and Root Device Type to select
Instance store.
4. (Optional) To get additional information to help you make your choice, choose the Show/Hide
Columns icon, update the columns to display, and choose Close.
5. Choose an AMI and write down its AMI ID.

To verify the type of the root device volume of an AMI using the command line
You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• describe-images (AWS CLI)

• Get-EC2Image (AWS Tools for Windows PowerShell)

Determine the root device type of your instance

New console

To determine the root device type of an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

23
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the root volume to persist

2. In the navigation pane, choose Instances, and select the instance.

3. On the Storage tab, under Root device details, check the value of Root device type as follows:

• If the value is EBS, this is an Amazon EBS-backed instance.

• If the value is INSTANCE-STORE, this is an instance store-backed instance.

Old console

To determine the root device type of an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and select the instance.
3. On the Description tab, check the value of Root device type as follows:

• If the value is ebs, this is an Amazon EBS-backed instance.

• If the value is instance store, this is an instance store-backed instance.

To determine the root device type of an instance using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• describe-instances (AWS CLI)

• Get-EC2Instance (AWS Tools for Windows PowerShell)

Change the root volume to persist

By default, the root volume for an AMI backed by Amazon EBS is deleted when the instance terminates.
You can change the default behavior to ensure that the volume persists after the instance terminates. To
change the default behavior, set the DeleteOnTermination attribute to false using a block device
mapping.

Topics
• Configure the root volume to persist during instance launch (p. 24)
• Configure the root volume to persist for an existing instance (p. 25)
• Confirm that a root volume is configured to persist (p. 26)

Configure the root volume to persist during instance launch

You can configure the root volume to persist when you launch an instance using the Amazon EC2 console
or the command line tools.

To configure the root volume to persist when you launch an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and then choose Launch instances.
3. On the Choose an Amazon Machine Image (AMI) page, select the AMI to use and choose Select.
4. Follow the wizard to complete the Choose an Instance Type and Configure Instance Details pages.
5. On the Add Storage page, deselect Delete On Termination for the root volume.
6. Complete the remaining wizard pages, and then choose Launch.

24
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the root volume to persist

To configure the root volume to persist when you launch an instance using the AWS CLI

Use the run-instances command and include a block device mapping that sets the
DeleteOnTermination attribute to false.

$ aws ec2 run-instances --block-device-mappings file://mapping.json ...other parameters...

Specify the following in mapping.json.

[
{
"DeviceName": "/dev/sda1",
"Ebs": {
"DeleteOnTermination": false
}
}
]

To configure the root volume to persist when you launch an instance using the Tools for Windows
PowerShell

Use the New-EC2Instance command and include a block device mapping that sets the
DeleteOnTermination attribute to false.

C:\> $ebs = New-Object Amazon.EC2.Model.EbsBlockDevice

C:\> $ebs.DeleteOnTermination = $false
C:\> $bdm = New-Object Amazon.EC2.Model.BlockDeviceMapping
C:\> $bdm.DeviceName = "dev/xvda"
C:\> $bdm.Ebs = $ebs
C:\> New-EC2Instance -ImageId ami-0abcdef1234567890 -BlockDeviceMapping $bdm ...other
parameters...

Configure the root volume to persist for an existing instance

You can configure the root volume to persist for a running instance using the command line tools only.

To configure the root volume to persist for an existing instance using the AWS CLI

Use the modify-instance-attribute command with a block device mapping that sets the
DeleteOnTermination attribute to false.

aws ec2 modify-instance-attribute --instance-id i-1234567890abcdef0 --block-device-mappings

"[{\"DeviceName\": \"/dev/xvda\",\"Ebs\":{\"DeleteOnTermination\":false}}]"

To configure the root volume to persist for an existing instance using the AWS Tools for Windows
PowerShell

Use the Edit-EC2InstanceAttribute command with a block device mapping that sets the
DeleteOnTermination attribute to false.

C:\> $ebs = New-Object Amazon.EC2.Model.EbsInstanceBlockDeviceSpecification

C:\> $ebs.DeleteOnTermination = $false
C:\> $bdm = New-Object Amazon.EC2.Model.InstanceBlockDeviceMappingSpecification
C:\> $bdm.DeviceName = "/dev/xvda"
C:\> $bdm.Ebs = $ebs
C:\> Edit-EC2InstanceAttribute -InstanceId i-1234567890abcdef0 -BlockDeviceMapping $bdm

25
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the initial size of the root volume

Confirm that a root volume is configured to persist

You can confirm that a root volume is configured to persist using the Amazon EC2 console or the
command line tools.

New console

To confirm that a root volume is configured to persist using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and then select the instance.
3. In the Storage tab, under Block devices, locate the entry for the root volume. If Delete on
termination is No, the volume is configured to persist.

Old console

To confirm that a root volume is configured to persist using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and then select the instance.
3. In the Description tab, choose the entry for Root device. If Delete on termination is False, the
volume is configured to persist.

To confirm that a root volume is configured to persist using the AWS CLI

Use the describe-instances command and verify that the DeleteOnTermination attribute in the
BlockDeviceMappings response element is set to false.

$ aws ec2 describe-instances --instance-id i-1234567890abcdef0

...
"BlockDeviceMappings": [
{
"DeviceName": "/dev/sda1",
"Ebs": {
"Status": "attached",
"DeleteOnTermination": false,
"VolumeId": "vol-1234567890abcdef0",
"AttachTime": "2013-07-19T02:42:39.000Z"
}
}
...

To confirm that a root volume is configured to persist using the AWS Tools for Windows PowerShell

Use the Get-EC2Instance and verify that the DeleteOnTermination attribute in the
BlockDeviceMappings response element is set to false.

C:\> (Get-EC2Instance -InstanceId i-i-1234567890abcdef0).Instances.BlockDeviceMappings.Ebs

Change the initial size of the root volume

By default, the size of the root volume is determined by the size of the snapshot. You can increase the
initial size of the root volume using the block device mapping of the instance as follows.

26
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the initial size of the root volume

1. Determine the device name of the root volume specified in the AMI, as described in View the EBS
volumes in an AMI block device mapping (p. 1298).
2. Confirm the size of the snapshot specified in the AMI block device mapping, as described in View
Amazon EBS snapshot information (p. 1138).
3. Override the size of the root volume using the instance block device mapping, as described in Update
the block device mapping when launching an instance (p. 1299), specifying a volume size that is
larger than the snapshot size.

For example, the following entry for the instance block device mapping increases the size of the root
volume, /dev/xvda, to 100 GiB. You can omit the snapshot ID in the instance block device mapping
because the snapshot ID is already specified in the AMI block device mapping.

{
"DeviceName": "/dev/xvda",
"Ebs": {
"VolumeSize": 100
}
}

For more information, see Block device mapping (p. 1294).

27
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Sign up for AWS

Set up to use Amazon EC2

Complete the tasks in this section to get set up for launching an Amazon EC2 instance for the first time:

1. Sign up for AWS (p. 28)

2. Create a key pair (p. 28)
3. Create a security group (p. 29)

When you are finished, you will be ready for the Amazon EC2 Getting started (p. 32) tutorial.

Sign up for AWS

When you sign up for Amazon Web Services (AWS), your AWS account is automatically signed up for all
services in AWS, including Amazon EC2. You are charged only for the services that you use.

With Amazon EC2, you pay only for what you use. If you are a new AWS customer, you can get started
with Amazon EC2 for free. For more information, see AWS Free Tier.

If you have an AWS account already, skip to the next task. If you don't have an AWS account, use the
following procedure to create one.

To create an AWS account

1. Open https://portal.aws.amazon.com/billing/signup.
2. Follow the online instructions.

Part of the sign-up procedure involves receiving a phone call and entering a verification code on the
phone keypad.

Create a key pair

AWS uses public-key cryptography to secure the login information for your instance. A Linux instance has
no password; you use a key pair to log in to your instance securely. You specify the name of the key pair
when you launch your instance, then provide the private key when you log in using SSH.

If you haven't created a key pair already, you can create one using the Amazon EC2 console. Note that
if you plan to launch instances in multiple Regions, you'll need to create a key pair in each Region. For
more information about Regions, see Regions and Zones (p. 7).

You can create a key pair using one of the following methods.

New console

To create your key pair

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Key Pairs.
3. Choose Create key pair.
4. For Name, enter a descriptive name for the key pair. Amazon EC2 associates the public key with
the name that you specify as the key name. A key name can include up to 255 ASCII characters.
It can’t include leading or trailing spaces.

28
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create a security group

5. For File format, choose the format in which to save the private key. To save the private key in
a format that can be used with OpenSSH, choose pem. To save the private key in a format that
can be used with PuTTY, choose ppk.
6. Choose Create key pair.
7. The private key file is automatically downloaded by your browser. The base file name is the
name you specified as the name of your key pair, and the file name extension is determined by
the file format you chose. Save the private key file in a safe place.
Important
This is the only chance for you to save the private key file.
8. If you will use an SSH client on a macOS or Linux computer to connect to your Linux instance,
use the following command to set the permissions of your private key file so that only you can
read it.

chmod 400 my-key-pair.pem

If you do not set these permissions, then you cannot connect to your instance using this key pair.
For more information, see Error: Unprotected private key file (p. 1336).

Old console

To create your key pair

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, under NETWORK & SECURITY, choose Key Pairs.
Note
The navigation pane is on the left side of the Amazon EC2 console. If you do not see
the pane, it might be minimized; choose the arrow to expand the pane.
3. Choose Create Key Pair.
4. For Key pair name, enter a name for the new key pair, and then choose Create. The name can
include up to 255 ASCII characters. It can’t include leading or trailing spaces.
5. The private key file is automatically downloaded by your browser. The base file name is the
name you specified as the name of your key pair, and the file name extension is .pem. Save the
private key file in a safe place.
Important
This is the only chance for you to save the private key file.
6. If you will use an SSH client on a macOS or Linux computer to connect to your Linux instance,
use the following command to set the permissions of your private key file so that only you can
read it.

chmod 400 my-key-pair.pem

If you do not set these permissions, then you cannot connect to your instance using this key pair.
For more information, see Error: Unprotected private key file (p. 1336).

For more information, see Amazon EC2 key pairs and Linux instances (p. 1044).

Create a security group

Security groups act as a firewall for associated instances, controlling both inbound and outbound traffic
at the instance level. You must add rules to a security group that enable you to connect to your instance

29
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create a security group

from your IP address using SSH. You can also add rules that allow inbound and outbound HTTP and
HTTPS access from anywhere.

Note that if you plan to launch instances in multiple Regions, you'll need to create a security group in
each Region. For more information about Regions, see Regions and Zones (p. 7).

Prerequisites

You'll need the public IPv4 address of your local computer. The security group editor in the Amazon
EC2 console can automatically detect the public IPv4 address for you. Alternatively, you can use the
search phrase "what is my IP address" in an Internet browser, or use the following service: Check IP. If
you are connecting through an Internet service provider (ISP) or from behind a firewall without a static IP
address, you need to find out the range of IP addresses used by client computers.

You can create a custom security group using one of the following methods.

New console

To create a security group with least privilege

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select a Region for the security group. Security groups are specific to a
Region, so you should select the same Region in which you created your key pair.
3. In the navigation pane, choose Security Groups.
4. Choose Create security group.
5. In the Basic details section, do the following:

a. Enter a name for the new security group and a description. Use a name that is easy for
you to remember, such as your user name, followed by _SG_, plus the Region name. For
example, me_SG_uswest2.
b. In the VPC list, select your default VPC for the Region.
6. In the Inbound rules section, create the following rules (choose Add rule for each new rule):

• Choose HTTP from the Type list, and make sure that Source is set to Anywhere (0.0.0.0/0).
• Choose HTTPS from the Type list, and make sure that Source is set to Anywhere
(0.0.0.0/0).
• Choose SSH from the Type list. In the Source box, choose My IP to automatically populate
the field with the public IPv4 address of your local computer. Alternatively, choose Custom
and specify the public IPv4 address of your computer or network in CIDR notation. To
specify an individual IP address in CIDR notation, add the routing suffix /32, for example,
203.0.113.25/32. If your company allocates addresses from a range, specify the entire
range, such as 203.0.113.0/24.
Warning
For security reasons, do not allow SSH access from all IPv4 addresses (0.0.0.0/0) to
your instance, except for testing purposes and only for a short time.
7. Choose Create security group.

Old console

To create a security group with least privilege

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Choose Create Security Group.

30
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create a security group

4. Enter a name for the new security group and a description. Use a name that is easy for you
to remember, such as your user name, followed by _SG_, plus the Region name. For example,
me_SG_uswest2.
5. In the VPC list, select your default VPC for the Region.
6. On the Inbound tab, create the following rules (choose Add rule for each new rule):

• Choose HTTP from the Type list, and make sure that Source is set to Anywhere (0.0.0.0/0).
• Choose HTTPS from the Type list, and make sure that Source is set to Anywhere
(0.0.0.0/0).
• Choose SSH from the Type list. In the Source box, choose My IP to automatically populate
the field with the public IPv4 address of your local computer. Alternatively, choose Custom
and specify the public IPv4 address of your computer or network in CIDR notation. To
specify an individual IP address in CIDR notation, add the routing suffix /32, for example,
203.0.113.25/32. If your company allocates addresses from a range, specify the entire
range, such as 203.0.113.0/24.
Warning
For security reasons, we don't recommend that you allow SSH access from all IPv4
addresses (0.0.0.0/0) to your instance, except for testing purposes and only for a
short time.
7. Choose Create.

Command line

To create a security group with least privilege

Use one of the following commands:

• create-security-group (AWS CLI)

• New-EC2SecurityGroup (AWS Tools for Windows PowerShell)

For more information, see Amazon EC2 security groups for Linux instances (p. 1058).

31
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Overview

Tutorial: Get started with Amazon

EC2 Linux instances
Use this tutorial to get started with Amazon Elastic Compute Cloud (Amazon EC2). You'll learn how
to launch, connect to, and use a Linux instance. An instance is a virtual server in the AWS cloud. With
Amazon EC2, you can set up and configure the operating system and applications that run on your
instance.

To get started with a Windows instance, see Getting started with Amazon EC2 Windows instances.

When you sign up for AWS, you can get started with Amazon EC2 using the AWS Free Tier. If you created
your AWS account less than 12 months ago, and have not already exceeded the free tier benefits for
Amazon EC2, it will not cost you anything to complete this tutorial, because we help you select options
that are within the free tier benefits. Otherwise, you'll incur the standard Amazon EC2 usage fees from
the time that you launch the instance until you terminate the instance (which is the final task of this
tutorial), even if it remains idle.

Contents
• Overview (p. 32)
• Prerequisites (p. 33)
• Step 1: Launch an instance (p. 33)
• Step 2: Connect to your instance (p. 34)
• Step 3: Clean up your instance (p. 34)
• Next steps (p. 35)

Overview
The instance is an Amazon EBS-backed instance (meaning that the root volume is an EBS volume).
You can either specify the Availability Zone in which your instance runs, or let Amazon EC2 select an
Availability Zone for you. When you launch your instance, you secure it by specifying a key pair and
security group. When you connect to your instance, you must specify the private key of the key pair that
you specified when launching your instance.

32
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Prerequisites

Tasks

To complete this tutorial, perform the following tasks:

1. Launch an instance (p. 33)

2. Connect to Your Instance (p. 34)
3. Clean up your instance (p. 34)

Related tutorials

• If you'd prefer to launch a Windows instance, see this tutorial in the Amazon EC2 User Guide for
Windows Instances: Getting started with Amazon EC2 Windows instances.
• If you'd prefer to use the command line, see this tutorial in the AWS Command Line Interface User
Guide: Using Amazon EC2 through the AWS CLI.

Prerequisites
Before you begin, be sure that you've completed the steps in Set up to use Amazon EC2 (p. 28).

Step 1: Launch an instance

You can launch a Linux instance using the AWS Management Console as described in the following
procedure. This tutorial is intended to help you launch your first instance quickly, so it doesn't cover all
possible options. For more information about the advanced options, see Launch an instance using the
Launch Instance Wizard (p. 526). For information about other ways to launch your instance, see Launch
your instance (p. 525).

To launch an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the console dashboard, choose Launch Instance.
3. The Choose an Amazon Machine Image (AMI) page displays a list of basic configurations, called
Amazon Machine Images (AMIs), that serve as templates for your instance. Select an HVM version of
Amazon Linux 2. Notice that these AMIs are marked "Free tier eligible."
4. On the Choose an Instance Type page, you can select the hardware configuration of your instance.
Select the t2.micro instance type, which is selected by default. The t2.micro instance type
is eligible for the free tier. In Regions where t2.micro is unavailable, you can use a t3.micro
instance under the free tier. For more information, see AWS Free Tier.
5. Choose Review and Launch to let the wizard complete the other configuration settings for you.
6. On the Review Instance Launch page, under Security Groups, you'll see that the wizard created and
selected a security group for you. You can use this security group, or alternatively you can select the
security group that you created when getting set up using the following steps:

a. Choose Edit security groups.

b. On the Configure Security Group page, ensure that Select an existing security group is
selected.
c. Select your security group from the list of existing security groups, and then choose Review and
Launch.
7. On the Review Instance Launch page, choose Launch.

33
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Connect to your instance

8. When prompted for a key pair, select Choose an existing key pair, then select the key pair that you
created when getting set up.
Warning
Don't select Proceed without a key pair. If you launch your instance without a key pair,
then you can't connect to it.

When you are ready, select the acknowledgement check box, and then choose Launch Instances.
9. A confirmation page lets you know that your instance is launching. Choose View Instances to close
the confirmation page and return to the console.
10. On the Instances screen, you can view the status of the launch. It takes a short time for an instance
to launch. When you launch an instance, its initial state is pending. After the instance starts, its
state changes to running and it receives a public DNS name. (If the Public DNS (IPv4) column is
hidden, choose Show/Hide Columns (the gear-shaped icon) in the top right corner of the page and
then select Public DNS (IPv4).)
11. It can take a few minutes for the instance to be ready so that you can connect to it. Check that your
instance has passed its status checks; you can view this information in the Status Checks column.

Step 2: Connect to your instance

There are several ways to connect to your Linux instance. For more information, see Connect to your
Linux instance (p. 598).
Important
You can't connect to your instance unless you launched it with a key pair for which you have the
.pem file and you launched it with a security group that allows SSH access from your computer.
If you can't connect to your instance, see Troubleshoot connecting to your instance (p. 1329) for
assistance.

Step 3: Clean up your instance

After you've finished with the instance that you created for this tutorial, you should clean up by
terminating the instance. If you want to do more with this instance before you clean up, see Next
steps (p. 35).
Important
Terminating an instance effectively deletes it; you can't reconnect to an instance after you've
terminated it.

If you launched an instance that is not within the AWS Free Tier, you'll stop incurring charges for that
instance as soon as the instance status changes to shutting down or terminated. If you'd like to keep
your instance for later, but not incur charges, you can stop the instance now and then start it again later.
For more information, see Stopping Instances.

To terminate your instance

1. In the navigation pane, choose Instances. In the list of instances, select the instance.
2. Choose Instance state, Terminate instance.
3. Choose Terminate when prompted for confirmation.

Amazon EC2 shuts down and terminates your instance. After your instance is terminated, it remains
visible on the console for a short while, and then the entry is automatically deleted. You cannot
remove the terminated instance from the console display yourself.

34
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Next steps

Next steps
After you start your instance, you might want to try some of the following exercises:

• Learn how to remotely manage your EC2 instance using Run Command. For more information, see
AWS Systems Manager Run Command in the AWS Systems Manager User Guide.
• Configure a CloudWatch alarm to notify you if your usage exceeds the Free Tier. For more information,
see Tracking your AWS Free Tier usage in the AWS Billing and Cost Management User Guide.
• Add an EBS volume. For more information, see Create an Amazon EBS volume (p. 1104) and Attach an
Amazon EBS volume to an instance (p. 1106).
• Install the LAMP stack. For more information, see Tutorial: Install a LAMP web server on Amazon Linux
2 (p. 38).

35
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Best practices for Amazon EC2

This list of practices will help you get the maximum benefit from Amazon EC2.

Security

• Manage access to AWS resources and APIs using identity federation, IAM users, and IAM roles. Establish
credential management policies and procedures for creating, distributing, rotating, and revoking AWS
access credentials. For more information, see IAM Best Practices in the IAM User Guide.
• Implement the least permissive rules for your security group. For more information, see Security group
rules (p. 1060).
• Regularly patch, update, and secure the operating system and applications on your instance. For more
information about updating Amazon Linux 2 or the Amazon Linux AMI, see Managing Software on
Your Linux Instance in the Amazon EC2 User Guide for Linux Instances.

Storage

• Understand the implications of the root device type for data persistence, backup, and recovery. For
more information, see Storage for the root device (p. 102).
• Use separate Amazon EBS volumes for the operating system versus your data. Ensure that the volume
with your data persists after instance termination. For more information, see Preserve Amazon EBS
volumes on instance termination (p. 646).
• Use the instance store available for your instance to store temporary data. Remember that the data
stored in instance store is deleted when you stop, hibernate, or terminate your instance. If you use
instance store for database storage, ensure that you have a cluster with a replication factor that
ensures fault tolerance.
• Encrypt EBS volumes and snapshots. For more information, see Amazon EBS encryption (p. 1176).

Resource management

• Use instance metadata and custom resource tags to track and identify your AWS resources. For
more information, see Instance metadata and user data (p. 704) and Tag your Amazon EC2
resources (p. 1311).
• View your current limits for Amazon EC2. Plan to request any limit increases in advance of the time
that you'll need them. For more information, see Amazon EC2 service quotas (p. 1323).

Backup and recovery

• Regularly back up your EBS volumes using Amazon EBS snapshots (p. 1126), and create an Amazon
Machine Image (AMI) (p. 99) from your instance to save the configuration as a template for
launching future instances.
• Deploy critical components of your application across multiple Availability Zones, and replicate your
data appropriately.
• Design your applications to handle dynamic IP addressing when your instance restarts. For more
information, see Amazon EC2 instance IP addressing (p. 809).
• Monitor and respond to events. For more information, see Monitor Amazon EC2 (p. 740).
• Ensure that you are prepared to handle failover. For a basic solution, you can manually attach a
network interface or Elastic IP address to a replacement instance. For more information, see Elastic
network interfaces (p. 839). For an automated solution, you can use Amazon EC2 Auto Scaling. For
more information, see the Amazon EC2 Auto Scaling User Guide.

36
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

• Regularly test the process of recovering your instances and Amazon EBS volumes if they fail.

Networking

• Set the time-to-live (TTL) value for your applications to 255, for IPv4 and IPv6. If you use a smaller
value, there is a risk that the TTL will expire while application traffic is in transit, causing reachability
issues for your instances.

37
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Install a LAMP server (Amazon Linux 2)

Tutorials for Amazon EC2 Instances

Running Linux
The following tutorials show you how to perform common tasks using EC2 instances running Linux. AWS
provides Amazon Linux 2 and the Amazon Linux AMI. For more information, see Amazon Linux 2 and
Amazon Linux AMI. For video tutorials, see AWS Instructional Videos and Labs.

Tutorials
• Tutorial: Install a LAMP web server on Amazon Linux 2 (p. 38)
• Tutorial: Install a LAMP web server with the Amazon Linux AMI (p. 48)
• Tutorial: Host a WordPress blog with Amazon Linux (p. 58)
• Tutorial: Configure SSL/TLS on Amazon Linux 2 (p. 67)
• Tutorial: Configure SSL/TLS on Amazon Linux (p. 82)
• Tutorial: Increase the availability of your application on Amazon EC2 (p. 95)

Tutorial: Install a LAMP web server on Amazon

Linux 2
The following procedures help you install an Apache web server with PHP and MariaDB (a community-
developed fork of MySQL) support on your Amazon Linux 2 instance (sometimes called a LAMP
web server or LAMP stack). You can use this server to host a static website or deploy a dynamic PHP
application that reads and writes information to a database.
Important
To set up a LAMP web server on Amazon Linux AMI, see Tutorial: Install a LAMP web server with
the Amazon Linux AMI (p. 48).
If you are trying to set up a LAMP web server on an Ubuntu or Red Hat Enterprise Linux
instance, this tutorial will not work for you. For more information about other distributions,
see their specific documentation. For information about LAMP web servers on Ubuntu, see the
Ubuntu community documentation ApacheMySQLPHP topic.

Option: Complete this tutorial using automation

To complete this tutorial using AWS Systems Manager Automation instead of the following tasks, run the
AWSDocs-InstallALAMPServer-AL2 Automation document.

Tasks
• Step 1: Prepare the LAMP server (p. 39)
• Step 2: Test your LAMP server (p. 42)
• Step 3: Secure the database server (p. 43)
• Step 4: (Optional) Install phpMyAdmin (p. 44)
• Troubleshoot (p. 47)

38
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 1: Prepare the LAMP server

• Related topics (p. 47)

Step 1: Prepare the LAMP server

Prerequisites

• This tutorial assumes that you have already launched a new instance using Amazon Linux 2, with a
public DNS name that is reachable from the internet. For more information, see Step 1: Launch an
instance (p. 33). You must also have configured your security group to allow SSH (port 22), HTTP (port
80), and HTTPS (port 443) connections. For more information about these prerequisites, see Authorize
inbound traffic for your Linux instances (p. 1041).
• The following procedure installs the latest PHP version available on Amazon Linux 2, currently PHP
7.2. If you plan to use PHP applications other than those described in this tutorial, you should check
their compatibility with PHP 7.2.

To prepare the LAMP server

1. Connect to your instance (p. 34).

2. To ensure that all of your software packages are up to date, perform a quick software update on
your instance. This process may take a few minutes, but it is important to make sure that you have
the latest security updates and bug fixes.

The -y option installs the updates without asking for confirmation. If you would like to examine the
updates before installing, you can omit this option.

[ec2-user ~]$ sudo yum update -y

3. Install the lamp-mariadb10.2-php7.2 and php7.2 Amazon Linux Extras repositories to get the
latest versions of the LAMP MariaDB and PHP packages for Amazon Linux 2.

[ec2-user ~]$ sudo amazon-linux-extras install -y lamp-mariadb10.2-php7.2 php7.2

If you receive an error stating sudo: amazon-linux-extras: command not found, then your
instance was not launched with an Amazon Linux 2 AMI (perhaps you are using the Amazon Linux
AMI instead). You can view your version of Amazon Linux using the following command.

cat /etc/system-release

To set up a LAMP web server on Amazon Linux AMI , see Tutorial: Install a LAMP web server with the
Amazon Linux AMI (p. 48).
4. Now that your instance is current, you can install the Apache web server, MariaDB, and PHP software
packages.

Use the yum install command to install multiple software packages and all related dependencies at
the same time.

[ec2-user ~]$ sudo yum install -y httpd mariadb-server

You can view the current versions of these packages using the following command:

yum info package_name

5. Start the Apache web server.

39
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 1: Prepare the LAMP server

[ec2-user ~]$ sudo systemctl start httpd

6. Use the systemctl command to configure the Apache web server to start at each system boot.

[ec2-user ~]$ sudo systemctl enable httpd

You can verify that httpd is on by running the following command:

[ec2-user ~]$ sudo systemctl is-enabled httpd

7. Add a security rule to allow inbound HTTP (port 80) connections to your instance if you have not
already done so. By default, a launch-wizard-N security group was set up for your instance during
initialization. This group contains a single rule to allow SSH connections.

a. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

b. Choose Instances and select your instance.
c. On the Security tab, view the inbound rules. You should see the following rule:

Port range Protocol Source

22 tcp 0.0.0.0/0

d. Choose the link for the security group. Using the procedures in Add rules to a security
group (p. 1066), add a new inbound security rule with the following values:

• Type: HTTP
• Protocol: TCP
• Port Range: 80
• Source: Custom
8. Test your web server. In a web browser, type the public DNS address (or the public IP address) of
your instance. If there is no content in /var/www/html, you should see the Apache test page. You
can get the public DNS for your instance using the Amazon EC2 console (check the Public DNS
column; if this column is hidden, choose Show/Hide Columns (the gear-shaped icon) and choose
Public DNS).

Verify that the security group for the instance contains a rule to allow HTTP traffic on port 80. For
more information, see Add rules to a security group (p. 1066).
Important
If you are not using Amazon Linux, you may also need to configure the firewall on your
instance to allow these connections. For more information about how to configure the
firewall, see the documentation for your specific distribution.

40
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 1: Prepare the LAMP server

Apache httpd serves files that are kept in a directory called the Apache document root. The Amazon
Linux Apache document root is /var/www/html, which by default is owned by root.

To allow the ec2-user account to manipulate files in this directory, you must modify the ownership and
permissions of the directory. There are many ways to accomplish this task. In this tutorial, you add ec2-
user to the apache group, to give the apache group ownership of the /var/www directory and assign
write permissions to the group.

To set file permissions

1. Add your user (in this case, ec2-user) to the apache group.

[ec2-user ~]$ sudo usermod -a -G apache ec2-user

2. Log out and then log back in again to pick up the new group, and then verify your membership.

a. Log out (use the exit command or close the terminal window):

[ec2-user ~]$ exit

b. To verify your membership in the apache group, reconnect to your instance, and then run the
following command:

[ec2-user ~]$ groups

ec2-user adm wheel apache systemd-journal

3. Change the group ownership of /var/www and its contents to the apache group.

[ec2-user ~]$ sudo chown -R ec2-user:apache /var/www

41
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Test your LAMP server

4. To add group write permissions and to set the group ID on future subdirectories, change the
directory permissions of /var/www and its subdirectories.

[ec2-user ~]$ sudo chmod 2775 /var/www && find /var/www -type d -exec sudo chmod 2775
{} \;

5. To add group write permissions, recursively change the file permissions of /var/www and its
subdirectories:

[ec2-user ~]$ find /var/www -type f -exec sudo chmod 0664 {} \;

Now, ec2-user (and any future members of the apache group) can add, delete, and edit files in the
Apache document root, enabling you to add content, such as a static website or a PHP application.

To secure your web server (Optional)

A web server running the HTTP protocol provides no transport security for the data that it sends or
receives. When you connect to an HTTP server using a web browser, the URLs that you visit, the content
of webpages that you receive, and the contents (including passwords) of any HTML forms that you
submit are all visible to eavesdroppers anywhere along the network pathway. The best practice for
securing your web server is to install support for HTTPS (HTTP Secure), which protects your data with
SSL/TLS encryption.

For information about enabling HTTPS on your server, see Tutorial: Configure SSL/TLS on Amazon Linux
2 (p. 67).

Step 2: Test your LAMP server

If your server is installed and running, and your file permissions are set correctly, your ec2-user account
should be able to create a PHP file in the /var/www/html directory that is available from the internet.

To test your LAMP server

1. Create a PHP file in the Apache document root.

[ec2-user ~]$ echo "<?php phpinfo(); ?>" > /var/www/html/phpinfo.php

If you get a "Permission denied" error when trying to run this command, try logging out and
logging back in again to pick up the proper group permissions that you configured in To set file
permissions (p. 41).
2. In a web browser, type the URL of the file that you just created. This URL is the public DNS address
of your instance followed by a forward slash and the file name. For example:

http://my.public.dns.amazonaws.com/phpinfo.php

You should see the PHP information page:

42
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 3: Secure the database server

If you do not see this page, verify that the /var/www/html/phpinfo.php file was created properly
in the previous step. You can also verify that all of the required packages were installed with the
following command.

[ec2-user ~]$ sudo yum list installed httpd mariadb-server php-mysqlnd

If any of the required packages are not listed in your output, install them with the sudo yum install
package command. Also verify that the php7.2 and lamp-mariadb10.2-php7.2 extras are
enabled in the output of the amazon-linux-extras command.
3. Delete the phpinfo.php file. Although this can be useful information, it should not be broadcast to
the internet for security reasons.

[ec2-user ~]$ rm /var/www/html/phpinfo.php

You should now have a fully functional LAMP web server. If you add content to the Apache document
root at /var/www/html, you should be able to view that content at the public DNS address for your
instance.

Step 3: Secure the database server

The default installation of the MariaDB server has several features that are great for testing
and development, but they should be disabled or removed for production servers. The
mysql_secure_installation command walks you through the process of setting a root password and
removing the insecure features from your installation. Even if you are not planning on using the MariaDB
server, we recommend performing this procedure.

43
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 4: (Optional) Install phpMyAdmin

To secure the MariaDB server

1. Start the MariaDB server.

[ec2-user ~]$ sudo systemctl start mariadb

2. Run mysql_secure_installation.

[ec2-user ~]$ sudo mysql_secure_installation

a. When prompted, type a password for the root account.

i. Type the current root password. By default, the root account does not have a password set.
Press Enter.
ii. Type Y to set a password, and type a secure password twice. For more information about
creating a secure password, see https://identitysafe.norton.com/password-generator/.
Make sure to store this password in a safe place.

Setting a root password for MariaDB is only the most basic measure for securing your
database. When you build or install a database-driven application, you typically create a
database service user for that application and avoid using the root account for anything but
database administration.
b. Type Y to remove the anonymous user accounts.
c. Type Y to disable the remote root login.
d. Type Y to remove the test database.
e. Type Y to reload the privilege tables and save your changes.
3. (Optional) If you do not plan to use the MariaDB server right away, stop it. You can restart it when
you need it again.

[ec2-user ~]$ sudo systemctl stop mariadb

4. (Optional) If you want the MariaDB server to start at every boot, type the following command.

[ec2-user ~]$ sudo systemctl enable mariadb

Step 4: (Optional) Install phpMyAdmin

phpMyAdmin is a web-based database management tool that you can use to view and edit the MySQL
databases on your EC2 instance. Follow the steps below to install and configure phpMyAdmin on your
Amazon Linux instance.
Important
We do not recommend using phpMyAdmin to access a LAMP server unless you have enabled
SSL/TLS in Apache; otherwise, your database administrator password and other data are
transmitted insecurely across the internet. For security recommendations from the developers,
see Securing your phpMyAdmin installation. For general information about securing a web
server on an EC2 instance, see Tutorial: Configure SSL/TLS on Amazon Linux 2 (p. 67).

To install phpMyAdmin

1. Install the required dependencies.

[ec2-user ~]$ sudo yum install php-mbstring -y

44
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 4: (Optional) Install phpMyAdmin

2. Restart Apache.

[ec2-user ~]$ sudo systemctl restart httpd

3. Restart php-fpm.

[ec2-user ~]$ sudo systemctl restart php-fpm

4. Navigate to the Apache document root at /var/www/html.

[ec2-user ~]$ cd /var/www/html

5. Select a source package for the latest phpMyAdmin release from https://www.phpmyadmin.net/
downloads. To download the file directly to your instance, copy the link and paste it into a wget
command, as in this example:

[ec2-user html]$ wget https://www.phpmyadmin.net/downloads/phpMyAdmin-latest-all-

languages.tar.gz

6. Create a phpMyAdmin folder and extract the package into it with the following command.

[ec2-user html]$ mkdir phpMyAdmin && tar -xvzf phpMyAdmin-latest-all-languages.tar.gz -

C phpMyAdmin --strip-components 1

7. Delete the phpMyAdmin-latest-all-languages.tar.gz tarball.

[ec2-user html]$ rm phpMyAdmin-latest-all-languages.tar.gz

8. (Optional) If the MySQL server is not running, start it now.

[ec2-user ~]$ sudo systemctl start mariadb

9. In a web browser, type the URL of your phpMyAdmin installation. This URL is the public DNS
address (or the public IP address) of your instance followed by a forward slash and the name of your
installation directory. For example:

http://my.public.dns.amazonaws.com/phpMyAdmin

You should see the phpMyAdmin login page:

45
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 4: (Optional) Install phpMyAdmin

10. Log in to your phpMyAdmin installation with the root user name and the MySQL root password you
created earlier.

Your installation must still be configured before you put it into service. To configure phpMyAdmin,
you can manually create a configuration file, use the setup script, or combine both approaches.

For information about using phpMyAdmin, see the phpMyAdmin User Guide.

46
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot

Troubleshoot
This section offers suggestions for resolving common problems you may encounter while setting up a
new LAMP server.

I can't connect to my server using a web browser

Perform the following checks to see if your Apache web server is running and accessible.

• Is the web server running?

You can verify that httpd is on by running the following command:

[ec2-user ~]$ sudo systemctl is-enabled httpd

If the httpd process is not running, repeat the steps described in To prepare the LAMP
server (p. 39).
• Is the firewall correctly configured?

Verify that the security group for the instance contains a rule to allow HTTP traffic on port 80. For
more information, see Add rules to a security group (p. 1066).

I can't connect to my server using HTTPS

Perform the following checks to see if your Apache web server is configured to support HTTPS.

• Is the web server correctly configured?

After you install Apache, the server is configured for HTTP traffic. To support HTTPS, enable TLS on
the server and install an SSL certificate. For information, see Tutorial: Configure SSL/TLS on Amazon
Linux 2 (p. 67).
• Is the firewall correctly configured?

Verify that the security group for the instance contains a rule to allow HTTPS traffic on port 443. For
more information, see Add rules to a security group (p. 1066).

Related topics
For more information about transferring files to your instance or installing a WordPress blog on your web
server, see the following documentation:

• Transfer files to your Linux instance using WinSCP (p. 617)

• Transfer files to Linux instances from Linux using SCP (p. 602)
• Tutorial: Host a WordPress blog with Amazon Linux (p. 58)

For more information about the commands and software used in this tutorial, see the following
webpages:

• Apache web server: http://httpd.apache.org/

• MariaDB database server: https://mariadb.org/
• PHP programming language: http://php.net/
• The chmod command: https://en.wikipedia.org/wiki/Chmod

47
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Install a LAMP server (Amazon Linux AMI)

• The chown command: https://en.wikipedia.org/wiki/Chown

For more information about registering a domain name for your web server, or transferring an existing
domain name to this host, see Creating and Migrating Domains and Subdomains to Amazon Route 53 in
the Amazon Route 53 Developer Guide.

Tutorial: Install a LAMP web server with the

Amazon Linux AMI
The following procedures help you install an Apache web server with PHP and MySQL support on your
Amazon Linux instance (sometimes called a LAMP web server or LAMP stack). You can use this server
to host a static website or deploy a dynamic PHP application that reads and writes information to a
database.
Important
To set up a LAMP web server on Amazon Linux 2, see Tutorial: Install a LAMP web server on
Amazon Linux 2 (p. 38).
If you are trying to set up a LAMP web server on an Ubuntu or Red Hat Enterprise Linux
instance, this tutorial will not work for you. For more information about other distributions,
see their specific documentation. For information about LAMP web servers on Ubuntu, see the
Ubuntu community documentation ApacheMySQLPHP topic.

Option: Complete this tutorial using automation

To complete this tutorial using AWS Systems Manager Automation instead of the following tasks, run the
AWSDocs-InstallALAMPServer-AL Automation document.

Tasks
• Step 1: Prepare the LAMP server (p. 48)
• Step 2: Test your Lamp server (p. 52)
• Step 3: Secure the database server (p. 53)
• Step 4: (Optional) Install phpMyAdmin (p. 54)
• Troubleshoot (p. 57)
• Related topics (p. 58)

Step 1: Prepare the LAMP server

Prerequisites

This tutorial assumes that you have already launched a new instance using the Amazon Linux AMI, with
a public DNS name that is reachable from the internet. For more information, see Step 1: Launch an
instance (p. 33). You must also have configured your security group to allow SSH (port 22), HTTP (port
80), and HTTPS (port 443) connections. For more information about these prerequisites, see Authorize
inbound traffic for your Linux instances (p. 1041).

To install and start the LAMP web server with the Amazon Linux AMI

1. Connect to your instance (p. 34).

2. To ensure that all of your software packages are up to date, perform a quick software update on
your instance. This process may take a few minutes, but it is important to make sure that you have
the latest security updates and bug fixes.

48
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 1: Prepare the LAMP server

The -y option installs the updates without asking for confirmation. If you would like to examine the
updates before installing, you can omit this option.

[ec2-user ~]$ sudo yum update -y

3. Now that your instance is current, you can install the Apache web server, MySQL, and PHP software
packages.
Important
Some applications may not be compatible with the following recommended software
environment. Before installing these packages, check whether your LAMP applications
are compatible with them. If there is a problem, you may need to install an alternative
environment. For more information, see The application software I want to run on my
server is incompatible with the installed PHP version or other software (p. 57)

Use the yum install command to install multiple software packages and all related dependencies at
the same time.

[ec2-user ~]$ sudo yum install -y httpd24 php72 mysql57-server php72-mysqlnd

If you receive the error No package package-name available, then your instance was not
launched with the Amazon Linux AMI (perhaps you are using Amazon Linux 2 instead). You can view
your version of Amazon Linux with the following command.

cat /etc/system-release

4. Start the Apache web server.

[ec2-user ~]$ sudo service httpd start

Starting httpd: [ OK ]

5. Use the chkconfig command to configure the Apache web server to start at each system boot.

[ec2-user ~]$ sudo chkconfig httpd on

The chkconfig command does not provide any confirmation message when you successfully use it to
enable a service.

You can verify that httpd is on by running the following command:

[ec2-user ~]$ chkconfig --list httpd

httpd 0:off 1:off 2:on 3:on 4:on 5:on 6:off

Here, httpd is on in runlevels 2, 3, 4, and 5 (which is what you want to see).

6. Add a security rule to allow inbound HTTP (port 80) connections to your instance if you have not
already done so. By default, a launch-wizard-N security group was set up for your instance during
initialization. This group contains a single rule to allow SSH connections.

a. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

b. Choose Instances and select your instance.
c. On the Security tab, view the inbound rules. You should see the following rule:

Port range Protocol Source

22 tcp 0.0.0.0/0

49
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 1: Prepare the LAMP server

d. Choose the link for the security group. Using the procedures in Add rules to a security
group (p. 1066), add a new inbound security rule with the following values:

• Type: HTTP
• Protocol: TCP
• Port Range: 80
• Source: Custom
7. Test your web server. In a web browser, type the public DNS address (or the public IP address) of
your instance. You can get the public DNS address for your instance using the Amazon EC2 console.
If there is no content in /var/www/html, you should see the Apache test page. When you add
content to the document root, your content appears at the public DNS address of your instance
instead of this test page.

Verify that the security group for the instance contains a rule to allow HTTP traffic on port 80. For
more information, see Add rules to a security group (p. 1066).

If you are not using Amazon Linux, you may also need to configure the firewall on your instance
to allow these connections. For more information about how to configure the firewall, see the
documentation for your specific distribution.

50
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 1: Prepare the LAMP server

Apache httpd serves files that are kept in a directory called the Apache document root. The Amazon
Linux Apache document root is /var/www/html, which by default is owned by root.

[ec2-user ~]$ ls -l /var/www

total 16
drwxr-xr-x 2 root root 4096 Jul 12 01:00 cgi-bin
drwxr-xr-x 3 root root 4096 Aug 7 00:02 error
drwxr-xr-x 2 root root 4096 Jan 6 2012 html
drwxr-xr-x 3 root root 4096 Aug 7 00:02 icons
drwxr-xr-x 2 root root 4096 Aug 7 21:17 noindex

To allow the ec2-user account to manipulate files in this directory, you must modify the ownership and
permissions of the directory. There are many ways to accomplish this task. In this tutorial, you add ec2-
user to the apache group, to give the apache group ownership of the /var/www directory and assign
write permissions to the group.

To set file permissions

1. Add your user (in this case, ec2-user) to the apache group.

[ec2-user ~]$ sudo usermod -a -G apache ec2-user

2. Log out and then log back in again to pick up the new group, and then verify your membership.

a. Log out (use the exit command or close the terminal window):

[ec2-user ~]$ exit

b. To verify your membership in the apache group, reconnect to your instance, and then run the
following command:

[ec2-user ~]$ groups

ec2-user wheel apache

3. Change the group ownership of /var/www and its contents to the apache group.

[ec2-user ~]$ sudo chown -R ec2-user:apache /var/www

4. To add group write permissions and to set the group ID on future subdirectories, change the
directory permissions of /var/www and its subdirectories.

[ec2-user ~]$ sudo chmod 2775 /var/www

[ec2-user ~]$ find /var/www -type d -exec sudo chmod 2775 {} \;

5. To add group write permissions, recursively change the file permissions of /var/www and its
subdirectories:

[ec2-user ~]$ find /var/www -type f -exec sudo chmod 0664 {} \;

Now, ec2-user (and any future members of the apache group) can add, delete, and edit files in the
Apache document root, enabling you to add content, such as a static website or a PHP application.

(Optional) Secure your web server

A web server running the HTTP protocol provides no transport security for the data that it sends or
receives. When you connect to an HTTP server using a web browser, the URLs that you visit, the content

51
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Test your Lamp server

of webpages that you receive, and the contents (including passwords) of any HTML forms that you
submit are all visible to eavesdroppers anywhere along the network pathway. The best practice for
securing your web server is to install support for HTTPS (HTTP Secure), which protects your data with
SSL/TLS encryption.

For information about enabling HTTPS on your server, see Tutorial: Configure SSL/TLS on Amazon
Linux (p. 82).

Step 2: Test your Lamp server

If your server is installed and running, and your file permissions are set correctly, your ec2-user account
should be able to create a PHP file in the /var/www/html directory that is available from the internet.

To test your LAMP web server

1. Create a PHP file in the Apache document root.

[ec2-user ~]$ echo "<?php phpinfo(); ?>" > /var/www/html/phpinfo.php

If you get a "Permission denied" error when trying to run this command, try logging out and logging
back in again to pick up the proper group permissions that you configured in Step 1: Prepare the
LAMP server (p. 48).
2. In a web browser, type the URL of the file that you just created. This URL is the public DNS address
of your instance followed by a forward slash and the file name. For example:

http://my.public.dns.amazonaws.com/phpinfo.php

You should see the PHP information page:

52
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 3: Secure the database server

If you do not see this page, verify that the /var/www/html/phpinfo.php file was created properly
in the previous step. You can also verify that all of the required packages were installed with the
following command. The package versions in the second column do not need to match this example
output.

[ec2-user ~]$ sudo yum list installed httpd24 php72 mysql57-server php72-mysqlnd
Loaded plugins: priorities, update-motd, upgrade-helper
Installed Packages
httpd24.x86_64 2.4.25-1.68.amzn1 @amzn-
updates
mysql56-server.x86_64 5.6.35-1.23.amzn1 @amzn-
updates
php70.x86_64 7.0.14-1.20.amzn1 @amzn-
updates
php70-mysqlnd.x86_64 7.0.14-1.20.amzn1 @amzn-
updates

If any of the required packages are not listed in your output, install them using the sudo yum install
package command.
3. Delete the phpinfo.php file. Although this can be useful information, it should not be broadcast to
the internet for security reasons.

[ec2-user ~]$ rm /var/www/html/phpinfo.php

Step 3: Secure the database server

The default installation of the MySQL server has several features that are great for testing
and development, but they should be disabled or removed for production servers. The
mysql_secure_installation command walks you through the process of setting a root password and
removing the insecure features from your installation. Even if you are not planning on using the MySQL
server, we recommend performing this procedure.

To secure the database server

1. Start the MySQL server.

[ec2-user ~]$ sudo service mysqld start

Initializing MySQL database:
...

PLEASE REMEMBER TO SET A PASSWORD FOR THE MySQL root USER !

...

Starting mysqld: [ OK ]

2. Run mysql_secure_installation.

[ec2-user ~]$ sudo mysql_secure_installation

a. When prompted, type a password for the root account.

i. Type the current root password. By default, the root account does not have a password set.
Press Enter.

53
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 4: (Optional) Install phpMyAdmin

ii. Type Y to set a password, and type a secure password twice. For more information about
creating a secure password, see https://identitysafe.norton.com/password-generator/.
Make sure to store this password in a safe place.

Setting a root password for MySQL is only the most basic measure for securing your
database. When you build or install a database-driven application, you typically create a
database service user for that application and avoid using the root account for anything but
database administration.
b. Type Y to remove the anonymous user accounts.
c. Type Y to disable the remote root login.
d. Type Y to remove the test database.
e. Type Y to reload the privilege tables and save your changes.
3. (Optional) If you do not plan to use the MySQL server right away, stop it. You can restart it when you
need it again.

[ec2-user ~]$ sudo service mysqld stop

Stopping mysqld: [ OK ]

4. (Optional) If you want the MySQL server to start at every boot, type the following command.

[ec2-user ~]$ sudo chkconfig mysqld on

You should now have a fully functional LAMP web server. If you add content to the Apache document
root at /var/www/html, you should be able to view that content at the public DNS address for your
instance.

Step 4: (Optional) Install phpMyAdmin

To install phpMyAdmin

phpMyAdmin is a web-based database management tool that you can use to view and edit the MySQL
databases on your EC2 instance. Follow the steps below to install and configure phpMyAdmin on your
Amazon Linux instance.
Important
We do not recommend using phpMyAdmin to access a LAMP server unless you have enabled
SSL/TLS in Apache; otherwise, your database administrator password and other data are
transmitted insecurely across the internet. For security recommendations from the developers,
see Securing your phpMyAdmin installation.
Note
The Amazon Linux package management system does not currently support the automatic
installation of phpMyAdmin in a PHP 7 environment. This tutorial describes how to install
phpMyAdmin manually.

1. Log in to your EC2 instance using SSH.

2. Install the required dependencies.

[ec2-user ~]$ sudo yum install php72-mbstring.x86_64 -y

3. Restart Apache.

[ec2-user ~]$ sudo service httpd restart

Stopping httpd: [ OK ]

54
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 4: (Optional) Install phpMyAdmin

Starting httpd: [ OK ]

4. Navigate to the Apache document root at /var/www/html.

[ec2-user ~]$ cd /var/www/html

[ec2-user html]$

5. Select a source package for the latest phpMyAdmin release from https://www.phpmyadmin.net/
downloads. To download the file directly to your instance, copy the link and paste it into a wget
command, as in this example:

[ec2-user html]$ wget https://www.phpmyadmin.net/downloads/phpMyAdmin-latest-all-

languages.tar.gz

6. Create a phpMyAdmin folder and extract the package into it using the following command.

[ec2-user html]$ mkdir phpMyAdmin && tar -xvzf phpMyAdmin-latest-all-languages.tar.gz -

C phpMyAdmin --strip-components 1

7. Delete the phpMyAdmin-latest-all-languages.tar.gz tarball.

[ec2-user html]$ rm phpMyAdmin-latest-all-languages.tar.gz

8. (Optional) If the MySQL server is not running, start it now.

[ec2-user ~]$ sudo service mysqld start

Starting mysqld: [ OK ]

9. In a web browser, type the URL of your phpMyAdmin installation. This URL is the public DNS
address (or the public IP address) of your instance followed by a forward slash and the name of your
installation directory. For example:

http://my.public.dns.amazonaws.com/phpMyAdmin

You should see the phpMyAdmin login page:

55
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 4: (Optional) Install phpMyAdmin

10. Log in to your phpMyAdmin installation with the root user name and the MySQL root password you
created earlier.

Your installation must still be configured before you put it into service. To configure phpMyAdmin,
you can manually create a configuration file, use the setup console, or combine both approaches.

For information about using phpMyAdmin, see the phpMyAdmin User Guide.

56
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot

Troubleshoot
This section offers suggestions for resolving common problems you may encounter while setting up a
new LAMP server.

I can't connect to my server using a web browser.

Perform the following checks to see if your Apache web server is running and accessible.

• Is the web server running?

You can verify that httpd is on by running the following command:

[ec2-user ~]$ chkconfig --list httpd

httpd 0:off 1:off 2:on 3:on 4:on 5:on 6:off

Here, httpd is on in runlevels 2, 3, 4, and 5 (which is what you want to see).

If the httpd process is not running, repeat the steps described in Step 1: Prepare the LAMP
server (p. 48).
• Is the firewall correctly configured?

Verify that the security group for the instance contains a rule to allow HTTP traffic on port 80. For
more information, see Add rules to a security group (p. 1066).

The application software I want to run on my server is

incompatible with the installed PHP version or other software
This tutorial recommends installing the most up-to-date versions of Apache HTTP Server, PHP, and
MySQL. Before installing an additional LAMP application, check its requirements to confirm that it is
compatible with your installed environment. If the latest version of PHP is not supported, it is possible
(and entirely safe) to downgrade to an earlier supported configuration. You can also install more than
one version of PHP in parallel, which solves certain compatibility problems with a minimum of effort. For
information about configuring a preference among multiple installed PHP versions, see Amazon Linux
AMI 2016.09 Release Notes.

How to downgrade

The well-tested previous version of this tutorial called for the following core LAMP packages:

• httpd24
• php56
• mysql55-server
• php56-mysqlnd

If you have already installed the latest packages as recommended at the start of this tutorial, you must
first uninstall these packages and other dependencies as follows:

[ec2-user ~]$ sudo yum remove -y httpd24 php72 mysql57-server php72-mysqlnd perl-DBD-
MySQL57

Next, install the replacement environment:

57
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Related topics

[ec2-user ~]$ sudo yum install -y httpd24 php56 mysql55-server php56-mysqlnd

If you decide later to upgrade to the recommended environment, you must first remove the customized
packages and dependencies:

[ec2-user ~]$ sudo yum remove -y httpd24 php56 mysql55-server php56-mysqlnd perl-DBD-
MySQL56

Now you can install the latest packages, as described earlier.

Related topics
For more information about transferring files to your instance or installing a WordPress blog on your web
server, see the following documentation:

• Transfer files to your Linux instance using WinSCP (p. 617)

• Transfer files to Linux instances from Linux using SCP (p. 602)
• Tutorial: Host a WordPress blog with Amazon Linux (p. 58)

For more information about the commands and software used in this tutorial, see the following
webpages:

• Apache web server: http://httpd.apache.org/

• MySQL database server: http://www.mysql.com/
• PHP programming language: http://php.net/
• The chmod command: https://en.wikipedia.org/wiki/Chmod
• The chown command: https://en.wikipedia.org/wiki/Chown

For more information about registering a domain name for your web server, or transferring an existing
domain name to this host, see Creating and Migrating Domains and Subdomains to Amazon Route 53 in
the Amazon Route 53 Developer Guide.

Tutorial: Host a WordPress blog with Amazon

Linux
The following procedures will help you install, configure, and secure a WordPress blog on your Amazon
Linux instance. This tutorial is a good introduction to using Amazon EC2 in that you have full control over
a web server that hosts your WordPress blog, which is not typical with a traditional hosting service.

You are responsible for updating the software packages and maintaining security patches for your server.
For a more automated WordPress installation that does not require direct interaction with the web server
configuration, the AWS CloudFormation service provides a WordPress template that can also get you
started quickly. For more information, see Get started in the AWS CloudFormation User Guide. If you'd
prefer to host your WordPress blog on a Windows instance, see Deploying a WordPress blog on your
Amazon EC2 Windows instance in the Amazon EC2 User Guide for Windows Instances. If you need a high-
availability solution with a decoupled database, see Deploying a high-availability WordPress website in
the AWS Elastic Beanstalk Developer Guide.
Important
These procedures are intended for use with Amazon Linux. For more information about other
distributions, see their specific documentation. Many steps in this tutorial do not work on

58
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Prerequisites

Ubuntu instances. For help installing WordPress on an Ubuntu instance, see WordPress in the
Ubuntu documentation.

Option: Complete this tutorial using automation

To complete this tutorial using AWS Systems Manager Automation instead of the following tasks, run
one of the following Automation documents: AWSDocs-HostingAWordPressBlog-AL (Amazon Linux) or
AWSDocs-HostingAWordPressBlog-AL2 (Amazon Linux 2).

Topics
• Prerequisites (p. 59)
• Install WordPress (p. 59)
• Next steps (p. 66)
• Help! My public DNS name changed and now my blog is broken (p. 66)

Prerequisites
This tutorial assumes that you have launched an Amazon Linux instance with a functional web server
with PHP and database (either MySQL or MariaDB) support by following all of the steps in Tutorial:
Install a LAMP web server with the Amazon Linux AMI (p. 48) for Amazon Linux AMI or Tutorial: Install
a LAMP web server on Amazon Linux 2 (p. 38) for Amazon Linux 2. This tutorial also has steps for
configuring a security group to allow HTTP and HTTPS traffic, as well as several steps to ensure that file
permissions are set properly for your web server. For information about adding rules to your security
group, see Add rules to a security group (p. 1066).

We strongly recommend that you associate an Elastic IP address (EIP) to the instance you are using
to host a WordPress blog. This prevents the public DNS address for your instance from changing and
breaking your installation. If you own a domain name and you want to use it for your blog, you can
update the DNS record for the domain name to point to your EIP address (for help with this, contact your
domain name registrar). You can have one EIP address associated with a running instance at no charge.
For more information, see Elastic IP addresses (p. 831).

If you don't already have a domain name for your blog, you can register a domain name with Route 53
and associate your instance's EIP address with your domain name. For more information, see Registering
domain names using Amazon Route 53 in the Amazon Route 53 Developer Guide.

Install WordPress
Connect to your instance, and download the WordPress installation package.

To download and unzip the WordPress installation package

1. Download the latest WordPress installation package with the wget command. The following
command should always download the latest release.

[ec2-user ~]$ wget https://wordpress.org/latest.tar.gz

2. Unzip and unarchive the installation package. The installation folder is unzipped to a folder called
wordpress.

[ec2-user ~]$ tar -xzf latest.tar.gz

59
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Install WordPress

To create a database user and database for your WordPress installation

Your WordPress installation needs to store information, such as blog posts and user comments, in a
database. This procedure helps you create your blog's database and a user that is authorized to read and
save information to it.

1. Start the database server.

• Amazon Linux 2

[ec2-user ~]$ sudo systemctl start mariadb

• Amazon Linux AMI

[ec2-user ~]$ sudo service mysqld start

2. Log in to the database server as the root user. Enter your database root password when prompted;
this may be different than your root system password, or it might even be empty if you have not
secured your database server.

If you have not secured your database server yet, it is important that you do so. For more
information, see To secure the MariaDB server (p. 44) (Amazon Linux 2) or To secure the database
server (p. 53) (Amazon Linux AMI).

[ec2-user ~]$ mysql -u root -p

3. Create a user and password for your MySQL database. Your WordPress installation uses these values
to communicate with your MySQL database. Enter the following command, substituting a unique
user name and password.

CREATE USER 'wordpress-user'@'localhost' IDENTIFIED BY 'your_strong_password';

Make sure that you create a strong password for your user. Do not use the single quote character
( ' ) in your password, because this will break the preceding command. For more information about
creating a secure password, go to http://www.pctools.com/guides/password/. Do not reuse an
existing password, and make sure to store this password in a safe place.
4. Create your database. Give your database a descriptive, meaningful name, such as wordpress-db.
Note
The punctuation marks surrounding the database name in the command below are called
backticks. The backtick (`) key is usually located above the Tab key on a standard keyboard.
Backticks are not always required, but they allow you to use otherwise illegal characters,
such as hyphens, in database names.

CREATE DATABASE `wordpress-db`;

5. Grant full privileges for your database to the WordPress user that you created earlier.

GRANT ALL PRIVILEGES ON `wordpress-db`.* TO "wordpress-user"@"localhost";

6. Flush the database privileges to pick up all of your changes.

FLUSH PRIVILEGES;

7. Exit the mysql client.

exit

60
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Install WordPress

To create and edit the wp-config.php file

The WordPress installation folder contains a sample configuration file called wp-config-sample.php.
In this procedure, you copy this file and edit it to fit your specific configuration.

1. Copy the wp-config-sample.php file to a file called wp-config.php. This creates a new
configuration file and keeps the original sample file intact as a backup.

[ec2-user ~]$ cp wordpress/wp-config-sample.php wordpress/wp-config.php

2. Edit the wp-config.php file with your favorite text editor (such as nano or vim) and enter values
for your installation. If you do not have a favorite text editor, nano is suitable for beginners.

[ec2-user ~]$ nano wordpress/wp-config.php

a. Find the line that defines DB_NAME and change database_name_here to the database
name that you created in Step 4 (p. 60) of To create a database user and database for your
WordPress installation (p. 60).

define('DB_NAME', 'wordpress-db');

b. Find the line that defines DB_USER and change username_here to the database user that
you created in Step 3 (p. 60) of To create a database user and database for your WordPress
installation (p. 60).

define('DB_USER', 'wordpress-user');

c. Find the line that defines DB_PASSWORD and change password_here to the strong password
that you created in Step 3 (p. 60) of To create a database user and database for your
WordPress installation (p. 60).

define('DB_PASSWORD', 'your_strong_password');

d. Find the section called Authentication Unique Keys and Salts. These KEY and SALT
values provide a layer of encryption to the browser cookies that WordPress users store on their
local machines. Basically, adding long, random values here makes your site more secure. Visit
https://api.wordpress.org/secret-key/1.1/salt/ to randomly generate a set of key values that
you can copy and paste into your wp-config.php file. To paste text into a PuTTY terminal,
place the cursor where you want to paste the text and right-click your mouse inside the PuTTY
terminal.

For more information about security keys, go to https://wordpress.org/support/article/editing-

wp-config-php/#security-keys.
Note
The values below are for example purposes only; do not use these values for your
installation.

define('AUTH_KEY', ' #U$$+[RXN8:b^-L 0(WU_+ c+WFkI~c]o]-bHw+)/

Aj[wTwSiZ<Qb[mghEXcRh-');
define('SECURE_AUTH_KEY', 'Zsz._P=l/|y.Lq)XjlkwS1y5NJ76E6EJ.AV0pCKZZB,*~*r ?6OP
$eJT@;+(ndLg');
define('LOGGED_IN_KEY', 'ju}qwre3V*+8f_zOWf?{LlGsQ]Ye@2Jh^,8x>)Y |;(^[Iw]Pi
+LG#A4R?7N`YB3');
define('NONCE_KEY', 'P(g62HeZxEes|LnI^i=H,[XwK9I&[2s|:?0N}VJM%?;v2v]v+;
+^9eXUahg@::Cj');
define('AUTH_SALT', 'C$DpB4Hj[JK:?{ql`sRVa:{:7yShy(9A@5wg+`JJVb1fk%_-
Bx*M4(qc[Qg%JT!h');

61
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Install WordPress

define('SECURE_AUTH_SALT', 'd!uRu#}+q#{f$Z?Z9uFPG.${+S{n~1M&%@~gL>U>NV<zpD-@2-
Es7Q1O-bp28EKv');
define('LOGGED_IN_SALT', ';j{00P*owZf)kVD+FVLn-~ >.|Y%Ug4#I^*LVd9QeZ^&XmK|e(76miC
+&W&+^0P/');
define('NONCE_SALT', '-97r*V/cgxLmp?Zy4zUU4r99QQ_rGs2LTd%P;|
_e1tS)8_B/,.6[=UK<J_y9?JWG');

e. Save the file and exit your text editor.

To install your WordPress files under the Apache document root

• Now that you've unzipped the installation folder, created a MySQL database and user, and
customized the WordPress configuration file, you are ready to copy your installation files to your
web server document root so you can run the installation script that completes your installation.
The location of these files depends on whether you want your WordPress blog to be available
at the actual root of your web server (for example, my.public.dns.amazonaws.com) or in a
subdirectory or folder under the root (for example, my.public.dns.amazonaws.com/blog).
• If you want WordPress to run at your document root, copy the contents of the wordpress
installation directory (but not the directory itself) as follows:

[ec2-user ~]$ cp -r wordpress/* /var/www/html/

• If you want WordPress to run in an alternative directory under the document root, first create
that directory, and then copy the files to it. In this example, WordPress will run from the
directory blog:

[ec2-user ~]$ mkdir /var/www/html/blog

[ec2-user ~]$ cp -r wordpress/* /var/www/html/blog/

Important
For security purposes, if you are not moving on to the next procedure immediately, stop the
Apache web server (httpd) now. After you move your installation under the Apache document
root, the WordPress installation script is unprotected and an attacker could gain access to your
blog if the Apache web server were running. To stop the Apache web server, enter the command
sudo service httpd stop. If you are moving on to the next procedure, you do not need to stop
the Apache web server.

To allow WordPress to use permalinks

WordPress permalinks need to use Apache .htaccess files to work properly, but this is not enabled by
default on Amazon Linux. Use this procedure to allow all overrides in the Apache document root.

1. Open the httpd.conf file with your favorite text editor (such as nano or vim). If you do not have a
favorite text editor, nano is suitable for beginners.

[ec2-user ~]$ sudo vim /etc/httpd/conf/httpd.conf

2. Find the section that starts with <Directory "/var/www/html">.

<Directory "/var/www/html">
#
# Possible values for the Options directive are "None", "All",
# or any combination of:
# Indexes Includes FollowSymLinks SymLinksifOwnerMatch ExecCGI MultiViews
#
# Note that "MultiViews" must be named *explicitly* --- "Options All"
# doesn't give it to you.

62
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Install WordPress

#
# The Options directive is both complicated and important. Please see
# http://httpd.apache.org/docs/2.4/mod/core.html#options
# for more information.
#
Options Indexes FollowSymLinks

#
# AllowOverride controls what directives may be placed in .htaccess files.
# It can be "All", "None", or any combination of the keywords:
# Options FileInfo AuthConfig Limit
#
AllowOverride None

#
# Controls who can get stuff from this server.
#
Require all granted
</Directory>

3. Change the AllowOverride None line in the above section to read AllowOverride All.
Note
There are multiple AllowOverride lines in this file; be sure you change the line in the
<Directory "/var/www/html"> section.

AllowOverride All

4. Save the file and exit your text editor.

To install the PHP graphics drawing library on Amazon Linux 2

The GD library for PHP enables you to modify images. Install this library if you need to crop the header
image for your blog. The version of phpMyAdmin that you install might require a specific minimum
version of this library (for example, version 7.2).

Use the following command to install the PHP graphics drawing library on Amazon Linux 2. For example,
if you installed php7.2 from amazon-linux-extras as part of installing the LAMP stack, this command
installs version 7.2 of the PHP graphics drawing library.

[ec2-user ~]$ sudo yum install php-gd

To verify the installed version, use the following command:

[ec2-user ~]$ sudo yum list installed | grep php-gd

The following is example output:

php-gd.x86_64 7.2.30-1.amzn2 @amzn2extra-php7.2

To install the PHP graphics drawing library on the Amazon Linux AMI

The GD library for PHP enables you to modify images. Install this library if you need to crop the header
image for your blog. The version of phpMyAdmin that you install might require a specific minimum
version of this library (for example, version 7.2).

To verify which versions are available, use the following command:

[ec2-user ~]$ yum list | grep php-gd

63
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Install WordPress

The following is an example line from the output for the PHP graphics drawing library (version 7.2):

php72-gd.x86_64 7.2.30-1.22.amzn1 amzn-updates

Use the following command to install a specific version of the PHP graphics drawing library (for example,
version 7.2) on the Amazon Linux AMI:

[ec2-user ~]$ sudo yum install php72-gd

To fix file permissions for the Apache web server

Some of the available features in WordPress require write access to the Apache document root (such
as uploading media though the Administration screens). If you have not already done so, apply the
following group memberships and permissions (as described in greater detail in the LAMP web server
tutorial (p. 48)).

1. Grant file ownership of /var/www and its contents to the apache user.

[ec2-user ~]$ sudo chown -R apache /var/www

2. Grant group ownership of /var/www and its contents to the apache group.

[ec2-user ~]$ sudo chgrp -R apache /var/www

3. Change the directory permissions of /var/www and its subdirectories to add group write
permissions and to set the group ID on future subdirectories.

[ec2-user ~]$ sudo chmod 2775 /var/www

[ec2-user ~]$ find /var/www -type d -exec sudo chmod 2775 {} \;

4. Recursively change the file permissions of /var/www and its subdirectories to add group write
permissions.

[ec2-user ~]$ find /var/www -type f -exec sudo chmod 0664 {} \;

5. Restart the Apache web server to pick up the new group and permissions.
• Amazon Linux 2

[ec2-user ~]$ sudo systemctl restart httpd

• Amazon Linux AMI

[ec2-user ~]$ sudo service httpd restart

To run the WordPress installation script with Amazon Linux 2

You are ready to install WordPress. The commands that you use depend on the operating system. The
commands in this procedure are for use with Amazon Linux 2. Use the procedure that follows this one
with Amazon Linux AMI.

1. Use the systemctl command to ensure that the httpd and database services start at every system
boot.

[ec2-user ~]$ sudo systemctl enable httpd && sudo systemctl enable mariadb

64
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Install WordPress

2. Verify that the database server is running.

[ec2-user ~]$ sudo systemctl status mariadb

If the database service is not running, start it.

[ec2-user ~]$ sudo systemctl start mariadb

3. Verify that your Apache web server (httpd) is running.

[ec2-user ~]$ sudo systemctl status httpd

If the httpd service is not running, start it.

[ec2-user ~]$ sudo systemctl start httpd

4. In a web browser, type the URL of your WordPress blog (either the public DNS address for your
instance, or that address followed by the blog folder). You should see the WordPress installation
script. Provide the information required by the WordPress installation. Choose Install WordPress to
complete the installation. For more information, see Step 5: Run the Install Script on the WordPress
website.

To run the WordPress installation script with Amazon Linux AMI

1. Use the chkconfig command to ensure that the httpd and database services start at every system
boot.

[ec2-user ~]$ sudo chkconfig httpd on && sudo chkconfig mysqld on

2. Verify that the database server is running.

[ec2-user ~]$ sudo service mysqld status

If the database service is not running, start it.

[ec2-user ~]$ sudo service mysqld start

3. Verify that your Apache web server (httpd) is running.

[ec2-user ~]$ sudo service httpd status

If the httpd service is not running, start it.

[ec2-user ~]$ sudo service httpd start

4. In a web browser, type the URL of your WordPress blog (either the public DNS address for your
instance, or that address followed by the blog folder). You should see the WordPress installation
script. Provide the information required by the WordPress installation. Choose Install WordPress to
complete the installation. For more information, see Step 5: Run the Install Script on the WordPress
website.

65
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Next steps

Next steps
After you have tested your WordPress blog, consider updating its configuration.

Use a custom domain name

If you have a domain name associated with your EC2 instance's EIP address, you can configure your blog
to use that name instead of the EC2 public DNS address. For more information, see Changing The Site
URL on the WordPress website.

Configure your blog

You can configure your blog to use different themes and plugins to offer a more personalized experience
for your readers. However, sometimes the installation process can backfire, causing you to lose your
entire blog. We strongly recommend that you create a backup Amazon Machine Image (AMI) of your
instance before attempting to install any themes or plugins so you can restore your blog if anything goes
wrong during installation. For more information, see Create your own AMI (p. 100).

Increase capacity

If your WordPress blog becomes popular and you need more compute power or storage, consider the
following steps:

• Expand the storage space on your instance. For more information, see Amazon EBS Elastic
Volumes (p. 1164).
• Move your MySQL database to Amazon RDS to take advantage of the service's ability to scale easily.
• Migrate to a larger instance type. For more information, see Change the instance type (p. 311).
• Add additional instances. For more information, see Tutorial: Increase the availability of your
application on Amazon EC2 (p. 95).

Learn more about WordPress

For information about WordPress, see the WordPress Codex help documentation at http://
codex.wordpress.org/. For more information about troubleshooting your installation, go to https://
wordpress.org/support/article/how-to-install-wordpress/#common-installation-problems. For
information about making your WordPress blog more secure, go to https://wordpress.org/support/
article/hardening-wordpress/. For information about keeping your WordPress blog up-to-date, go to
https://wordpress.org/support/article/updating-wordpress/.

Help! My public DNS name changed and now my blog

is broken
Your WordPress installation is automatically configured using the public DNS address for your EC2
instance. If you stop and restart the instance, the public DNS address changes (unless it is associated with
an Elastic IP address) and your blog will not work anymore because it references resources at an address
that no longer exists (or is assigned to another EC2 instance). A more detailed description of the problem
and several possible solutions are outlined in https://wordpress.org/support/article/changing-the-site-
url/.

If this has happened to your WordPress installation, you may be able to recover your blog with the
procedure below, which uses the wp-cli command line interface for WordPress.

To change your WordPress site URL with the wp-cli

1. Connect to your EC2 instance with SSH.

2. Note the old site URL and the new site URL for your instance. The old site URL is likely the public
DNS name for your EC2 instance when you installed WordPress. The new site URL is the current

66
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tutorial: Configure SSL/TLS on Amazon Linux 2

public DNS name for your EC2 instance. If you are not sure of your old site URL, you can use curl to
find it with the following command.

[ec2-user ~]$ curl localhost | grep wp-content

You should see references to your old public DNS name in the output, which will look like this (old
site URL in red):

<script type='text/javascript' src='http://ec2-52-8-139-223.us-

west-1.compute.amazonaws.com/wp-content/themes/twentyfifteen/js/functions.js?
ver=20150330'></script>

3. Download the wp-cli with the following command.

[ec2-user ~]$ curl -O https://raw.githubusercontent.com/wp-cli/builds/gh-pages/phar/wp-

cli.phar

4. Search and replace the old site URL in your WordPress installation with the following command.
Substitute the old and new site URLs for your EC2 instance and the path to your WordPress
installation (usually /var/www/html or /var/www/html/blog).

[ec2-user ~]$ php wp-cli.phar search-replace 'old_site_url' 'new_site_url' --path=/

path/to/wordpress/installation --skip-columns=guid

5. In a web browser, enter the new site URL of your WordPress blog to verify that the site is working
properly again. If it is not, see https://wordpress.org/support/article/changing-the-site-url/ and
https://wordpress.org/support/article/how-to-install-wordpress/#common-installation-problems
for more information.

Tutorial: Configure SSL/TLS on Amazon Linux 2

Secure Sockets Layer/Transport Layer Security (SSL/TLS) creates an encrypted channel between a web
server and web client that protects data in transit from being eavesdropped on. This tutorial explains
how to add support manually for SSL/TLS on an EC2 instance with Amazon Linux 2 and Apache web
server. This tutorial assumes that you are not using a load balancer. If you are using a load balancer, you
can choose to use the AWS Certificate Manager instead.

For historical reasons, web encryption is often referred to simply as SSL. While web browsers still
support SSL, its successor protocol TLS is less vulnerable to attack. Amazon Linux 2 disables server-
side support for all versions of SSL by default. Security standards bodies consider TLS 1.0 to be unsafe,
and both TLS 1.0 and TLS 1.1 are on track to be formally deprecated by the IETF. This tutorial contains
guidance based exclusively on enabling TLS 1.2. (A newer TLS 1.3 protocol exists, but it is not installed
by default on Amazon Linux 2.) For more information about the updated encryption standards, see RFC
7568 and RFC 8446.

This tutorial refers to modern web encryption simply as TLS.

Important
These procedures are intended for use with Amazon Linux 2. We also assume that you are
starting with a fresh Amazon EC2 instance. If you are trying to set up a LAMP web server on
an instance with a different distribution, or if you are reusing an older, existing instance, some
procedures in this tutorial might not work for you. For information about LAMP web servers on
Ubuntu, see the Ubuntu community documentation ApacheMySQLPHP. For information about
Red Hat Enterprise Linux, see the Customer Portal topic Web Servers.

Contents
• Prerequisites (p. 68)

67
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Prerequisites

• Step 1: Enable TLS on the server (p. 68)

• Step 2: Obtain a CA-signed certificate (p. 70)
• Step 3: Test and harden the security configuration (p. 75)
• Troubleshoot (p. 77)
• Certificate automation: Let's Encrypt with Certbot on Amazon Linux 2 (p. 78)

Prerequisites
Before you begin this tutorial, complete the following steps:

• Launch an EBS-backed Amazon Linux 2 instance. For more information, see Step 1: Launch an
instance (p. 33).
• Configure your security groups to allow your instance to accept connections on the following TCP
ports:
• SSH (port 22)
• HTTP (port 80)
• HTTPS (port 443)

For more information, see Authorize inbound traffic for your Linux instances (p. 1041).
• Install the Apache web server. For step-by-step instructions, see Tutorial: Install a LAMP Web Server on
Amazon Linux 2 (p. 38). Only the httpd package and its dependencies are needed, so you can ignore
the instructions involving PHP and MariaDB.
• To identify and authenticate websites, the TLS public key infrastructure (PKI) relies on the Domain
Name System (DNS). To use your EC2 instance to host a public website, you need to register a domain
name for your web server or transfer an existing domain name to your Amazon EC2 host. Numerous
third-party domain registration and DNS hosting services are available for this, or you can use Amazon
Route 53.

Step 1: Enable TLS on the server

This procedure takes you through the process of setting up TLS on Amazon Linux 2 with a self-signed
digital certificate.
Note
A self-signed certificate is acceptable for testing but not production. If you expose your self-
signed certificate to the internet, visitors to your site are greeted by security warnings.

To enable TLS on a server

1. Connect to your instance (p. 34) and confirm that Apache is running.

[ec2-user ~]$ sudo systemctl is-enabled httpd

If the returned value is not "enabled," start Apache and set it to start each time the system boots.

[ec2-user ~]$ sudo systemctl start httpd && sudo systemctl enable httpd

2. To ensure that all of your software packages are up to date, perform a quick software update on
your instance. This process may take a few minutes, but it is important to make sure that you have
the latest security updates and bug fixes.

68
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 1: Enable TLS on the server

Note
The -y option installs the updates without asking for confirmation. If you would like to
examine the updates before installing, you can omit this option.

[ec2-user ~]$ sudo yum update -y

3. Now that your instance is current, add TLS support by installing the Apache module mod_ssl.

[ec2-user ~]$ sudo yum install -y mod_ssl

Your instance now has the following files that you use to configure your secure server and create a
certificate for testing:

• /etc/httpd/conf.d/ssl.conf

The configuration file for mod_ssl. It contains directives telling Apache where to find encryption
keys and certificates, the TLS protocol versions to allow, and the encryption ciphers to accept.
• /etc/pki/tls/certs/make-dummy-cert

A script to generate a self-signed X.509 certificate and private key for your server host. This
certificate is useful for testing that Apache is properly set up to use TLS. Because it offers no proof
of identity, it should not be used in production. If used in production, it triggers warnings in Web
browsers.
4. Run the script to generate a self-signed dummy certificate and key for testing.

[ec2-user ~]$ cd /etc/pki/tls/certs

sudo ./make-dummy-cert localhost.crt

This generates a new file localhost.crt in the /etc/pki/tls/certs/ directory. The specified
file name matches the default that is assigned in the SSLCertificateFile directive in /etc/httpd/
conf.d/ssl.conf.

This file contains both a self-signed certificate and the certificate's private key. Apache requires the
certificate and key to be in PEM format, which consists of Base64-encoded ASCII characters framed
by "BEGIN" and "END" lines, as in the following abbreviated example.

-----BEGIN PRIVATE KEY-----

MIIEvgIBADANBgkqhkiG9w0BAQEFAASCBKgwggSkAgEAAoIBAQD2KKx/8Zk94m1q
3gQMZF9ZN66Ls19+3tHAgQ5Fpo9KJDhzLjOOCI8u1PTcGmAah5kEitCEc0wzmNeo
BCl0wYR6G0rGaKtK9Dn7CuIjvubtUysVyQoMVPQ97ldeakHWeRMiEJFXg6kZZ0vr
GvwnKoMh3DlK44D9dX7IDua2PlYx5+eroA+1Lqf32ZSaAO0bBIMIYTHigwbHMZoT
...
56tE7THvH7vOEf4/iUOsIrEzaMaJ0mqkmY1A70qQGQKBgBF3H1qNRNHuyMcPODFs
27hDzPDinrquSEvoZIggkDMlh2irTiipJ/GhkvTpoQlv0fK/VXw8vSgeaBuhwJvS
LXU9HvYq0U6O4FgD3nAyB9hI0BE13r1HjUvbjT7moH+RhnNz6eqqdscCS09VtRAo
4QQvAqOa8UheYeoXLdWcHaLP
-----END PRIVATE KEY-----

-----BEGIN CERTIFICATE-----
MIIEazCCA1OgAwIBAgICWxQwDQYJKoZIhvcNAQELBQAwgbExCzAJBgNVBAYTAi0t
MRIwEAYDVQQIDAlTb21lU3RhdGUxETAPBgNVBAcMCFNvbWVDaXR5MRkwFwYDVQQK
DBBTb21lT3JnYW5pemF0aW9uMR8wHQYDVQQLDBZTb21lT3JnYW5pemF0aW9uYWxV
bml0MRkwFwYDVQQDDBBpcC0xNzItMzEtMjAtMjM2MSQwIgYJKoZIhvcNAQkBFhVy
...
z5rRUE/XzxRLBZOoWZpNWTXJkQ3uFYH6s/sBwtHpKKZMzOvDedREjNKAvk4ws6F0
CuIjvubtUysVyQoMVPQ97ldeakHWeRMiEJFXg6kZZ0vrGvwnKoMh3DlK44D9dlU3
WanXWehT6FiSZvB4sTEXXJN2jdw8g+sHGnZ8zCOsclknYhHrCVD2vnBlZJKSZvak
3ZazhBxtQSukFMOnWPP2a0DMMFGYUHOd0BQE8sBJxg==

69
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Obtain a CA-signed certificate

-----END CERTIFICATE-----

The file names and extensions are a convenience and have no effect on function. For example, you
can call a certificate cert.crt, cert.pem, or any other file name, so long as the related directive in
the ssl.conf file uses the same name.
Note
When you replace the default TLS files with your own customized files, be sure that they are
in PEM format.
5. Open the /etc/httpd/conf.d/ssl.conf file and comment out the following line, because the
self-signed dummy certificate also contains the key. If you do not comment out this line before you
complete the next step, the Apache service fails to start.

SSLCertificateKeyFile /etc/pki/tls/private/localhost.key

6. Restart Apache.

[ec2-user ~]$ sudo systemctl restart httpd

Note
Make sure that TCP port 443 is accessible on your EC2 instance, as previously described.
7. Your Apache web server should now support HTTPS (secure HTTP) over port 443. Test it by entering
the IP address or fully qualified domain name of your EC2 instance into a browser URL bar with the
prefix https://.

Because you are connecting to a site with a self-signed, untrusted host certificate, your browser may
display a series of security warnings. Override the warnings and proceed to the site.

If the default Apache test page opens, it means that you have successfully configured TLS on your
server. All data passing between the browser and server is now encrypted.
Note
To prevent site visitors from encountering warning screens, you must obtain a trusted, CA-
signed certificate that not only encrypts, but also publicly authenticates you as the owner
of the site.

Step 2: Obtain a CA-signed certificate

You can use the following process to obtain a CA-signed certificate:

• Generate a certificate signing request (CSR) from a private key

• Submit the CSR to a certificate authority (CA)
• Obtain a signed host certificate
• Configure Apache to use the certificate

A self-signed TLS X.509 host certificate is cryptologically identical to a CA-signed certificate. The
difference is social, not mathematical. A CA promises, at a minimum, to validate a domain's ownership
before issuing a certificate to an applicant. Each web browser contains a list of CAs trusted by the
browser vendor to do this. An X.509 certificate consists primarily of a public key that corresponds to
your private server key, and a signature by the CA that is cryptographically tied to the public key. When a
browser connects to a web server over HTTPS, the server presents a certificate for the browser to check
against its list of trusted CAs. If the signer is on the list, or accessible through a chain of trust consisting
of other trusted signers, the browser negotiates a fast encrypted data channel with the server and loads
the page.

70
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Obtain a CA-signed certificate

Important
Certificates generally cost money because of the labor involved in validating the requests, so it
pays to shop around. A list of well-known CAs can be found at dmoztools.net. A few CAs offer
basic-level certificates free of charge. The most notable of these CAs is the Let's Encrypt project,
which also supports the automation of the certificate creation and renewal process. For more
information about using Let's Encrypt as your CA, see Certificate automation: Let's Encrypt with
Certbot on Amazon Linux 2 (p. 78).

Underlying the host certificate is the key. As of 2019, government and industry groups recommend using
a minimum key (modulus) size of 2048 bits for RSA keys intended to protect documents, through 2030.
The default modulus size generated by OpenSSL in Amazon Linux 2 is 2048 bits, which is suitable for use
in a CA-signed certificate. In the following procedure, an optional step provided for those who want a
customized key, for example, one with a larger modulus or using a different encryption algorithm.

These instructions for acquiring a CA-signed host certificate do not work unless you own a registered and
hosted DNS domain.

To obtain a CA-signed certificate

1. Connect to your instance (p. 34) and navigate to /etc/pki/tls/private/. This is the directory where
you store the server's private key for TLS. If you prefer to use an existing host key to generate the
CSR, skip to Step 3.
2. (Optional) Generate a new private key. Here are some examples of key configurations. Any of the
resulting keys works with your web server, but they vary in the degree and type of security that they
implement.

• Example 1: Create a default RSA host key. The resulting file, custom.key, is a 2048-bit RSA
private key.

[ec2-user ~]$ sudo openssl genrsa -out custom.key

• Example 2: Create a stronger RSA key with a bigger modulus. The resulting file, custom.key, is a
4096-bit RSA private key.

[ec2-user ~]$ sudo openssl genrsa -out custom.key 4096

• Example 3: Create a 4096-bit encrypted RSA key with password protection. The resulting file,
custom.key, is a 4096-bit RSA private key encrypted with the AES-128 cipher.
Important
Encrypting the key provides greater security, but because an encrypted key requires a
password, services depending on it cannot be auto-started. Each time you use this key,
you must supply the password (in the preceding example, "abcde12345") over an SSH
connection.

[ec2-user ~]$ sudo openssl genrsa -aes128 -passout pass:abcde12345 -out custom.key
4096

• Example 4: Create a key using a non-RSA cipher. RSA cryptography can be relatively slow because
of the size of its public keys, which are based on the product of two large prime numbers.
However, it is possible to create keys for TLS that use non-RSA ciphers. Keys based on the
mathematics of elliptic curves are smaller and computationally faster when delivering an
equivalent level of security.

[ec2-user ~]$ sudo openssl ecparam -name prime256v1 -out custom.key -genkey

The result is a 256-bit elliptic curve private key using prime256v1, a "named curve" that OpenSSL
supports. Its cryptographic strength is slightly greater than a 2048-bit RSA key, according to NIST.

71
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Obtain a CA-signed certificate

Note
Not all CAs provide the same level of support for elliptic-curve-based keys as for RSA
keys.

Make sure that the new private key has highly restrictive ownership and permissions (owner=root,
group=root, read/write for owner only). The commands would be as shown in the following
example.

[ec2-user ~]$ sudo chown root:root custom.key

[ec2-user ~]$ sudo chmod 600 custom.key
[ec2-user ~]$ ls -al custom.key

The preceding commands yield the following result.

-rw------- root root custom.key

After you have created and configured a satisfactory key, you can create a CSR.
3. Create a CSR using your preferred key. The following example uses custom.key.

[ec2-user ~]$ sudo openssl req -new -key custom.key -out csr.pem

OpenSSL opens a dialog and prompts you for the information shown in the following table. All of
the fields except Common Name are optional for a basic, domain-validated host certificate.

Name Description Example

Country Name The two-letter ISO abbreviation for your US (=United States)
country.

State or The name of the state or province where your Washington

Province organization is located. This name cannot be
Name abbreviated.

Locality Name The location of your organization, such as a city. Seattle

Organization The full legal name of your organization. Do not Example Corporation
Name abbreviate your organization name.

Organizational Additional organizational information, if any. Example Dept

Unit Name

Common This value must exactly match the web www.example.com

Name address that you expect users to enter into
a browser. Usually, this means a domain
name with a prefixed hostname or alias in
the form www.example.com. In testing
with a self-signed certificate and no DNS
resolution, the common name may consist
of the hostname alone. CAs also offer more
expensive certificates that accept wild-card
names such as *.example.com.

Email Address The server administrator's email address. [email protected]

72
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Obtain a CA-signed certificate

Finally, OpenSSL prompts you for an optional challenge password. This password applies only to the
CSR and to transactions between you and your CA, so follow the CA's recommendations about this
and the other optional field, optional company name. The CSR challenge password has no effect on
server operation.

The resulting file csr.pem contains your public key, your digital signature of your public key, and
the metadata that you entered.
4. Submit the CSR to a CA. This usually consists of opening your CSR file in a text editor and copying
the contents into a web form. At this time, you may be asked to supply one or more subject
alternate names (SANs) to be placed on the certificate. If www.example.com is the common name,
then example.com would be a good SAN, and vice versa. A visitor to your site entering either of
these names would see an error-free connection. If your CA web form allows it, include the common
name in the list of SANs. Some CAs include it automatically.

After your request has been approved, you receive a new host certificate signed by the CA. You
might also be instructed to download an intermediate certificate file that contains additional
certificates needed to complete the CA's chain of trust.
Note
Your CA might send you files in multiple formats intended for various purposes. For this
tutorial, you should only use a certificate file in PEM format, which is usually (but not
always) marked with a .pem or .crt file extension. If you are uncertain which file to use,
open the files with a text editor and find the one containing one or more blocks beginning
with the following line.

- - - - -BEGIN CERTIFICATE - - - - -

The file should also end with the following line.

- - - -END CERTIFICATE - - - - -

You can also test the file at the command line as shown in the following.

[ec2-user certs]$ openssl x509 -in certificate.crt -text

Verify that these lines appear in the file. Do not use files ending with .p7b, .p7c, or similar
file extensions.
5. Place the new CA-signed certificate and any intermediate certificates in the /etc/pki/tls/certs
directory.
Note
There are several ways to upload your new certificate to your EC2 instance, but the most
straightforward and informative way is to open a text editor (for example, vi, nano,
or notepad) on both your local computer and your instance, and then copy and paste
the file contents between them. You need root [sudo] permissions when performing
these operations on the EC2 instance. This way, you can see immediately if there are any
permission or path problems. Be careful, however, not to add any additional lines while
copying the contents, or to change them in any way.

From inside the /etc/pki/tls/certs directory, check that the file ownership, group, and
permission settings match the highly restrictive Amazon Linux 2 defaults (owner=root, group=root,
read/write for owner only). The following example shows the commands to use.

[ec2-user certs]$ sudo chown root:root custom.crt

[ec2-user certs]$ sudo chmod 600 custom.crt

73
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Obtain a CA-signed certificate

[ec2-user certs]$ ls -al custom.crt

These commands should yield the following result.

-rw------- root root custom.crt

The permissions for the intermediate certificate file are less stringent (owner=root, group=root,
owner can write, group can read, world can read). The following example shows the commands to
use.

[ec2-user certs]$ sudo chown root:root intermediate.crt

[ec2-user certs]$ sudo chmod 644 intermediate.crt
[ec2-user certs]$ ls -al intermediate.crt

These commands should yield the following result.

-rw-r--r-- root root intermediate.crt

6. Place the private key that you used to create the CSR in the /etc/pki/tls/private/ directory.
Note
There are several ways to upload your custom key to your EC2 instance, but the most
straightforward and informative way is to open a text editor (for example, vi, nano,
or notepad) on both your local computer and your instance, and then copy and paste
the file contents between them. You need root [sudo] permissions when performing
these operations on the EC2 instance. This way, you can see immediately if there are any
permission or path problems. Be careful, however, not to add any additional lines while
copying the contents, or to change them in any way.

From inside the /etc/pki/tls/private directory, use the following commands to verify that the
file ownership, group, and permission settings match the highly restrictive Amazon Linux 2 defaults
(owner=root, group=root, read/write for owner only).

[ec2-user private]$ sudo chown root:root custom.key

[ec2-user private]$ sudo chmod 600 custom.key
[ec2-user private]$ ls -al custom.key

These commands should yield the following result.

-rw------- root root custom.key

7. Edit /etc/httpd/conf.d/ssl.conf to reflect your new certificate and key files.

a. Provide the path and file name of the CA-signed host certificate in Apache's
SSLCertificateFile directive:

SSLCertificateFile /etc/pki/tls/certs/custom.crt

b. If you received an intermediate certificate file (intermediate.crt in this example), provide its
path and file name using Apache's SSLCACertificateFile directive:

SSLCACertificateFile /etc/pki/tls/certs/intermediate.crt

74
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 3: Test and harden the security configuration

Note
Some CAs combine the host certificate and the intermediate certificates in a single file,
making the SSLCACertificateFile directive unnecessary. Consult the instructions
provided by your CA.
c. Provide the path and file name of the private key (custom.key in this example) in Apache's
SSLCertificateKeyFile directive:

SSLCertificateKeyFile /etc/pki/tls/private/custom.key

8. Save /etc/httpd/conf.d/ssl.conf and restart Apache.

[ec2-user ~]$ sudo systemctl restart httpd

9. Test your server by entering your domain name into a browser URL bar with the prefix https://.
Your browser should load the test page over HTTPS without generating errors.

Step 3: Test and harden the security configuration

After your TLS is operational and exposed to the public, you should test how secure it really is. This is
easy to do using online services such as Qualys SSL Labs, which performs a free and thorough analysis
of your security setup. Based on the results, you may decide to harden the default security configuration
by controlling which protocols you accept, which ciphers you prefer, and which you exclude. For more
information, see how Qualys formulates its scores.
Important
Real-world testing is crucial to the security of your server. Small configuration errors may lead
to serious security breaches and loss of data. Because recommended security practices change
constantly in response to research and emerging threats, periodic security audits are essential to
good server administration.

On the Qualys SSL Labs site, enter the fully qualified domain name of your server, in the form
www.example.com. After about two minutes, you receive a grade (from A to F) for your site and a
detailed breakdown of the findings. The following table summarizes the report for a domain with
settings identical to the default Apache configuration on Amazon Linux 2, and with a default Certbot
certificate.

Overall rating B

Certificate 100%

Protocol support 95%

Key exchange 70%

Cipher strength 90%

Though the overview shows that the configuration is mostly sound, the detailed report flags several
potential problems, listed here in order of severity:

✗ The RC4 cipher is supported for use by certain older browsers. A cipher is the mathematical core of
an encryption algorithm. RC4, a fast cipher used to encrypt TLS data-streams, is known to have several
serious weaknesses. Unless you have very good reasons to support legacy browsers, you should disable
this.

✗ Old TLS versions are supported. The configuration supports TLS 1.0 (already deprecated) and TLS 1.1
(on a path to deprecation). Only TLS 1.2 has been recommended since 2018.

75
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 3: Test and harden the security configuration

✗ Forward secrecy is not fully supported. Forward secrecy is a feature of algorithms that encrypt using
temporary (ephemeral) session keys derived from the private key. This means in practice that attackers
cannot decrypt HTTPS data even if they possess a web server's long-term private key.

To correct and future-proof the TLS configuration

1. Open the configuration file /etc/httpd/conf.d/ssl.conf in a text editor and comment out the
following line by entering "#" at the beginning of the line.

#SSLProtocol all -SSLv3

2. Add the following directive:

#SSLProtocol all -SSLv3

SSLProtocol -SSLv2 -SSLv3 -TLSv1 -TLSv1.1 +TLSv1.2

This directive explicitly disables SSL versions 2 and 3, as well as TLS versions 1.0 and 1.1. The
server now refuses to accept encrypted connections with clients using anything except TLS 1.2.
The verbose wording in the directive conveys more clearly, to a human reader, what the server is
configured to do.
Note
Disabling TLS versions 1.0 and 1.1 in this manner blocks a small percentage of outdated
web browsers from accessing your site.

To modify the list of allowed ciphers

1. In the configuration file /etc/httpd/conf.d/ssl.conf, find the section with the

SSLCipherSuite directive and comment out the existing line by entering "#" at the beginning of
the line.

#SSLCipherSuite HIGH:MEDIUM:!aNULL:!MD5

2. Specify explicit cipher suites and a cipher order that prioritizes forward secrecy and avoids insecure
ciphers. The SSLCipherSuite directive used here is based on output from the Mozilla SSL
Configuration Generator, which tailors a TLS configuration to the specific software running on
your server. (For more information, see Mozilla's useful resource Security/Server Side TLS.) First
determine your Apache and OpenSSL versions by using the output from the following commands.

[ec2-user ~]$ yum list installed | grep httpd

[ec2-user ~]$ yum list installed | grep openssl

For example, if the returned information is Apache 2.4.34 and OpenSSL 1.0.2, we enter this into
the generator. If you choose the "modern" compatibility model, this creates an SSLCipherSuite
directive that aggressively enforces security but still works for most browsers. If your software
doesn't support the modern configuration, you can update your software or choose the
"intermediate" configuration instead.

SSLCipherSuite ECDHE-ECDSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-GCM-SHA384:ECDHE-ECDSA-
CHACHA20-POLY1305:
ECDHE-RSA-CHACHA20-POLY1305:ECDHE-ECDSA-AES128-GCM-SHA256:ECDHE-RSA-AES128-GCM-SHA256:
ECDHE-ECDSA-AES256-SHA384:ECDHE-RSA-AES256-SHA384:ECDHE-ECDSA-AES128-SHA256:ECDHE-RSA-
AES128-SHA256

76
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot

The selected ciphers have ECDHE in their names, an abbreviation for Elliptic Curve Diffie-Hellman
Ephemeral . The term ephemeral indicates forward secrecy. As a by-product, these ciphers do not
support RC4.

We recommend that you use an explicit list of ciphers instead of relying on defaults or terse
directives whose content isn't visible.

Copy the generated directive into /etc/httpd/conf.d/ssl.conf.

Note
Though shown here on several lines for readability, the directive must be on a single line
when copied to /etc/httpd/conf.d/ssl.conf, with only a colon (no spaces) between
cipher names.
3. Finally, uncomment the following line by removing the "#" at the beginning of the line.

#SSLHonorCipherOrder on

This directive forces the server to prefer high-ranking ciphers, including (in this case) those that
support forward secrecy. With this directive turned on, the server tries to establish a strong secure
connection before falling back to allowed ciphers with lesser security.

After completing both of these procedures, save the changes to /etc/httpd/conf.d/ssl.conf and
restart Apache.

If you test the domain again on Qualys SSL Labs, you should see that the RC4 vulnerability and other
warnings are gone and the summary looks something like the following.

Overall rating A

Certificate 100%

Protocol support 100%

Key exchange 90%

Cipher strength 90%

Important
Each update to OpenSSL introduces new ciphers and removes support for old ones. Keep your
EC2 Amazon Linux 2 instance up-to-date, watch for security announcements from OpenSSL,
and be alert to reports of new security exploits in the technical press. For more information,
see Predefined SSL Security Policies for Elastic Load Balancing in the User Guide for Classic Load
Balancers.

Troubleshoot
• My Apache webserver doesn't start unless I supply a password

This is expected behavior if you installed an encrypted, password-protected, private server key.

You can remove the encryption and password requirement from the key. Assuming that you have a
private encrypted RSA key called custom.key in the default directory, and that the password on it is
abcde12345, run the following commands on your EC2 instance to generate an unencrypted version
of the key.

77
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Certificate automation: Let's Encrypt
with Certbot on Amazon Linux 2
[ec2-user ~]$ cd /etc/pki/tls/private/
[ec2-user private]$ sudo cp custom.key custom.key.bak
[ec2-user private]$ sudo openssl rsa -in custom.key -passin pass:abcde12345 -out
custom.key.nocrypt
[ec2-user private]$ sudo mv custom.key.nocrypt custom.key
[ec2-user private]$ sudo chown root:root custom.key
[ec2-user private]$ sudo chmod 600 custom.key
[ec2-user private]$ sudo systemctl restart httpd

Apache should now start without prompting you for a password.

• I get errors when I run sudo yum install -y mod_ssl.

When you are installing the required packages for SSL, you may see errors similar to the following.

Error: httpd24-tools conflicts with httpd-tools-2.2.34-1.16.amzn1.x86_64

Error: httpd24 conflicts with httpd-2.2.34-1.16.amzn1.x86_64

This typically means that your EC2 instance is not running Amazon Linux 2. This tutorial only supports
instances freshly created from an official Amazon Linux 2 AMI.

Certificate automation: Let's Encrypt with Certbot on

Amazon Linux 2
The Let's Encrypt certificate authority is the centerpiece of an effort by the Electronic Frontier
Foundation (EFF) to encrypt the entire internet. In line with that goal, Let's Encrypt host certificates
are designed to be created, validated, installed, and maintained with minimal human intervention. The
automated aspects of certificate management are carried out by a software agent running on your
web server. After you install and configure the agent, it communicates securely with Let's Encrypt and
performs administrative tasks on Apache and the key management system. This tutorial uses the free
Certbot agent because it allows you either to supply a customized encryption key as the basis for your
certificates, or to allow the agent itself to create a key based on its defaults. You can also configure
Certbot to renew your certificates on a regular basis without human interaction, as described in To
automate Certbot (p. 81). For more information, consult the Certbot User Guide and man page.

Certbot is not officially supported on Amazon Linux 2, but is available for download and functions
correctly when installed. We recommend that you make the following backups to protect your data and
avoid inconvenience:

• Before you begin, take a snapshot of your Amazon EBS root volume. This allows you to restore the
original state of your EC2 instance. For information about creating EBS snapshots, see Create Amazon
EBS snapshots (p. 1129).
• The procedure below requires you to edit your httpd.conf file, which controls Apache's operation.
Certbot makes its own automated changes to this and other configuration files. Make a backup copy of
your entire /etc/httpd directory in case you need to restore it.

Prepare to install
Complete the following procedures before you install Certbot.

1. Download the Extra Packages for Enterprise Linux (EPEL) 7 repository packages. These are required
to supply dependencies needed by Certbot.

78
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Certificate automation: Let's Encrypt
with Certbot on Amazon Linux 2
a. Navigate to your home directory (/home/ec2-user). Download EPEL with the following
command.

[ec2-user ~]$ sudo wget -r --no-parent -A 'epel-release-*.rpm' https://

dl.fedoraproject.org/pub/epel/7/x86_64/Packages/e/

b. Install the repository packages as shown in the following command.

[ec2-user ~]$ sudo rpm -Uvh dl.fedoraproject.org/pub/epel/7/x86_64/Packages/e/epel-

release-*.rpm

c. Enable EPEL as shown in the following command.

[ec2-user ~]$ sudo yum-config-manager --enable epel*

You can confirm that EPEL is enabled with the following command. It should return information
similar to the following.

[ec2-user ~]$ sudo yum repolist all

...
epel/x86_64 Extra Packages for Enterprise Linux 7 - x86_64
enabled: 12949+175
epel-debuginfo/x86_64 Extra Packages for Enterprise Linux 7 - x86_64
- Debug enabled: 2890
epel-source/x86_64 Extra Packages for Enterprise Linux 7 - x86_64
- Source enabled: 0
epel-testing/x86_64 Extra Packages for Enterprise Linux 7 -
Testing - x86_64 enabled: 778+12
epel-testing-debuginfo/x86_64 Extra Packages for Enterprise Linux 7 -
Testing - x86_64 - Debug enabled: 107
epel-testing-source/x86_64 Extra Packages for Enterprise Linux 7 -
Testing - x86_64 - Source enabled: 0
...

2. Edit the main Apache configuration file, /etc/httpd/conf/httpd.conf. Locate the "Listen 80"
directive and add the following lines after it, replacing the example domain names with the actual
Common Name and Subject Alternative Name (SAN).

<VirtualHost *:80>
DocumentRoot "/var/www/html"
ServerName "example.com"
ServerAlias "www.example.com"
</VirtualHost>

Save the file and restart Apache.

[ec2-user ~]$ sudo systemctl restart httpd

Install and run Certbot

This procedure is based on the EFF documentation for installing Certbot on Fedora and on RHEL 7. It
describes the default use of Certbot, resulting in a certificate based on a 2048-bit RSA key.

1. Install Certbot packages and dependencies using the following command.

79
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Certificate automation: Let's Encrypt
with Certbot on Amazon Linux 2
[ec2-user ~]$ sudo yum install -y certbot python2-certbot-apache

2. Run Certbot.

[ec2-user ~]$ sudo certbot

3. At the prompt "Enter email address (used for urgent renewal and security notices)," enter a contact
address and press Enter.
4. Agree to the Let's Encrypt Terms of Service at the prompt. Enter "A" and press Enter to proceed.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Please read the Terms of Service at
https://letsencrypt.org/documents/LE-SA-v1.2-November-15-2017.pdf. You must
agree in order to register with the ACME server at
https://acme-v02.api.letsencrypt.org/directory
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
(A)gree/(C)ancel: A

5. At the authorization for EFF to put you on their mailing list, enter "Y" or "N" and press Enter.
6. Certbot displays the Common Name and Subject Alternative Name (SAN) that you provided in the
VirtualHost block.

Which names would you like to activate HTTPS for?

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
1: example.com
2: www.example.com
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Select the appropriate numbers separated by commas and/or spaces, or leave input
blank to select all options shown (Enter 'c' to cancel):

Leave the input blank and press Enter.

7. Certbot displays the following output as it creates certificates and configures Apache. It then
prompts you about redirecting HTTP queries to HTTPS.

Obtaining a new certificate

Performing the following challenges:
http-01 challenge for example.com
http-01 challenge for www.example.com
Waiting for verification...
Cleaning up challenges
Created an SSL vhost at /etc/httpd/conf/httpd-le-ssl.conf
Deploying Certificate for example.com to VirtualHost /etc/httpd/conf/httpd-le-ssl.conf
Enabling site /etc/httpd/conf/httpd-le-ssl.conf by adding Include to root configuration
Deploying Certificate for www.example.com to VirtualHost /etc/httpd/conf/httpd-le-
ssl.conf

Please choose whether or not to redirect HTTP traffic to HTTPS, removing HTTP access.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
1: No redirect - Make no further changes to the webserver configuration.
2: Redirect - Make all requests redirect to secure HTTPS access. Choose this for
new sites, or if you're confident your site works on HTTPS. You can undo this
change by editing your web server's configuration.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Select the appropriate number [1-2] then [enter] (press 'c' to cancel):

To allow visitors to connect to your server via unencrypted HTTP, enter "1". If you want to accept
only encrypted connections via HTTPS, enter "2". Press Enter to submit your choice.
8. Certbot completes the configuration of Apache and reports success and other information.

80
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Certificate automation: Let's Encrypt
with Certbot on Amazon Linux 2
Congratulations! You have successfully enabled https://example.com and
https://www.example.com

You should test your configuration at:

https://www.ssllabs.com/ssltest/analyze.html?d=example.com
https://www.ssllabs.com/ssltest/analyze.html?d=www.example.com
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

IMPORTANT NOTES:
- Congratulations! Your certificate and chain have been saved at:
/etc/letsencrypt/live/certbot.oneeyedman.net/fullchain.pem
Your key file has been saved at:
/etc/letsencrypt/live/certbot.oneeyedman.net/privkey.pem
Your cert will expire on 2019-08-01. To obtain a new or tweaked
version of this certificate in the future, simply run certbot again
with the "certonly" option. To non-interactively renew *all* of
your certificates, run "certbot renew"
- Your account credentials have been saved in your Certbot
configuration directory at /etc/letsencrypt. You should make a
secure backup of this folder now. This configuration directory will
also contain certificates and private keys obtained by Certbot so
making regular backups of this folder is ideal.

9. After you complete the installation, test and optimize the security of your server as described in
Step 3: Test and harden the security configuration (p. 75).

Configure automated certificate renewal

Certbot is designed to become an invisible, error-resistant part of your server system. By default, it
generates host certificates with a short, 90-day expiration time. If you have not configured your system
to call the command automatically, you must re-run the certbot command manually before expiration.
This procedure shows how to automate Certbot by setting up a cron job.

To automate Certbot

1. Open the /etc/crontab file in a text editor, such as vim or nano, using sudo. Alternatively, use
sudo crontab -e.
2. Add a line similar to the following and save the file.

39 1,13 * * * root certbot renew --no-self-upgrade

Here is an explanation of each component:

39 1,13 * * *

Schedules a command to be run at 01:39 and 13:39 every day. The selected values are arbitrary,
but the Certbot developers suggest running the command at least twice daily. This guarantees
that any certificate found to be compromised is promptly revoked and replaced.
root

The command runs with root permissions.

certbot renew --no-self-upgrade

The command to be run. The renew subcommand causes Certbot to check any previously
obtained certificates and to renew those that are approaching expiration. The --no-self-
upgrade flag prevents Certbot from upgrading itself without your intervention.
3. Restart the cron daemon.

81
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tutorial: Configure SSL/TLS on Amazon Linux

[ec2-user ~]$ sudo systemctl restart crond

Tutorial: Configure SSL/TLS on Amazon Linux

Secure Sockets Layer/Transport Layer Security (SSL/TLS) creates an encrypted channel between a web
server and web client that protects data in transit from being eavesdropped on. This tutorial explains
how to add support manually for SSL/TLS on an EC2 instance with the Amazon Linux AMI and Apache
web server. If you plan to offer commercial-grade services, the AWS Certificate Manager, which is not
discussed here, is a good option.

For historical reasons, web encryption is often referred to simply as SSL. While web browsers still
support SSL, its successor protocol TLS is less vulnerable to attack. The Amazon Linux AMI disables
server-side support all versions of SSL by default. Security standards bodies consider TLS 1.0 to be
unsafe, and both TLS 1.0 and TLS 1.1 are on track to be formally deprecated by the IETF. This tutorial
contains guidance based exclusively on enabling TLS 1.2. (A newer TLS 1.3 protocol exists in draft form,
but is not supported on Amazon Linux.) For more information about the updated encryption standards,
see RFC 7568 and RFC 8446.

This tutorial refers to modern web encryption simply as TLS.

Important
These procedures are intended for use with the Amazon Linux AMI. If you are trying to set up a
LAMP web server on an instance with a different distribution, some procedures in this tutorial
might not work for you. For information about LAMP web servers on Ubuntu, see the Ubuntu
community documentation ApacheMySQLPHP. For information about Red Hat Enterprise Linux,
see the Customer Portal documentation Web Servers.

Contents
• Prerequisites (p. 82)
• Step 1: Enable TLS on the server (p. 83)
• Step 2: Obtain a CA-signed certificate (p. 84)
• Step 3: Test and harden the security configuration (p. 89)
• Troubleshoot (p. 91)
• Certificate automation: Let's Encrypt with Certbot on Amazon Linux (p. 91)

Prerequisites
Before you begin this tutorial, complete the following steps:

• Launch an EBS-backed instance using the Amazon Linux AMI. For more information, see Step 1: Launch
an instance (p. 33).
• Configure your security group to allow your instance to accept connections on the following TCP ports:
• SSH (port 22)
• HTTP (port 80)
• HTTPS (port 443)

For more information, see Authorize inbound traffic for your Linux instances (p. 1041).
• Install Apache web server. For step-by-step instructions, see Tutorial: Installing a LAMP Web Server on
Amazon Linux (p. 48). Only the http24 package and its dependencies are needed; you can ignore
the instructions involving PHP and MySQL.

82
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 1: Enable TLS on the server

• To identify and authenticate web sites, the TLS public key infrastructure (PKI) relies on the Domain
Name System (DNS). To use your EC2 instance to host a public web site, you need to register a domain
name for your web server or transfer an existing domain name to your Amazon EC2 host. Numerous
third-party domain registration and DNS hosting services are available for this, or you can use Amazon
Route 53.

Step 1: Enable TLS on the server

This procedure takes you through the process of setting up TLS on Amazon Linux with a self-signed
digital certificate.
Note
A self-signed certificate is acceptable for testing but not production. If you expose your self-
signed certificate to the internet, visitors to your site receive security warnings.

To enable TLS on a server

1. Connect to your instance (p. 34) and confirm that Apache is running.

[ec2-user ~]$ sudo service httpd status

If necessary, start Apache.

[ec2-user ~]$ sudo service httpd start

2. To ensure that all of your software packages are up to date, perform a quick software update on
your instance. This process may take a few minutes, but it is important to make sure you have the
latest security updates and bug fixes.
Note
The -y option installs the updates without asking for confirmation. If you would like to
examine the updates before installing, you can omit this option.

[ec2-user ~]$ sudo yum update -y

3. Now that your instance is current, add TLS support by installing the Apache module mod_ssl:

[ec2-user ~]$ sudo yum install -y mod24_ssl

Your instance now has the following files that you use to configure your secure server and create a
certificate for testing:

/etc/httpd/conf.d/ssl.conf

The configuration file for mod_ssl. It contains "directives" telling Apache where to find
encryption keys and certificates, the TLS protocol versions to allow, and the encryption ciphers
to accept.
/etc/pki/tls/private/localhost.key

An automatically generated, 2048-bit RSA private key for your Amazon EC2 host. During
installation, OpenSSL used this key to generate a self-signed host certificate, and you can also
use this key to generate a certificate signing request (CSR) to submit to a certificate authority
(CA).

83
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Obtain a CA-signed certificate

/etc/pki/tls/certs/localhost.crt

An automatically generated, self-signed X.509 certificate for your server host. This certificate is
useful for testing that Apache is properly set up to use TLS.

The .key and .crt files are both in PEM format, which consists of Base64-encoded ASCII characters
framed by "BEGIN" and "END" lines, as in this abbreviated example of a certificate:

-----BEGIN CERTIFICATE-----
MIIEazCCA1OgAwIBAgICWxQwDQYJKoZIhvcNAQELBQAwgbExCzAJBgNVBAYTAi0t
MRIwEAYDVQQIDAlTb21lU3RhdGUxETAPBgNVBAcMCFNvbWVDaXR5MRkwFwYDVQQK
DBBTb21lT3JnYW5pemF0aW9uMR8wHQYDVQQLDBZTb21lT3JnYW5pemF0aW9uYWxV
bml0MRkwFwYDVQQDDBBpcC0xNzItMzEtMjAtMjM2MSQwIgYJKoZIhvcNAQkBFhVy
...
z5rRUE/XzxRLBZOoWZpNWTXJkQ3uFYH6s/sBwtHpKKZMzOvDedREjNKAvk4ws6F0
WanXWehT6FiSZvB4sTEXXJN2jdw8g+sHGnZ8zCOsclknYhHrCVD2vnBlZJKSZvak
3ZazhBxtQSukFMOnWPP2a0DMMFGYUHOd0BQE8sBJxg==
-----END CERTIFICATE-----

The file names and extensions are a convenience and have no effect on function; you can call a
certificate cert.crt, cert.pem, or any other file name, so long as the related directive in the
ssl.conf file uses the same name.
Note
When you replace the default TLS files with your own customized files, be sure that they are
in PEM format.
4. Restart Apache.

[ec2-user ~]$ sudo service httpd restart

5. Your Apache web server should now support HTTPS (secure HTTP) over port 443. Test it by typing
the IP address or fully qualified domain name of your EC2 instance into a browser URL bar with the
prefix https://. Because you are connecting to a site with a self-signed, untrusted host certificate,
your browser may display a series of security warnings.

Override the warnings and proceed to the site. If the default Apache test page opens, it means that
you have successfully configured TLS on your server. All data passing between the browser and
server is now safely encrypted.

To prevent site visitors from encountering warning screens, you need to obtain a certificate that not
only encrypts, but also publicly authenticates you as the owner of the site.

Step 2: Obtain a CA-signed certificate

You can use the following process to obtain a CA-signed certificate:

• Generate a certificate signing request (CSR) from a private key

• Submit the CSR to a certificate authority (CA)
• Obtain a signed host certificate
• Configure Apache to use the certificate

A self-signed TLS X.509 host certificate is cryptologically identical to a CA-signed certificate. The
difference is social, not mathematical; a CA promises to validate, at a minimum, a domain's ownership
before issuing a certificate to an applicant. Each web browser contains a list of CAs trusted by the
browser vendor to do this. An X.509 certificate consists primarily of a public key that corresponds to

84
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Obtain a CA-signed certificate

your private server key, and a signature by the CA that is cryptographically tied to the public key. When a
browser connects to a web server over HTTPS, the server presents a certificate for the browser to check
against its list of trusted CAs. If the signer is on the list, or accessible through a chain of trust consisting
of other trusted signers, the browser negotiates a fast encrypted data channel with the server and loads
the page.

Certificates generally cost money because of the labor involved in validating the requests, so it pays
to shop around. A list of well-known CAs can be found at dmoztools.net. A few CAs offer basic-level
certificates free of charge. The most notable of these is the Let's Encrypt project, which also supports
automation of the certificate creation and renewal process. For more information about using Let's
Encrypt as your CA, see Certificate automation: Let's Encrypt with Certbot on Amazon Linux (p. 91).

Underlying the host certificate is the key. As of 2017, government and industry groups recommend using
a minimum key (modulus) size of 2048 bits for RSA keys intended to protect documents through 2030.
The default modulus size generated by OpenSSL in Amazon Linux is 2048 bits, which means that the
existing auto-generated key is suitable for use in a CA-signed certificate. An alternative procedure is
described below for those who desire a customized key, for instance, one with a larger modulus or using
a different encryption algorithm.

These instructions for acquiring a CA-signed host certificate do not work unless you own a registered and
hosted DNS domain.

To obtain a CA-signed certificate

1. Connect to your instance (p. 34) and navigate to /etc/pki/tls/private/. This is the directory where
the server's private key for TLS is stored. If you prefer to use your existing host key to generate the
CSR, skip to Step 3.
2. (Optional) Generate a new private key. Here are some examples of key configurations. Any of
the resulting keys work with your web server, but they vary in how (and how much) security they
implement.

• Example 1: Create a default RSA host key. The resulting file, custom.key, is a 2048-bit RSA
private key.

[ec2-user ~]$ sudo openssl genrsa -out custom.key

• Example 2: Create a stronger RSA key with a bigger modulus. The resulting file, custom.key, is a
4096-bit RSA private key.

[ec2-user ~]$ sudo openssl genrsa -out custom.key 4096

• Example 3: Create a 4096-bit encrypted RSA key with password protection. The resulting file,
custom.key, is a 4096-bit RSA private key encrypted with the AES-128 cipher.
Important
Encrypting the key provides greater security, but because an encrypted key requires a
password, services depending on it cannot be auto-started. Each time you use this key,
you must supply the password (in the preceding example, "abcde12345") over an SSH
connection.

[ec2-user ~]$ sudo openssl genrsa -aes128 -passout pass:abcde12345 -out custom.key
4096

• Example 4: Create a key using a non-RSA cipher. RSA cryptography can be relatively slow because
of the size of its public keys, which are based on the product of two large prime numbers.
However, it is possible to create keys for TLS that use non-RSA ciphers. Keys based on the
mathematics of elliptic curves are smaller and computationally faster when delivering an
equivalent level of security.

85
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Obtain a CA-signed certificate

[ec2-user ~]$ sudo openssl ecparam -name prime256v1 -out custom.key -genkey

The result is a 256-bit elliptic curve private key using prime256v1, a "named curve" that OpenSSL
supports. Its cryptographic strength is slightly greater than a 2048-bit RSA key, according to NIST.
Note
Not all CAs provide the same level of support for elliptic-curve-based keys as for RSA
keys.

Make sure that the new private key has highly restrictive ownership and permissions (owner=root,
group=root, read/write for owner only). The commands would be as follows:

[ec2-user ~]$ sudo chown root.root custom.key

[ec2-user ~]$ sudo chmod 600 custom.key
[ec2-user ~]$ ls -al custom.key

The commands above should yield the following result:

-rw------- root root custom.key

After you have created and configured a satisfactory key, you can create a CSR.
3. Create a CSR using your preferred key; the example below uses custom.key:

[ec2-user ~]$ sudo openssl req -new -key custom.key -out csr.pem

OpenSSL opens a dialog and prompts you for the information shown in the following table. All of
the fields except Common Name are optional for a basic, domain-validated host certificate.

Name Description Example

Country Name The two-letter ISO abbreviation for your US (=United States)
country.

State or The name of the state or province where your Washington

Province organization is located. This name cannot be
Name abbreviated.

Locality Name The location of your organization, such as a city. Seattle

Organization The full legal name of your organization. Do not Example Corporation
Name abbreviate your organization name.

Organizational Additional organizational information, if any. Example Dept

Unit Name

Common This value must exactly match the web address www.example.com
Name that you expect users to type into a browser.
Usually, this means a domain name with
a prefixed host name or alias in the form
www.example.com. In testing with a self-
signed certificate and no DNS resolution,
the common name may consist of the host
name alone. CAs also offer more expensive

86
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Obtain a CA-signed certificate

Name Description Example

certificates that accept wild-card names such as
*.example.com.

Email Address The server administrator's email address. [email protected]

Finally, OpenSSL prompts you for an optional challenge password. This password applies only to the
CSR and to transactions between you and your CA, so follow the CA's recommendations about this
and the other optional field, optional company name. The CSR challenge password has no effect on
server operation.

The resulting file csr.pem contains your public key, your digital signature of your public key, and
the metadata that you entered.
4. Submit the CSR to a CA. This usually consists of opening your CSR file in a text editor and copying
the contents into a web form. At this time, you may be asked to supply one or more subject
alternate names (SANs) to be placed on the certificate. If www.example.com is the common name,
then example.com would be a good SAN, and vice versa. A visitor to your site typing in either of
these names would see an error-free connection. If your CA web form allows it, include the common
name in the list of SANs. Some CAs include it automatically.

After your request has been approved, you receive a new host certificate signed by the CA. You
might also be instructed to download an intermediate certificate file that contains additional
certificates needed to complete the CA's chain of trust.
Note
Your CA may send you files in multiple formats intended for various purposes. For this
tutorial, you should only use a certificate file in PEM format, which is usually (but not
always) marked with a .pem or .crt extension. If you are uncertain which file to use, open
the files with a text editor and find the one containing one or more blocks beginning with
the following:

- - - - -BEGIN CERTIFICATE - - - - -

The file should also end with the following:

- - - -END CERTIFICATE - - - - -

You can also test a file at the command line as follows:

[ec2-user certs]$ openssl x509 -in certificate.crt -text

Verify that these lines appear in the file. Do not use files ending with .p7b, .p7c, or similar
file extensions.
5. Place the new CA-signed certificate and any intermediate certificates in the /etc/pki/tls/certs
directory.
Note
There are several ways to upload your custom key to your EC2 instance, but the most
straightforward and informative way is to open a text editor (for example, vi, nano,
or notepad) on both your local computer and your instance, and then copy and paste
the file contents between them. You need root [sudo] permissions when performing
these operations on the EC2 instance. This way, you can see immediately if there are any
permission or path problems. Be careful, however, not to add any additional lines while
copying the contents, or to change them in any way.

87
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 2: Obtain a CA-signed certificate

From inside the /etc/pki/tls/certs directory, use the following commands to verify that the
file ownership, group, and permission settings match the highly restrictive Amazon Linux defaults
(owner=root, group=root, read/write for owner only).

[ec2-user certs]$ sudo chown root.root custom.crt

[ec2-user certs]$ sudo chmod 600 custom.crt
[ec2-user certs]$ ls -al custom.crt

The commands above should yield the following result:

-rw------- root root custom.crt

The permissions for the intermediate certificate file are less stringent (owner=root, group=root,
owner can write, group can read, world can read). The commands would be:

[ec2-user certs]$ sudo chown root.root intermediate.crt

[ec2-user certs]$ sudo chmod 644 intermediate.crt
[ec2-user certs]$ ls -al intermediate.crt

The commands above should yield the following result:

-rw-r--r-- root root intermediate.crt

6. If you used a custom key to create your CSR and the resulting host certificate, remove or rename the
old key from the /etc/pki/tls/private/ directory, and then install the new key there.
Note
There are several ways to upload your custom key to your EC2 instance, but the most
straightforward and informative way is to open a text editor (vi, nano, notepad, etc.) on
both your local computer and your instance, and then copy and paste the file contents
between them. You need root [sudo] privileges when performing these operations on
the EC2 instance. This way, you can see immediately if there are any permission or path
problems. Be careful, however, not to add any additional lines while copying the contents,
or to change them in any way.

From inside the /etc/pki/tls/private directory, check that the file ownership, group, and
permission settings match the highly restrictive Amazon Linux defaults (owner=root, group=root,
read/write for owner only). The commands would be as follows:

[ec2-user private]$ sudo chown root.root custom.key

[ec2-user private]$ sudo chmod 600 custom.key
[ec2-user private]$ ls -al custom.key

The commands above should yield the following result:

-rw------- root root custom.key

7. Edit /etc/httpd/conf.d/ssl.conf to reflect your new certificate and key files.

a. Provide the path and file name of the CA-signed host certificate in Apache's
SSLCertificateFile directive:

SSLCertificateFile /etc/pki/tls/certs/custom.crt

88
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 3: Test and harden the security configuration

b. If you received an intermediate certificate file (intermediate.crt in this example), provide its
path and file name using Apache's SSLCACertificateFile directive:

SSLCACertificateFile /etc/pki/tls/certs/intermediate.crt

Note
Some CAs combine the host certificate and the intermediate certificates in a single file,
making this directive unnecessary. Consult the instructions provided by your CA.
c. Provide the path and file name of the private key in Apache's SSLCertificateKeyFile
directive:

SSLCertificateKeyFile /etc/pki/tls/private/custom.key

8. Save /etc/httpd/conf.d/ssl.conf and restart Apache.

[ec2-user ~]$ sudo service httpd restart

9. Test your server by entering your domain name into a browser URL bar with the prefix https://.
Your browser should load the test page over HTTPS without generating errors.

Step 3: Test and harden the security configuration

After your TLS is operational and exposed to the public, you should test how secure it really is. This is
easy to do using online services such as Qualys SSL Labs, which performs a free and thorough analysis
of your security setup. Based on the results, you may decide to harden the default security configuration
by controlling which protocols you accept, which ciphers you prefer, and which you exclude. For more
information, see how Qualys formulates its scores.
Important
Real-world testing is crucial to the security of your server. Small configuration errors may lead
to serious security breaches and loss of data. Because recommended security practices change
constantly in response to research and emerging threats, periodic security audits are essential to
good server administration.

On the Qualys SSL Labs site, type the fully qualified domain name of your server, in the form
www.example.com. After about two minutes, you receive a grade (from A to F) for your site and a
detailed breakdown of the findings. Though the overview shows that the configuration is mostly sound,
the detailed report flags several potential problems. For example:

✗ The RC4 cipher is supported for use by certain older browsers. A cipher is the mathematical core of
an encryption algorithm. RC4, a fast cipher used to encrypt TLS data-streams, is known to have several
serious weaknesses. Unless you have very good reasons to support legacy browsers, you should disable
this.

✗ Old TLS versions are supported. The configuration supports TLS 1.0 (already deprecated) and TLS 1.1
(on a path to deprecation). Only TLS 1.2 has been recommended since 2018.

To correct the TLS configuration

1. Open the configuration file /etc/httpd/conf.d/ssl.conf in a text editor and comment out the
following lines by typing "#" at the beginning of each:

#SSLProtocol all -SSLv3

#SSLProxyProtocol all -SSLv3

2. Add the following directives:

89
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Step 3: Test and harden the security configuration

SSLProtocol -SSLv2 -SSLv3 -TLSv1 -TLSv1.1 +TLSv1.2

SSLProxyProtocol -SSLv2 -SSLv3 -TLSv1 -TLSv1.1 +TLSv1.2

These directives explicitly disable SSL versions 2 and 3, as well as TLS versions 1.0 and 1.1. The
server now refuses to accept encrypted connections with clients using anything except TLS 1.2. The
verbose wording in the directive communicates more clearly, to a human reader, what the server is
configured to do.
Note
Disabling TLS versions 1.0 and 1.1 in this manner blocks a small percentage of outdated
web browsers from accessing your site.

To modify the list of allowed ciphers

1. Open the configuration file /etc/httpd/conf.d/ssl.conf and find the section with
commented-out examples for configuring SSLCipherSuite and SSLProxyCipherSuite.

#SSLCipherSuite HIGH:MEDIUM:!aNULL:!MD5
#SSLProxyCipherSuite HIGH:MEDIUM:!aNULL:!MD5

Leave these as they are, and below them add the following directives:
Note
Though shown here on several lines for readability, each of these two directives must be on
a single line without spaces between the cipher names.

SSLCipherSuite ECDHE-ECDSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-GCM-SHA384:ECDHE-ECDSA-
CHACHA20-POLY1305:
ECDHE-RSA-CHACHA20-POLY1305:ECDHE-ECDSA-AES128-GCM-SHA256:ECDHE-RSA-AES128-GCM-
SHA256:ECDHE-ECDSA-AES256-SHA384:
ECDHE-RSA-AES256-SHA384:ECDHE-ECDSA-AES128-SHA256:ECDHE-RSA-AES128-SHA256:AES:!aNULL:!
eNULL:!EXPORT:!DES:
!RC4:!MD5:!PSK:!aECDH:!EDH-DSS-DES-CBC3-SHA:!EDH-RSA-DES-CBC3-SHA:!KRB5-DES-CBC3-SHA

SSLProxyCipherSuite ECDHE-ECDSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-GCM-SHA384:ECDHE-
ECDSA-CHACHA20-POLY1305:
ECDHE-RSA-CHACHA20-POLY1305:ECDHE-ECDSA-AES128-GCM-SHA256:ECDHE-RSA-AES128-GCM-
SHA256:ECDHE-ECDSA-AES256-SHA384:
ECDHE-RSA-AES256-SHA384:ECDHE-ECDSA-AES128-SHA256:ECDHE-RSA-AES128-SHA256:AES:!aNULL:!
eNULL:!EXPORT:!DES:
!RC4:!MD5:!PSK:!aECDH:!EDH-DSS-DES-CBC3-SHA:!EDH-RSA-DES-CBC3-SHA:!KRB5-DES-CBC3-SHA

These ciphers are a subset of the much longer list of supported ciphers in OpenSSL. They were
selected and ordered according to the following criteria:

• Support for forward secrecy

• Strength
• Speed
• Specific ciphers before cipher families
• Allowed ciphers before denied ciphers

Note that the high-ranking ciphers have ECDHE in their names, for Elliptic Curve Diffie-Hellman
Ephemeral ; the ephemeral indicates forward secrecy. Also, RC4 is now among the forbidden ciphers
near the end.

90
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot

We recommend that you use an explicit list of ciphers instead relying on defaults or terse directives
whose content isn't visible.
Important
The cipher list shown here is just one of many possible lists; for instance, you might want to
optimize a list for speed rather than forward secrecy.
If you anticipate a need to support older clients, you can allow the DES-CBC3-SHA cipher
suite.
Finally, each update to OpenSSL introduces new ciphers and deprecates old ones. Keep your
EC2 Amazon Linux instance up to date, watch for security announcements from OpenSSL,
and be alert to reports of new security exploits in the technical press. For more information,
see Predefined SSL Security Policies for Elastic Load Balancing in the User Guide for Classic
Load Balancers.
2. Uncomment the following line by removing the "#":

#SSLHonorCipherOrder on

This command forces the server to prefer high-ranking ciphers, including (in this case) those that
support forward secrecy. With this directive turned on, the server tries to establish a strongly secure
connection before falling back to allowed ciphers with lesser security.
3. Restart Apache. If you test the domain again on Qualys SSL Labs, you should see that the RC4
vulnerability is gone.

Troubleshoot
• My Apache webserver won't start unless I supply a password

This is expected behavior if you installed an encrypted, password-protected, private server key.

You can remove the encryption and password requirement from the key. Assuming that you have a
private encrypted RSA key called custom.key in the default directory, and that the password on it is
abcde12345, run the following commands on your EC2 instance to generate an unencrypted version
of the key.

[ec2-user ~]$ cd /etc/pki/tls/private/

[ec2-user private]$ sudo cp custom.key custom.key.bak
[ec2-user private]$ sudo openssl rsa -in custom.key -passin pass:abcde12345 -out
custom.key.nocrypt
[ec2-user private]$ sudo mv custom.key.nocrypt custom.key
[ec2-user private]$ sudo chown root.root custom.key
[ec2-user private]$ sudo chmod 600 custom.key
[ec2-user private]$ sudo service httpd restart

Apache should now start without prompting you for a password.

Certificate automation: Let's Encrypt with Certbot on

Amazon Linux
The Let's Encrypt certificate authority is the centerpiece of the Electronic Frontier Foundation (EFF) effort
to encrypt the entire internet. In line with that goal, Let's Encrypt host certificates are designed to be
created, validated, installed, and maintained with minimal human intervention. The automated aspects
of certificate management are carried out by an agent running on the web server. After you install and

91
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Certificate automation: Let's Encrypt
with Certbot on Amazon Linux
configure the agent, it communicates securely with Let's Encrypt and performs administrative tasks on
Apache and the key management system. This tutorial uses the free Certbot agent because it allows
you either to supply a customized encryption key as the basis for your certificates, or to allow the agent
itself to create a key based on its defaults. You can also configure Certbot to renew your certificates on
a regular basis without human interaction, as described in To automate Certbot (p. 81). For more
information, consult the Certbot User Guide or man page.

Certbot is not officially supported on Amazon Linux AMI, but is available for download and functions
correctly when installed. We recommend that you make the following backups to protect your data and
avoid inconvenience:

• Before you begin, take a snapshot of your Amazon EBS root volume. This allows you to restore the
original state of your EC2 instance. For information about creating EBS snapshots, see Create Amazon
EBS snapshots (p. 1129).
• The procedure below requires you to edit your httpd.conf file, which controls Apache's operation.
Certbot makes its own automated changes to this and other configuration files. Make a backup copy of
your entire /etc/httpd directory in case you need to restore it.

To install and run Certbot

1. Enable the Extra Packages for Enterprise Linux (EPEL) repository from the Fedora project on your
instance. Packages from EPEL are required as dependencies when you run the Certbot installation
script.

[ec2-user ~]$ sudo yum-config-manager --enable epel

2. Download the latest release of Certbot from EFF onto your EC2 instance using the following
command.

[ec2-user ~]$ wget https://dl.eff.org/certbot-auto

3. Make the downloaded file executable.

[ec2-user ~]$ chmod a+x certbot-auto

4. Run the file with root permissions and the --debug flag.

[ec2-user ~]$ sudo ./certbot-auto --debug

5. At the prompt "Is this ok [y/d/N]," type "y" and press Enter.
6. At the prompt "Enter email address (used for urgent renewal and security notices)," type a contact
address and press Enter.
7. Agree to the Let's Encrypt Terms of Service at the prompt. Type "A" and press Enter to proceed:

-------------------------------------------------------------------------------
Please read the Terms of Service at
https://letsencrypt.org/documents/LE-SA-v1.1.1-August-1-2016.pdf. You must agree
in order to register with the ACME server at
https://acme-v01.api.letsencrypt.org/directory
-------------------------------------------------------------------------------
(A)gree/(C)ancel: A

8. Click through the authorization for EFF put you on their mailing list by typing "Y" or "N" and press
Enter.

92
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Certificate automation: Let's Encrypt
with Certbot on Amazon Linux
9. At the prompt shown below, type your Common Name (the name of your domain as described
above) and your Subject Alternative Name (SAN), separating the two names with a space or a
comma. Then press Enter. In this example, the names have been provided:

No names were found in your configuration files. Please enter in your domain
name(s) (comma and/or space separated) (Enter 'c' to cancel):example.com
www.example.com

10. On an Amazon Linux system with a default Apache configuration, you see output similar to the
example below, asking about the first name you provided. Type "1" and press Enter.

Obtaining a new certificate

Performing the following challenges:
tls-sni-01 challenge for example.com
tls-sni-01 challenge for www.example.com

We were unable to find a vhost with a ServerName or Address of example.com.

Which virtual host would you like to choose?
(note: conf files with multiple vhosts are not yet supported)
-------------------------------------------------------------------------------
1: ssl.conf | | HTTPS | Enabled
-------------------------------------------------------------------------------
Press 1 [enter] to confirm the selection (press 'c' to cancel): 1

11. Next, Certbot asks about the second name. Type "1" and press Enter.

We were unable to find a vhost with a ServerName or Address of www.example.com.

Which virtual host would you like to choose?
(note: conf files with multiple vhosts are not yet supported)
-------------------------------------------------------------------------------
1: ssl.conf | | HTTPS | Enabled
-------------------------------------------------------------------------------
Press 1 [enter] to confirm the selection (press 'c' to cancel): 1

At this point, Certbot creates your key and a CSR:

Waiting for verification...

Cleaning up challenges
Generating key (2048 bits): /etc/letsencrypt/keys/0000_key-certbot.pem
Creating CSR: /etc/letsencrypt/csr/0000_csr-certbot.pem

12. Authorize Certbot to create and all needed host certificates. When prompted for each name, type
"1" and press Enter as shown in the example:

We were unable to find a vhost with a ServerName or Address of example.com.

Which virtual host would you like to choose?
(note: conf files with multiple vhosts are not yet supported)
-------------------------------------------------------------------------------
1: ssl.conf | | HTTPS | Enabled
-------------------------------------------------------------------------------
Press 1 [enter] to confirm the selection (press 'c' to cancel): 1
Deploying Certificate for example.com to VirtualHost /etc/httpd/conf.d/ssl.conf

We were unable to find a vhost with a ServerName or Address of www.example.com.

Which virtual host would you like to choose?
(note: conf files with multiple vhosts are not yet supported)
-------------------------------------------------------------------------------
1: ssl.conf | example.com | HTTPS | Enabled
-------------------------------------------------------------------------------
Press 1 [enter] to confirm the selection (press 'c' to cancel): 1

93
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Certificate automation: Let's Encrypt
with Certbot on Amazon Linux
Deploying Certificate for www.example.com to VirtualHost /etc/httpd/conf.d/ssl.conf

13. Choose whether to allow insecure connections to your web server. If you choose option 2 (as shown
in the example), all connections to your server will either be encrypted or rejected.

Please choose whether HTTPS access is required or optional.

-------------------------------------------------------------------------------
1: Easy - Allow both HTTP and HTTPS access to these sites
2: Secure - Make all requests redirect to secure HTTPS access
-------------------------------------------------------------------------------
Select the appropriate number [1-2] then [enter] (press 'c' to cancel): 2

Certbot completes the configuration of Apache and reports success and other information:

Congratulations! You have successfully enabled https://example.com and

https://www.example.com

You should test your configuration at:

https://www.ssllabs.com/ssltest/analyze.html?d=example.com
https://www.ssllabs.com/ssltest/analyze.html?d=www.example.com
-------------------------------------------------------------------------------

IMPORTANT NOTES:
- Congratulations! Your certificate and chain have been saved at
/etc/letsencrypt/live/example.com/fullchain.pem. Your cert will
expire on 2017-07-19. To obtain a new or tweaked version of this
certificate in the future, simply run certbot-auto again with the
"certonly" option. To non-interactively renew *all* of your
certificates, run "certbot-auto renew"
....

14. After you complete the installation, test and optimize the security of your server as described in
Step 3: Test and harden the security configuration (p. 75).

Certbot is designed to become an invisible, error-resistant part of your server system. By default, it
generates host certificates with a short, 90-day expiration time. If you have not previously configured
your system to call the command automatically, you must re-run the certbot command manually. This
procedure shows how to automate Certbot by setting up a cron job.

To configure automated certificate renewal

1. Open the /etc/crontab file in a text editor, such as vim or nano, using sudo. Alternatively, use
sudo crontab -e.
2. Add a line similar to the following and save the file.

39 1,13 * * * root certbot renew --no-self-upgrade

Here is an explanation of each component:

39 1,13 * * *

Schedules a command to be run at 01:39 and 13:39 every day. The selected values are arbitrary,
but the Certbot developers suggest running the command at least twice daily. This guarantees
that any certificate found to be compromised is promptly revoked and replaced.
root

The command runs with root privileges.

94
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tutorial: Increase the availability of your application

certbot renew --no-self-upgrade

The command to be run. The renew subcommand causes Certbot to check any previously
obtained certificates and to renew those that are approaching expiration. The --no-self-
upgrade flag prevents Certbot from upgrading itself without your intervention.
3. Restart the cron daemon:

[ec2-user ~]$ sudo service crond restart

Tutorial: Increase the availability of your

application on Amazon EC2
Suppose that you start out running your app or website on a single EC2 instance, and over time, traffic
increases to the point that you require more than one instance to meet the demand. You can launch
multiple EC2 instances from your AMI and then use Elastic Load Balancing to distribute incoming traffic
for your application across these EC2 instances. This increases the availability of your application. Placing
your instances in multiple Availability Zones also improves the fault tolerance in your application. If one
Availability Zone experiences an outage, traffic is routed to the other Availability Zone.

You can use Amazon EC2 Auto Scaling to maintain a minimum number of running instances for your
application at all times. Amazon EC2 Auto Scaling can detect when your instance or application is
unhealthy and replace it automatically to maintain the availability of your application. You can also
use Amazon EC2 Auto Scaling to scale your Amazon EC2 capacity up or down automatically based on
demand, using criteria that you specify.

In this tutorial, we use Amazon EC2 Auto Scaling with Elastic Load Balancing to ensure that you maintain
a specified number of healthy EC2 instances behind your load balancer. Note that these instances do not
need public IP addresses, because traffic goes to the load balancer and is then routed to the instances.
For more information, see Amazon EC2 Auto Scaling and Elastic Load Balancing.

Contents

95
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Prerequisites

• Prerequisites (p. 96)

• Scale and load balance your application (p. 96)
• Test your load balancer (p. 98)

Prerequisites
This tutorial assumes that you have already done the following:

1. Create a virtual private cloud (VPC) with one public subnet in two or more Availability Zones.
2. Launch an instance in the VPC.
3. Connect to the instance and customized it. For example, installing software and applications,
copying data, and attaching additional EBS volumes. For information about setting up a web server
on your instance, see Tutorial: Install a LAMP web server with the Amazon Linux AMI (p. 48).
4. Test your application on your instance to ensure that your instance is configured correctly.
5. Create a custom Amazon Machine Image (AMI) from your instance. For more information, see Create
an Amazon EBS-backed Linux AMI (p. 125) or Create an instance store-backed Linux AMI (p. 130).
6. (Optional) Terminate the instance if you no longer need it.
7. Create an IAM role that grants your application the access to AWS it needs. For more information,
see To create an IAM role using the IAM console (p. 1034).

Scale and load balance your application

Use the following procedure to create a load balancer, create a launch configuration for your instances,
create an Auto Scaling group with two or more instances, and associate the load balancer with the Auto
Scaling group.

To scale and load-balance your application

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. On the navigation pane, under LOAD BALANCING, choose Load Balancers.
3. Choose Create Load Balancer.
4. For Application Load Balancer, choose Create.
5. On the Configure Load Balancer page, do the following:

a. For Name, enter a name for your load balancer. For example, my-lb.
b. For Scheme, keep the default value, internet-facing.
c. For Listeners, keep the default, which is a listener that accepts HTTP traffic on port 80.
d. For Availability Zones, select the VPC that you used for your instances. Select an Availability
Zone and then select the public subnet for that Availability Zone. Repeat for a second
Availability Zone.
e. Choose Next: Configure Security Settings.
6. For this tutorial, you are not using a secure listener. Choose Next: Configure Security Groups.
7. On the Configure Security Groups page, do the following:

a. Choose Create a new security group.

b. Type a name and description for the security group, or keep the default name and description.
This new security group contains a rule that allows traffic to the port configured for the listener.
c. Choose Next: Configure Routing.
8. On the Configure Routing page, do the following:

96
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scale and load balance your application

a. For Target group, keep the default, New target group.

b. For Name, enter a name for the target group.
c. Keep Protocol as HTTP, Port as 80, and Target type as instance.
d. For Health checks, keep the default protocol and path.
e. Choose Next: Register Targets.
9. On the Register Targets page, choose Next: Review to continue to the next page, as we'll use
Amazon EC2 Auto Scaling to add EC2 instances to the target group.
10. On the Review page, choose Create. After the load balancer is created, choose Close.
11. On the navigation pane, under AUTO SCALING, choose Launch Configurations.
• If you are new to Amazon EC2 Auto Scaling, you see a welcome page. Choose Create Auto
Scaling group to start the Create Auto Scaling Group wizard, and then choose Create launch
configuration.
• Otherwise, choose Create launch configuration.
12. For Launch configuration name, enter a name for your launch configuration (for example, my-
launch-config).
13. For Amazon machine image (AMI), under My AMIs, choose the AMI that you created in
Prerequisites (p. 96).
14. For Instance type, use Choose instance type to select an instance type.
15. (Optional) For Additional configuration, do the following as needed:

a. Choose Request Spot Instances. Otherwise, the instances are On-Demand instances.
b. For IAM instance profile, select the IAM role that you created in Prerequisites (p. 96).
c. Choose Enable EC2 instance detailed monitoring within CloudWatch. Otherwise, basic
monitoring is enabled.
d. To configure instance metadata, expand Advanced details, enable or disable instance metadata,
and configure the version and hop limit as needed.
e. To run a startup script, expand Advanced details and enter the script in User data.
f. To assign public IP addresses to your instances, expand Advanced details and set IP address
type as needed.
16. For Storage (volumes), you add volumes as needed. You can create empty EBS volumes or create
EBS volumes from EBS snapshots.
17. For Security groups, you can select an existing security group or create a new one. This security
group must allow HTTP traffic and health checks from the load balancer. If you assigned public IP
addresses to your instances, you can optionally allow SSH traffic so you can connect to them.
18. For Key pair (login), choose an existing key pair, create a new key pair, or proceed without a key pair.
Select the acknowledgment check box.
19. Choose Create launch configuration.
20. After the launch configuration is created, you must create an Auto Scaling group.
• If you are new to Amazon EC2 Auto Scaling and you are using the Create Auto Scaling group
wizard, you are taken to the next step automatically.
• Otherwise, select the Auto Scaling group and choose Actions, Create an Auto Scaling group.
21. On the Choose launch template or configuration page, enter a name for the Auto Scaling group.
For example, my-asg. Choose Next.
22. On the Configure settings page, choose your VPC and your two public subnets. Choose Next.
23. On the Configure advanced options page, select Enable load balancing and choose your target
group. Select the ELB health check type and choose Next.
24. For Group size, type the number of instances (for example, 2). Note that we recommend that you
maintain approximately the same number of instances in each Availability Zone.

97
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Test your load balancer

25. On the Configure group size and scaling policies page, you can configure the size of the group or
configure the group to scale dynamically based on demand. Choose Next.
26. (Optional) Choose Add notifications to configure SNS notifications for scaling activities. Choose
Next.
27. (Optional) Choose Add tag to add tags. Choose Next.
28. On the Review page, edit the details as needed, and then choose Create Auto Scaling group.

Test your load balancer

When a client sends a request to your load balancer, the load balancer routes the request to one of its
registered instances.

To test your load balancer

1. Verify that your instances are ready. From the Auto Scaling Groups page, select your Auto Scaling
group, and then choose the Instance management tab. Initially, your instances are in the Pending
state. When Lifecycle is InService, your instances are ready for use.
2. Verify that your instances are registered with the load balancer. From the Target Groups page,
choose the name of the target group to open its details page, and then choose Targets. If the state
of your instances is initial, it's possible that they are still registering. When the state of your
instances is healthy, they are ready for use. After your instances are ready, you can go to the next
step.
3. From the Load Balancers page, select your load balancer.
4. On the Description tab, locate the DNS name. This name has the following form:

my-lb-xxxxxxxxxx.us-west-2.elb.amazonaws.com

5. In a web browser, paste the DNS name for the load balancer into the address bar and press Enter.
You'll see your website displayed.

98
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Use an AMI

Amazon Machine Images (AMI)

An Amazon Machine Image (AMI) provides the information required to launch an instance. You must
specify an AMI when you launch an instance. You can launch multiple instances from a single AMI when
you need multiple instances with the same configuration. You can use different AMIs to launch instances
when you need instances with different configurations.

An AMI includes the following:

• One or more Amazon Elastic Block Store (Amazon EBS) snapshots, or, for instance-store-backed AMIs,
a template for the root volume of the instance (for example, an operating system, an application
server, and applications).
• Launch permissions that control which AWS accounts can use the AMI to launch instances.
• A block device mapping that specifies the volumes to attach to the instance when it's launched.

Topics
• Use an AMI (p. 99)
• Create your own AMI (p. 100)
• Buy, share, and sell AMIs (p. 100)
• Deregister your AMI (p. 101)
• Amazon Linux 2 and Amazon Linux AMI (p. 101)
• AMI types (p. 101)
• Linux AMI virtualization types (p. 104)
• Find a Linux AMI (p. 106)
• Shared AMIs (p. 111)
• Paid AMIs (p. 121)
• Create an Amazon EBS-backed Linux AMI (p. 125)
• Automate the EBS-backed AMI lifecycle (p. 130)
• Create an instance store-backed Linux AMI (p. 130)
• Use encryption with EBS-backed AMIs (p. 161)
• Copy an AMI (p. 166)
• Understanding AMI billing information (p. 172)
• Deregister your Linux AMI (p. 175)
• Amazon Linux (p. 178)
• User provided kernels (p. 196)
• Use the MATE desktop environment provided with Amazon Linux 2 (p. 201)

Use an AMI
The following diagram summarizes the AMI lifecycle. After you create and register an AMI, you can use it
to launch new instances. (You can also launch instances from an AMI if the AMI owner grants you launch

99
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create your own AMI

permissions.) You can copy an AMI within the same AWS Region or to different AWS Regions. When you
no longer require an AMI, you can deregister it.

You can search for an AMI that meets the criteria for your instance. You can search for AMIs provided by
AWS or AMIs provided by the community. For more information, see AMI types (p. 101) and Finding an
AMI.

After you launch an instance from an AMI, you can connect to it. When you are connected to an instance,
you can use it just like you use any other server. For information about launching, connecting, and using
your instance, see Tutorial: Get started with Amazon EC2 Linux instances (p. 32).

Create your own AMI

You can launch an instance from an existing AMI, customize the instance (for example, install
software (p. 654) on the instance), and then save this updated configuration as a custom AMI. Instances
launched from this new custom AMI include the customizations that you made when you created the
AMI.

The root storage device of the instance determines the process you follow to create an AMI. The root
volume of an instance is either an Amazon Elastic Block Store (Amazon EBS) volume or an instance store
volume. For more information about the root device volume, see Amazon EC2 root device volume (p. 21).

• To create an Amazon EBS-backed AMI, see Create an Amazon EBS-backed Linux AMI (p. 125).
• To create an instance store-backed AMI, see Create an instance store-backed Linux AMI (p. 130).

To help categorize and manage your AMIs, you can assign custom tags to them. For more information,
see Tag your Amazon EC2 resources (p. 1311).

Buy, share, and sell AMIs

After you create an AMI, you can keep it private so that only you can use it, or you can share it with a
specified list of AWS accounts. You can also make your custom AMI public so that the community can use
it. Building a safe, secure, usable AMI for public consumption is a fairly straightforward process, if you
follow a few simple guidelines. For information about how to create and use shared AMIs, see Shared
AMIs (p. 111).

You can purchase AMIs from a third party, including AMIs that come with service contracts from
organizations such as Red Hat. You can also create an AMI and sell it to other Amazon EC2 users. For
more information about buying or selling AMIs, see Paid AMIs (p. 121).

100
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Deregister your AMI

Deregister your AMI

You can deregister an AMI when you have finished with it. After you deregister an AMI, it can't be used to
launch new instances. Existing instances launched from the AMI are not affected. For more information,
see Deregister your Linux AMI (p. 175).

Amazon Linux 2 and Amazon Linux AMI

Amazon Linux 2 and the Amazon Linux AMI are supported and maintained Linux images provided by
AWS. The following are some of the features of Amazon Linux 2 and Amazon Linux AMI:

• A stable, secure, and high-performance execution environment for applications running on Amazon
EC2.
• Provided at no additional charge to Amazon EC2 users.
• Repository access to multiple versions of MySQL, PostgreSQL, Python, Ruby, Tomcat, and many more
common packages.
• Updated on a regular basis to include the latest components, and these updates are also made
available in the yum repositories for installation on running instances.
• Includes packages that enable easy integration with AWS services, such as the AWS CLI, Amazon EC2
API and AMI tools, the Boto library for Python, and the Elastic Load Balancing tools.

For more information, see Amazon Linux (p. 178).

AMI types
You can select an AMI to use based on the following characteristics:

• Region (see Regions and Zones (p. 7))

• Operating system
• Architecture (32-bit or 64-bit)
• Launch permissions (p. 101)
• Storage for the root device (p. 102)

Launch permissions
The owner of an AMI determines its availability by specifying launch permissions. Launch permissions fall
into the following categories.

Launch Description
permission

public The owner grants launch permissions to all AWS accounts.

explicit The owner grants launch permissions to specific AWS accounts.

implicit The owner has implicit launch permissions for an AMI.

101
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Storage for the root device

Amazon and the Amazon EC2 community provide a large selection of public AMIs. For more information,
see Shared AMIs (p. 111). Developers can charge for their AMIs. For more information, see Paid
AMIs (p. 121).

Storage for the root device

All AMIs are categorized as either backed by Amazon EBS or backed by instance store. The former means
that the root device for an instance launched from the AMI is an Amazon Elastic Block Store (Amazon
EBS) volume created from an Amazon EBS snapshot. The latter means that the root device for an
instance launched from the AMI is an instance store volume created from a template stored in Amazon
S3. For more information, see Amazon EC2 root device volume (p. 21).

The following table summarizes the important differences when using the two types of AMIs.

Characteristic Amazon EBS-backed AMI Amazon instance store-backed AMI

Boot time for an Usually less than 1 minute Usually less than 5 minutes
instance

Size limit for a root 16 TiB 10 GiB

device

Root device volume EBS volume Instance store volume

Data persistence By default, the root volume Data on any instance store volumes
is deleted when the instance persists only during the life of the
terminates.* Data on any other EBS instance.
volumes persists after instance
termination by default.

Modifications The instance type, kernel, RAM disk, Instance attributes are fixed for the
and user data can be changed while life of an instance.
the instance is stopped.

Charges You're charged for instance usage, You're charged for instance usage
EBS volume usage, and storing your and storing your AMI in Amazon S3.
AMI as an EBS snapshot.

AMI creation/bundling Uses a single command/call Requires installation and use of AMI
tools

Stopped state Can be in a stopped state. Even Cannot be in stopped state;

when the instance is stopped and instances are running or terminated
not running, the root volume is
persisted in Amazon EBS

* By default, EBS root volumes have the DeleteOnTermination flag set to true. For information
about how to change this flag so that the volume persists after termination, see Change the root volume
to persist (p. 24).

Determin the root device type of your AMI

To determine the root device type of an AMI using the console

1. Open the Amazon EC2 console.

2. In the navigation pane, click AMIs, and select the AMI.

102
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Storage for the root device

3. Check the value of Root Device Type in the Details tab as follows:

• If the value is ebs, this is an Amazon EBS-backed AMI.

• If the value is instance store, this is an instance store-backed AMI.

To determine the root device type of an AMI using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• describe-images (AWS CLI)

• Get-EC2Image (AWS Tools for Windows PowerShell)

Stopped state
You can stop an Amazon EBS-backed instance, but not an Amazon EC2 instance store-backed instance.
Stopping causes the instance to stop running (its status goes from running to stopping to stopped).
A stopped instance persists in Amazon EBS, which allows it to be restarted. Stopping is different from
terminating; you can't restart a terminated instance. Because Amazon EC2 instance store-backed
instances can't be stopped, they're either running or terminated. For more information about what
happens and what you can do while an instance is stopped, see Stop and start your instance (p. 623).

Default data storage and persistence

Instances that use an instance store volume for the root device automatically have instance store
available (the root volume contains the root partition and you can store additional data). You can add
persistent storage to your instance by attaching one or more EBS volumes. Any data on an instance
store volume is deleted when the instance fails or terminates. For more information, see Instance store
lifetime (p. 1269).

Instances that use Amazon EBS for the root device automatically have an EBS volume attached. The
volume appears in your list of volumes like any other. With most instance types, Amazon EBS-backed
instances don't have instance store volumes by default. You can add instance store volumes or additional
EBS volumes using a block device mapping. For more information, see Block device mapping (p. 1294).

Boot times
Instances launched from an Amazon EBS-backed AMI launch faster than instances launched from an
instance store-backed AMI. When you launch an instance from an instance store-backed AMI, all the
parts have to be retrieved from Amazon S3 before the instance is available. With an Amazon EBS-backed
AMI, only the parts required to boot the instance need to be retrieved from the snapshot before the
instance is available. However, the performance of an instance that uses an EBS volume for its root
device is slower for a short time while the remaining parts are retrieved from the snapshot and loaded
into the volume. When you stop and restart the instance, it launches quickly, because the state is stored
in an EBS volume.

AMI creation
To create Linux AMIs backed by instance store, you must create an AMI from your instance on the
instance itself using the Amazon EC2 AMI tools.

AMI creation is much easier for AMIs backed by Amazon EBS. The CreateImage API action creates your
Amazon EBS-backed AMI and registers it. There's also a button in the AWS Management Console that
lets you create an AMI from a running instance. For more information, see Create an Amazon EBS-backed
Linux AMI (p. 125).

103
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Virtualization types

How you're charged

With AMIs backed by instance store, you're charged for instance usage and storing your AMI in Amazon
S3. With AMIs backed by Amazon EBS, you're charged for instance usage, EBS volume storage and usage,
and storing your AMI as an EBS snapshot.

With Amazon EC2 instance store-backed AMIs, each time you customize an AMI and create a new one, all
of the parts are stored in Amazon S3 for each AMI. So, the storage footprint for each customized AMI is
the full size of the AMI. For Amazon EBS-backed AMIs, each time you customize an AMI and create a new
one, only the changes are stored. So the storage footprint for subsequent AMIs you customize after the
first is much smaller, resulting in lower AMI storage charges.

When an Amazon EBS-backed instance is stopped, you're not charged for instance usage; however, you're
still charged for volume storage. As soon as you start your instance, we charge a minimum of one minute
for usage. After one minute, we charge only for the seconds used. For example, if you run an instance for
20 seconds and then stop it, we charge for a full one minute. If you run an instance for 3 minutes and 40
seconds, we charge for exactly 3 minutes and 40 seconds of usage. We charge you for each second, with
a one-minute minimum, that you keep the instance running, even if the instance remains idle and you
don't connect to it.

Linux AMI virtualization types

Linux Amazon Machine Images use one of two types of virtualization: paravirtual (PV) or hardware
virtual machine (HVM). The main differences between PV and HVM AMIs are the way in which they boot
and whether they can take advantage of special hardware extensions (CPU, network, and storage) for
better performance.

For the best performance, we recommend that you use current generation instance types and HVM
AMIs when you launch your instances. For more information about current generation instance types,
see Amazon EC2 Instance Types. If you are using previous generation instance types and would like to
upgrade, see Upgrade Paths.

The following table compares HVM and PV AMIs.

  HVM PV

Description HVM AMIs are presented with a PV AMIs boot with a special boot
fully virtualized set of hardware loader called PV-GRUB, which
and boot by executing the starts the boot cycle and then
master boot record of the root chain loads the kernel specified
block device of your image. This in the menu.lst file on your
virtualization type provides image. Paravirtual guests can
the ability to run an operating run on host hardware that does
system directly on top of a not have explicit support for
virtual machine without any virtualization. Historically, PV
modification, as if it were run guests had better performance
on the bare-metal hardware. than HVM guests in many cases,
The Amazon EC2 host system but because of enhancements
emulates some or all of the in HVM virtualization and the
underlying hardware that is availability of PV drivers for
presented to the guest. HVM AMIs, this is no longer true.
For more information about PV-
GRUB and its use in Amazon
EC2, see Enabling Your Own
Linux Kernels (p. 196).

104
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Virtualization types

  HVM PV

Support for hardware extensions Yes. Unlike PV guests, HVM No, they cannot take advantage
guests can take advantage of special hardware extensions
of hardware extensions that such as enhanced networking or
provide fast access to the GPU processing.
underlying hardware on the host
system. For more information
on CPU virtualization extensions
available in Amazon EC2, see
Intel Virtualization Technology
on the Intel website. HVM AMIs
are required to take advantage
of enhanced networking and
GPU processing. In order to
pass through instructions to
specialized network and GPU
devices, the OS needs to be
able to have access to the
native hardware platform; HVM
virtualization provides this
access. For more information,
see Enhanced networking
on Linux (p. 864) and
Linux accelerated computing
instances (p. 290).

Supported instance types All current generation instance The following previous
types support HVM AMIs. generation instance types
support PV AMIs: C1, C3, HS1,
M1, M3, M2, and T1. Current
generation instance types do not
support PV AMIs.

Supported Regions All Regions support HVM Asia Pacific (Tokyo), Asia
instances. Pacific (Singapore), Asia Pacific
(Sydney), Europe (Frankfurt),
Europe (Ireland), South America
(São Paulo), US East (N. Virginia),
US West (N. California), and US
West (Oregon)

How to find Verify that the virtualization Verify that the virtualization
type of the AMI is set to hvm, type of the AMI is set to
using the console or the paravirtual, using the
describe-images command. console or the describe-images
command.

PV on HVM

Paravirtual guests traditionally performed better with storage and network operations than HVM guests
because they could leverage special drivers for I/O that avoided the overhead of emulating network
and disk hardware, whereas HVM guests had to translate these instructions to emulated hardware.
Now PV drivers are available for HVM guests, so operating systems that cannot be ported to run in
a paravirtualized environment can still see performance advantages in storage and network I/O by
using them. With these PV on HVM drivers, HVM guests can get the same, or better, performance than
paravirtual guests.

105
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Find a Linux AMI

Find a Linux AMI

Before you can launch an instance, you must select an AMI to use. As you select an AMI, consider the
following requirements you might have for the instances that you'll launch:

• The Region
• The operating system
• The architecture: 32-bit (i386), 64-bit (x86_64), or 64-bit ARM (arm64)
• The root device type: Amazon EBS or instance store
• The provider (for example, Amazon Web Services)
• Additional software (for example, SQL server)

If you need to find a Windows AMI, see Finding a Windows AMI in the Amazon EC2 User Guide for
Windows Instances.

Contents
• Find a Linux AMI using the Amazon EC2 console (p. 106)
• Find an AMI using the AWS CLI (p. 107)
• Find the latest Amazon Linux AMI using Systems Manager (p. 107)
• Use a Systems Manager parameter to find an AMI (p. 108)
• Find a Quick Start AMI (p. 110)

Find a Linux AMI using the Amazon EC2 console

You can find Linux AMIs using the Amazon EC2 console. You can select from the list of AMIs when you
use the launch wizard to launch an instance, or you can search through all available AMIs using the
Images page. AMI IDs are unique to each AWS Region.

To find a Linux AMI using the launch wizard

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region in which to launch your instances. You can select any
Region that's available to you, regardless of your location.
3. From the console dashboard, choose Launch instance.
4. On the Quick Start tab, select from one of the commonly used AMIs in the list. If you don't
see the AMI that you need, select the My AMIs, AWS Marketplace, or Community AMIs tab
to find additional AMIs. For more information, see Step 1: Choose an Amazon Machine Image
(AMI) (p. 527).

To find a Linux AMI using the Images page

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region in which to launch your instances. You can select any
Region that's available to you, regardless of your location.
3. In the navigation pane, choose AMIs.
4. (Optional) Use the Filter options to scope the list of displayed AMIs to see only the AMIs that
interest you. For example, to list all Linux AMIs provided by AWS, select Public images. Choose the
Search bar and select Owner from the menu, then select Amazon images. Choose the Search bar
again to select Platform and then the operating system from the list provided.

106
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Find an AMI using the AWS CLI

5. (Optional) Choose the Show/Hide Columns icon to select which image attributes to display, such as
the root device type. Alternatively, you can select an AMI from the list and view its properties in the
Details tab.
6. Before you select an AMI, it's important that you check whether it's backed by instance store or by
Amazon EBS and that you are aware of the effects of this difference. For more information, see
Storage for the root device (p. 102).
7. To launch an instance from this AMI, select it and then choose Launch. For more information about
launching an instance using the console, see Launching your instance from an AMI (p. 528). If
you're not ready to launch the instance now, make note of the AMI ID for later.

Find an AMI using the AWS CLI

You can use AWS CLI commands for Amazon EC2 to list only the Linux AMIs that meet your needs. After
locating an AMI that meets your needs, make note of its ID so that you can use it to launch instances. For
more information, see Launching an Instance Using the AWS CLI in the AWS Command Line Interface User
Guide.

The describe-images command supports filtering parameters. For example, use the --owners parameter
to display public AMIs owned by Amazon.

aws ec2 describe-images --owners self amazon

You can add the following filter to the previous command to display only AMIs backed by Amazon EBS.

--filters "Name=root-device-type,Values=ebs"

Important
Omitting the --owners flag from the describe-images command will return all images for
which you have launch permissions, regardless of ownership.

Find the latest Amazon Linux AMI using Systems

Manager
Amazon EC2 provides AWS Systems Manager public parameters for AWS-maintained public AMIs that
you can use when launching instances. For example, the EC2-provided parameter /aws/service/ami-
amazon-linux-latest/amzn2-ami-hvm-x86_64-gp2 is available in all Regions and always points to
the latest version of the Amazon Linux 2 AMI in a given Region.

The Amazon EC2 AMI public parameters are available from the following paths:

• /aws/service/ami-amazon-linux-latest
• /aws/service/ami-windows-latest

You can view a list of all Linux AMIs in the current AWS Region by using the following command in the
AWS CLI.

aws ssm get-parameters-by-path --path /aws/service/ami-amazon-linux-latest --query

"Parameters[].Name"

To launch an instance using a public parameter

The following example uses the EC2-provided public parameter to launch an m5.xlarge instance using
the latest Amazon Linux 2 AMI.

107
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Use a Systems Manager parameter to find an AMI

To specify the parameter in the command, use the following syntax: resolve:ssm:public-
parameter, where resolve:ssm is the standard prefix and public-parameter is the path and name
of the public parameter.

In this example, the --count and --security-group parameters are not included. For --count, the
default is 1. If you have a default VPC and a default security group, they are used.

aws ec2 run-instances

--image-id resolve:ssm:/aws/service/ami-amazon-linux-latest/amzn2-ami-hvm-x86_64-gp2
--instance-type m5.xlarge
--key-name MyKeyPair

For more information, see Using public parameters in the AWS Systems Manager User Guide and Query
for the latest Amazon Linux AMI IDs Using AWS Systems Manager Parameter Store.

Use a Systems Manager parameter to find an AMI

When you launch an instance using the EC2 launch wizard in the console, you can either select an AMI
from the list, or you can select an AWS Systems Manager parameter that points to an AMI ID. If you use
automation code to launch your instances, you can specify the Systems Manager parameter instead of
the AMI ID.

A Systems Manager parameter is a customer-defined key-value pair that you can create in Systems
Manager Parameter Store. The Parameter Store provides a central store to externalize your application
configuration values. For more information, see AWS Systems Manager Parameter Store in the AWS
Systems Manager User Guide.

When you create a parameter that points to an AMI ID, make sure that you specify the data type as
aws:ec2:image. This data type ensures that when the parameter is created or modified, the parameter
value is validated as an AMI ID. For more information, see Native parameter support for Amazon Machine
Image IDs in the AWS Systems Manager User Guide.

Contents
• Use cases (p. 108)
• Launch an instance using a Systems Manager parameter (p. 109)
• Permissions (p. 110)
• Limitations (p. 110)

Use cases
By using Systems Manager parameters to point to AMI IDs, you can make it easier for your users to select
the correct AMI when launching instances, and you can simplify the maintenance of automation code.

Easier for users

If you require instances to be launched using a specific AMI, and if that AMI is updated regularly, we
recommend that you require your users to select a Systems Manager parameter to find the AMI. By
requiring your users to select a Systems Manager parameter, you can ensure that the latest AMI is used
to launch instances.

For example, every month in your organization you might create a new version of your AMI that has
the latest operating system and application patches. You also require your users to launch instances
using the latest version of your AMI. To ensure that your users use the latest version, you can create a
Systems Manager parameter (for example, golden-ami) that points to the correct AMI ID. Each time a
new version of the AMI is created, you update the AMI ID value in the parameter so that it always points

108
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Use a Systems Manager parameter to find an AMI

to the latest AMI. Your users don't need to know about the periodic updates to the AMI, because they
continue to select the same Systems Manager parameter every time. By having users select a Systems
Manager parameter, you make it easier for them to select the correct AMI for an instance launch.

Simplify automation code maintenance

If you use automation code to launch your instances, you can specify the Systems Manager parameter
instead of the AMI ID. If a new version of the AMI is created, you change the AMI ID value in the
parameter so that it points to the latest AMI. The automation code that references the parameter doesn’t
need to be modified every time a new version of the AMI is created. This greatly simplifies maintenance
of automation and helps drive down deployment costs.
Note
Running instances are not affected when you change the AMI ID to which the Systems Manager
parameter points.

Launch an instance using a Systems Manager parameter

You can launch an instance using the console or the AWS CLI. Instead of specifying an AMI ID, you can
specify an AWS Systems Manager parameter that points to an AMI ID.

To find a Linux AMI using a Systems Manager parameter (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region in which to launch your instances. You can select any
Region that's available to you, regardless of your location.
3. From the console dashboard, choose Launch instance.
4. Choose Search by Systems Manager parameter (at top right).
5. For Systems Manager parameter, select a parameter. The corresponding AMI ID appears next to
Currently resolves to.
6. Choose Search. The AMIs that match the AMI ID appear in the list.
7. Select the AMI from the list, and choose Select.

For more information about launching an instance from an AMI using the launch wizard, see Step 1:
Choose an Amazon Machine Image (AMI) (p. 527).

To launch an instance using an AWS Systems Manager parameter instead of an AMI ID (AWS CLI)

The following example uses the Systems Manager parameter golden-ami to launch an m5.xlarge
instance. The parameter points to an AMI ID.

To specify the parameter in the command, use the following syntax: resolve:ssm:/parameter-name,
where resolve:ssm is the standard prefix and parameter-name is the unique parameter name. Note
that the parameter name is case-sensitive. Backslashes for the parameter name are only necessary when
the parameter is part of a hierarchy, for example, /amis/production/golden-ami. You can omit the
backslash if the parameter is not part of a hierarchy.

In this example, the --count and --security-group parameters are not included. For --count, the
default is 1. If you have a default VPC and a default security group, they are used.

aws ec2 run-instances

--image-id resolve:ssm:/golden-ami
--instance-type m5.xlarge
...

To launch an instance using a specific version of an AWS Systems Manager parameter (AWS CLI)

109
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Find a Quick Start AMI

Systems Manager parameters have version support. Each iteration of a parameter is assigned a unique
version number. You can reference the version of the parameter as follows resolve:ssm:parameter-
name:version, where version is the unique version number. By default, the latest version of the
parameter is used when no version is specified.

The following example uses version 2 of the parameter.

In this example, the --count and --security-group parameters are not included. For --count, the
default is 1. If you have a default VPC and a default security group, they are used.

aws ec2 run-instances

--image-id resolve:ssm:/golden-ami:2
--instance-type m5.xlarge
...

To launch an instance using a public parameter provided by AWS

Amazon EC2 provides Systems Manager public parameters for public AMIs provided by AWS. For
example, the public parameter /aws/service/ami-amazon-linux-latest/amzn2-ami-hvm-x86_64-gp2 is
available in all Regions and always points to the latest version of the Amazon Linux 2 AMI in the Region.

aws ec2 run-instances

--image-id resolve:ssm:/aws/service/ami-amazon-linux-latest/amzn2-ami-hvm-x86_64-gp2
--instance-type m5.xlarge
...

Permissions
If you use Systems Manager parameters that point to AMI IDs in the launch instance wizard,
you must add ssm:DescribeParameters and ssm:GetParameters to your IAM policy.
ssm:DescribeParameters grants your IAM users the permission to view and select Systems Manager
parameters. ssm:GetParameters grants your IAM users the permission to get the values of the
Systems Manager parameters. You can also restrict access to specific Systems Manager parameters. For
more information, see Use the EC2 launch wizard (p. 1025).

Limitations
AMIs and Systems Manager parameters are Region specific. To use the same Systems Manager parameter
name across Regions, create a Systems Manager parameter in each Region with the same name (for
example, golden-ami). In each Region, point the Systems Manager parameter to an AMI in that Region.

Find a Quick Start AMI

When you launch an instance using the Amazon EC2 console, the Choose an Amazon Machine Image
(AMI) page includes a list of popular AMIs on the Quick Start tab. If you want to automate launching an
instance using one of these quick start AMIs, you'll need to programatically locate the ID of the current
version of the AMI.

To locate the current version of a Quick Start AMI, you can enumerate all AMIs with its AMI name, and
then find the one with the most recent creation date.

Example Example: Find the current Amazon Linux 2 AMI

aws ec2 describe-images \

--owners amazon \
--filters "Name=name,Values=amzn2-ami-hvm-2.0.????????.?-x86_64-gp2"
"Name=state,Values=available" \

110
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Shared AMIs

--query "reverse(sort_by(Images, &CreationDate))[:1].ImageId" \

--output text

Example Example: Find the current Amazon Linux AMI

aws ec2 describe-images \

--owners amazon \
--filters "Name=name,Values=amzn-ami-hvm-????.??.?.????????-x86_64-gp2"
"Name=state,Values=available" \
--query "reverse(sort_by(Images, &CreationDate))[:1].ImageId" \
--output text

Example Example: Find the current Ubuntu Server 16.04 LTS AMI

aws ec2 describe-images \

--owners 099720109477 \
--filters "Name=name,Values=ubuntu/images/hvm-ssd/ubuntu-xenial-16.04-amd64-
server-????????" "Name=state,Values=available" \
--query "reverse(sort_by(Images, &CreationDate))[:1].ImageId" \
--output text

Example Example: Find the current Red Hat Enterprise Linux 7.5 AMI

aws ec2 describe-images \

--owners 309956199498 \
--filters "Name=name,Values=RHEL-7.5_HVM_GA*" "Name=state,Values=available" \
--query "reverse(sort_by(Images, &CreationDate))[:1].ImageId" \
--output text

Example Example: Find the current SUSE Linux Enterprise Server 15 AMI

aws ec2 describe-images \

--owners amazon \
--filters "Name=name,Values=suse-sles-15-v????????-hvm-ssd-x86_64"
"Name=state,Values=available" \
--query "reverse(sort_by(Images, &CreationDate))[:1].ImageId" \
--output text

Shared AMIs
A shared AMI is an AMI that a developer created and made available for other developers to use. One of
the easiest ways to get started with Amazon EC2 is to use a shared AMI that has the components you
need and then add custom content. You can also create your own AMIs and share them with others.

You use a shared AMI at your own risk. Amazon can't vouch for the integrity or security of AMIs shared
by other Amazon EC2 users. Therefore, you should treat shared AMIs as you would any foreign code that
you might consider deploying in your own data center and perform the appropriate due diligence. We
recommend that you get an AMI from a trusted source.

Amazon's public images have an aliased owner, which appears as amazon in the account field. This
enables you to find AMIs from Amazon easily. Other users can't alias their AMIs.

For information about creating an AMI, see Creating an instance store-backed Linux AMI or Creating an
Amazon EBS-backed Linux AMI. For more information about building, delivering, and maintaining your
applications on the AWS Marketplace, see the AWS Marketplace Documentation.

111
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Find shared AMIs

Contents
• Find shared AMIs (p. 112)
• Make an AMI public (p. 114)
• Share an AMI with specific AWS accounts (p. 115)
• Use bookmarks (p. 117)
• Guidelines for shared Linux AMIs (p. 117)

Find shared AMIs

You can use the Amazon EC2 console or the command line to find shared AMIs.

AMIs are a regional resource. Therefore, when searching for a shared AMI (public or private), you must
search for it from within the Region from which it is being shared. To make an AMI available in a different
Region, copy the AMI to the Region and then share it. For more information, see Copying an AMI.

Topics
• Find a shared AMI (console) (p. 112)
• Find a shared AMI (AWS CLI) (p. 112)
• Use shared AMIs (p. 113)

Find a shared AMI (console)

To find a shared private AMI using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose AMIs.
3. In the first filter, choose Private images. All AMIs that have been shared with you are listed. To
granulate your search, choose the Search bar and use the filter options provided in the menu.

To find a shared public AMI using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose AMIs.
3. In the first filter, choose Public images. To granulate your search, choose the Search bar and use the
filter options provided in the menu.
4. Use filters to list only the types of AMIs that interest you. For example, choose Owner : and then
choose Amazon images to display only Amazon's public images.

Find a shared AMI (AWS CLI)

Use the describe-images command (AWS CLI) to list AMIs. You can scope the list to the types of AMIs that
interest you, as shown in the following examples.

Example: List all public AMIs

The following command lists all public AMIs, including any public AMIs that you own.

aws ec2 describe-images --executable-users all

112
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Find shared AMIs

Example: List AMIs with explicit launch permissions

The following command lists the AMIs for which you have explicit launch permissions. This list does not
include any AMIs that you own.

aws ec2 describe-images --executable-users self

Example: List AMIs owned by Amazon

The following command lists the AMIs owned by Amazon. Amazon's public AMIs have an aliased owner,
which appears as amazon in the account field. This enables you to find AMIs from Amazon easily. Other
users can't alias their AMIs.

aws ec2 describe-images --owners amazon

Example: List AMIs owned by an account

The following command lists the AMIs owned by the specified AWS account.

aws ec2 describe-images --owners 123456789012

Example: Scope AMIs using a filter

To reduce the number of displayed AMIs, use a filter to list only the types of AMIs that interest you. For
example, use the following filter to display only EBS-backed AMIs.

--filters "Name=root-device-type,Values=ebs"

Use shared AMIs

Before you use a shared AMI, take the following steps to confirm that there are no pre-installed
credentials that would allow unwanted access to your instance by a third party and no pre-configured
remote logging that could transmit sensitive data to a third party. Check the documentation for the
Linux distribution used by the AMI for information about improving the security of the system.

To ensure that you don't accidentally lose access to your instance, we recommend that you initiate
two SSH sessions and keep the second session open until you've removed credentials that you don't
recognize and confirmed that you can still log into your instance using SSH.

1. Identify and disable any unauthorized public SSH keys. The only key in the file should be the key you
used to launch the AMI. The following command locates authorized_keys files:

[ec2-user ~]$ sudo find / -name "authorized_keys" -print -exec cat {} \;

2. Disable password-based authentication for the root user. Open the sshd_config file and edit the
PermitRootLogin line as follows:

PermitRootLogin without-password

Alternatively, you can disable the ability to log into the instance as the root user:

PermitRootLogin No

Restart the sshd service.

113
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Make an AMI public

3. Check whether there are any other user accounts that are able to log in to your instance. Accounts
with superuser privileges are particularly dangerous. Remove or lock the password of any unknown
accounts.
4. Check for open ports that you aren't using and running network services listening for incoming
connections.
5. To prevent preconfigured remote logging, you should delete the existing configuration file and
restart the rsyslog service. For example:

[ec2-user ~]$ sudo rm /etc/rsyslog.conf

[ec2-user ~]$ sudo service rsyslog restart

6. Verify that all cron jobs are legitimate.

If you discover a public AMI that you feel presents a security risk, contact the AWS security team. For
more information, see the AWS Security Center.

Make an AMI public

Amazon EC2 enables you to share your AMIs with other AWS accounts. You can allow all AWS accounts to
use the AMI to launch instances (by making the AMI public), or only allow a few specific accounts to use
the AMI to launch instances (see Share an AMI with specific AWS accounts (p. 115)). You are not billed
when your AMI is used by other AWS accounts to launch instances; only the accounts launching instances
using the AMI are billed for the instances they launch.

AMIs with encrypted volumes cannot be made public.

AMIs are a regional resource. Therefore, sharing an AMI makes it available in that Region. To make an AMI
available in a different Region, copy the AMI to the Region and then share it. For more information, see
Copy an AMI (p. 166).

To avoid exposing sensitive data when you share an AMI, read the security considerations in Guidelines
for shared Linux AMIs (p. 117) and follow the recommended actions.

If an AMI has a product code, or contains a snapshot of an encrypted volume, you can't make it public;
you can share the AMI only with specific AWS accounts.

Topics
• Share an AMI with all AWS accounts (console) (p. 114)
• Share an AMI with all AWS accounts (AWS CLI) (p. 115)

Share an AMI with all AWS accounts (console)

After you make an AMI public, it is available in Community AMIs when you launch an instance in the
same Region using the console. Note that it can take a short while for an AMI to appear in Community
AMIs after you make it public. It can also take a short while for an AMI to be removed from Community
AMIs after you make it private again.

To share a public AMI using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose AMIs.
3. Select your AMI from the list, and then choose Actions, Modify Image Permissions.
4. Choose Public and choose Save.

114
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Share an AMI with specific AWS accounts

Share an AMI with all AWS accounts (AWS CLI)

Each AMI has a launchPermission property that controls which AWS accounts, besides the owner's,
are allowed to use that AMI to launch instances. By modifying the launchPermission property of an
AMI, you can make the AMI public (which grants launch permissions to all AWS accounts) or share it with
only the AWS accounts that you specify.

You can add or remove account IDs from the list of accounts that have launch permissions for an AMI. To
make the AMI public, specify the all group. You can specify both public and explicit launch permissions.

To make an AMI public

1. Use the modify-image-attribute command as follows to add the all group to the
launchPermission list for the specified AMI.

aws ec2 modify-image-attribute \

--image-id ami-0abcdef1234567890 \
--launch-permission "Add=[{Group=all}]"

2. To verify the launch permissions of the AMI, use the describe-image-attribute command.

aws ec2 describe-image-attribute \

--image-id ami-0abcdef1234567890 \
--attribute launchPermission

3. (Optional) To make the AMI private again, remove the all group from its launch permissions.
Note that the owner of the AMI always has launch permissions and is therefore unaffected by this
command.

aws ec2 modify-image-attribute \

--image-id ami-0abcdef1234567890 \
--launch-permission "Remove=[{Group=all}]"

Share an AMI with specific AWS accounts

You can share an AMI with specific AWS accounts without making the AMI public. All you need is the AWS
account IDs. You can only share AMIs that have unencrypted volumes and volumes that are encrypted
with a customer managed CMK. If you share an AMI with encrypted volumes, you must also share any
CMKs used to encrypt them. For more information, see Share an Amazon EBS snapshot (p. 1139). You
cannot share an AMI that has volumes that are encrypted with a AWS managed CMK.

AMIs are a regional resource. Therefore, sharing an AMI makes it available in that Region. To make an AMI
available in a different Region, copy the AMI to the Region and then share it. For more information, see
Copy an AMI (p. 166).

There is no limit to the number of AWS accounts with which an AMI can be shared. User-defined tags
that you attach to a shared AMI are available only to your AWS account and not to the other accounts
that the AMI is shared with.

Share an AMI (console)

To grant explicit launch permissions using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose AMIs.

115
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Share an AMI with specific AWS accounts

3. Select your AMI in the list, and then choose Actions, Modify Image Permissions.
4. Specify the AWS account number of the user with whom you want to share the AMI in the AWS
Account Number field, then choose Add Permission.

To share this AMI with multiple users, repeat this step until you have added all the required users.
5. To allow create volume permissions for snapshots, select Add "create volume" permissions to the
following associated snapshots when creating permissions.
Note
You do not need to share the Amazon EBS snapshots that an AMI references in order to
share the AMI. Only the AMI itself needs to be shared; the system automatically provides
the instance access to the referenced Amazon EBS snapshots for the launch. However, you
do need to share any CMKs used to encrypt snapshots that the AMI references. For more
information, see Share an Amazon EBS snapshot (p. 1139).
6. Choose Save when you are done.
7. (Optional) To view the AWS account IDs with which you have shared the AMI, select the AMI in
the list, and choose the Permissions tab. To find AMIs that are shared with you, see Find shared
AMIs (p. 112).

Share an AMI (AWS CLI)

Use the modify-image-attribute command (AWS CLI) to share an AMI as shown in the following
examples.

To grant explicit launch permissions

The following command grants launch permissions for the specified AMI to the specified AWS account.

aws ec2 modify-image-attribute \

--image-id ami-0abcdef1234567890 \
--launch-permission "Add=[{UserId=123456789012}]"

The following command grants create volume permission for a snapshot.

aws ec2 modify-snapshot-attribute \

--snapshot-id snap-1234567890abcdef0 \
--attribute createVolumePermission \
--operation-type add \
--user-ids 123456789012

Note
You do not need to share the Amazon EBS snapshots that an AMI references in order to share
the AMI. Only the AMI itself needs to be shared; the system automatically provides the instance
access to the referenced Amazon EBS snapshots for the launch. However, you do need to share
any CMKs used to encrypt snapshots that the AMI references. For more information, see Share
an Amazon EBS snapshot (p. 1139).

To remove launch permissions for an account

The following command removes launch permissions for the specified AMI from the specified AWS
account:

aws ec2 modify-image-attribute \

--image-id ami-0abcdef1234567890 \
--launch-permission "Remove=[{UserId=123456789012}]"

116
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Use bookmarks

The following command removes create volume permission for a snapshot.

aws ec2 modify-snapshot-attribute \

--snapshot-id snap-1234567890abcdef0 \
--attribute createVolumePermission \
--operation-type remove \
--user-ids 123456789012

To remove all launch permissions

The following command removes all public and explicit launch permissions from the specified AMI. Note
that the owner of the AMI always has launch permissions and is therefore unaffected by this command.

aws ec2 reset-image-attribute \

--image-id ami-0abcdef1234567890 \
--attribute launchPermission

Use bookmarks
If you have created a public AMI, or shared an AMI with another AWS user, you can create a bookmark
that allows a user to access your AMI and launch an instance in their own account immediately. This is an
easy way to share AMI references, so users don't have to spend time finding your AMI in order to use it.

Note that your AMI must be public, or you must have shared it with the user to whom you want to send
the bookmark.

To create a bookmark for your AMI

1. Type a URL with the following information, where region is the Region in which your AMI resides:

https://console.aws.amazon.com/ec2/v2/home?
region=region#LaunchInstanceWizard:ami=ami_id

For example, this URL launches an instance from the ami-0abcdef1234567890 AMI in the us-east-1
Region:

https://console.aws.amazon.com/ec2/v2/home?region=us-
east-1#LaunchInstanceWizard:ami=ami-0abcdef1234567890

2. Distribute the link to users who want to use your AMI.

3. To use a bookmark, choose the link or copy and paste it into your browser. The launch wizard opens,
with the AMI already selected.

Guidelines for shared Linux AMIs

Use the following guidelines to reduce the attack surface and improve the reliability of the AMIs you
create.
Important
No list of security guidelines can be exhaustive. Build your shared AMIs carefully and take time
to consider where you might expose sensitive data.

Contents
• Update the AMI tools before using them (p. 118)
• Disable password-based remote logins for root (p. 118)

117
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Guidelines for shared Linux AMIs

• Disable local root access (p. 119)

• Remove SSH host key pairs (p. 119)
• Install public key credentials (p. 119)
• Disabling sshd DNS checks (optional) (p. 120)
• Identify yourself (p. 121)
• Protect yourself (p. 121)

If you are building AMIs for AWS Marketplace, see Building AMIs for AWS Marketplace for guidelines,
policies and best practices.

For additional information about sharing AMIs safely, see the following articles:

• How To Share and Use Public AMIs in A Secure Manner

• Public AMI Publishing: Hardening and Clean-up Requirements

Update the AMI tools before using them

For AMIs backed by instance store, we recommend that your AMIs download and upgrade the Amazon
EC2 AMI creation tools before you use them. This ensures that new AMIs based on your shared AMIs have
the latest AMI tools.

For Amazon Linux 2, install the aws-amitools-ec2 package and add the AMI tools to your PATH with
the following command. For the Amazon Linux AMI, aws-amitools-ec2 package is already installed by
default.

[ec2-user ~]$ sudo yum install -y aws-amitools-ec2 && export PATH=$PATH:/opt/aws/bin > /
etc/profile.d/aws-amitools-ec2.sh && . /etc/profile.d/aws-amitools-ec2.sh

Upgrade the AMI tools with the following command:

[ec2-user ~]$ sudo yum upgrade -y aws-amitools-ec2

For other distributions, make sure you have the latest AMI tools.

Disable password-based remote logins for root

Using a fixed root password for a public AMI is a security risk that can quickly become known. Even
relying on users to change the password after the first login opens a small window of opportunity for
potential abuse.

To solve this problem, disable password-based remote logins for the root user.

To disable password-based remote logins for root

1. Open the /etc/ssh/sshd_config file with a text editor and locate the following line:

#PermitRootLogin yes

2. Change the line to:

PermitRootLogin without-password

The location of this configuration file might differ for your distribution, or if you are not running
OpenSSH. If this is the case, consult the relevant documentation.

118
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Guidelines for shared Linux AMIs

Disable local root access

When you work with shared AMIs, a best practice is to disable direct root logins. To do this, log into your
running instance and issue the following command:

[ec2-user ~]$ sudo passwd -l root

Note
This command does not impact the use of sudo.

Remove SSH host key pairs

If you plan to share an AMI derived from a public AMI, remove the existing SSH host key pairs located in
/etc/ssh. This forces SSH to generate new unique SSH key pairs when someone launches an instance
using your AMI, improving security and reducing the likelihood of "man-in-the-middle" attacks.

Remove all of the following key files that are present on your system.

• ssh_host_dsa_key
• ssh_host_dsa_key.pub
• ssh_host_key
• ssh_host_key.pub
• ssh_host_rsa_key
• ssh_host_rsa_key.pub
• ssh_host_ecdsa_key
• ssh_host_ecdsa_key.pub
• ssh_host_ed25519_key
• ssh_host_ed25519_key.pub

You can securely remove all of these files with the following command.

[ec2-user ~]$ sudo shred -u /etc/ssh/*_key /etc/ssh/*_key.pub

Warning
Secure deletion utilities such as shred may not remove all copies of a file from your storage
media. Hidden copies of files may be created by journalling file systems (including Amazon Linux
default ext4), snapshots, backups, RAID, and temporary caching. For more information see the
shred documentation.
Important
If you forget to remove the existing SSH host key pairs from your public AMI, our routine
auditing process notifies you and all customers running instances of your AMI of the potential
security risk. After a short grace period, we mark the AMI private.

Install public key credentials

After configuring the AMI to prevent logging in using a password, you must make sure users can log in
using another mechanism.

Amazon EC2 allows users to specify a public-private key pair name when launching an instance. When
a valid key pair name is provided to the RunInstances API call (or through the command line API
tools), the public key (the portion of the key pair that Amazon EC2 retains on the server after a call to
CreateKeyPair or ImportKeyPair) is made available to the instance through an HTTP query against
the instance metadata.

119
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Guidelines for shared Linux AMIs

To log in through SSH, your AMI must retrieve the key value at boot and append it to /root/.ssh/
authorized_keys (or the equivalent for any other user account on the AMI). Users can launch instances
of your AMI with a key pair and log in without requiring a root password.

Many distributions, including Amazon Linux and Ubuntu, use the cloud-init package to inject public
key credentials for a configured user. If your distribution does not support cloud-init, you can add
the following code to a system start-up script (such as /etc/rc.local) to pull in the public key you
specified at launch for the root user.
Note
In the following example, the IP address http://169.254.169.254/ is a link-local address and is
valid only from the instance.

IMDSv2

if [ ! -d /root/.ssh ] ; then
mkdir -p /root/.ssh
chmod 700 /root/.ssh
fi
# Fetch public key using HTTP
TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-metadata-
token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/meta-
data/public-keys/0/openssh-key > /tmp/my-key
if [ $? -eq 0 ] ; then
cat /tmp/my-key >> /root/.ssh/authorized_keys
chmod 700 /root/.ssh/authorized_keys
rm /tmp/my-key
fi

IMDSv1

if [ ! -d /root/.ssh ] ; then
mkdir -p /root/.ssh
chmod 700 /root/.ssh
fi
# Fetch public key using HTTP
curl http://169.254.169.254/latest/meta-data/public-keys/0/openssh-key > /tmp/my-key
if [ $? -eq 0 ] ; then
cat /tmp/my-key >> /root/.ssh/authorized_keys
chmod 700 /root/.ssh/authorized_keys
rm /tmp/my-key
fi

This can be applied to any user account; you do not need to restrict it to root.
Note
Rebundling an instance based on this AMI includes the key with which it was launched. To
prevent the key's inclusion, you must clear out (or delete) the authorized_keys file or exclude
this file from rebundling.

Disabling sshd DNS checks (optional)

Disabling sshd DNS checks slightly weakens your sshd security. However, if DNS resolution fails, SSH
logins still work. If you do not disable sshd checks, DNS resolution failures prevent all logins.

To disable sshd DNS checks

1. Open the /etc/ssh/sshd_config file with a text editor and locate the following line:

120
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Paid AMIs

#UseDNS yes

2. Change the line to:

UseDNS no

Note
The location of this configuration file can differ for your distribution or if you are not running
OpenSSH. If this is the case, consult the relevant documentation.

Identify yourself
Currently, there is no easy way to know who provided a shared AMI, because each AMI is represented by
an account ID.

We recommend that you post a description of your AMI, and the AMI ID, in the Amazon EC2 forum. This
provides a convenient central location for users who are interested in trying new shared AMIs.

Protect yourself
We recommend against storing sensitive data or software on any AMI that you share. Users who launch a
shared AMI might be able to rebundle it and register it as their own. Follow these guidelines to help you
to avoid some easily overlooked security risks:

• We recommend using the --exclude directory option on ec2-bundle-vol to skip any

directories and subdirectories that contain secret information that you would not like to include
in your bundle. In particular, exclude all user-owned SSH public/private key pairs and SSH
authorized_keys files when bundling the image. The Amazon public AMIs store these in /
root/.ssh for the root account, and /home/user_name/.ssh/ for regular user accounts. For more
information, see ec2-bundle-vol (p. 148).
• Always delete the shell history before bundling. If you attempt more than one bundle upload in the
same AMI, the shell history contains your secret access key. The following example should be the last
command you run before bundling from within the instance.

[ec2-user ~]$ shred -u ~/.*history

Warning
The limitations of shred described in the warning above apply here as well.
Be aware that bash writes the history of the current session to the disk on exit. If you log out
of your instance after deleting ~/.bash_history, and then log back in, you will find that
~/.bash_history has been re-created and contains all of the commands you ran during
your previous session.
Other programs besides bash also write histories to disk, Use caution and remove or exclude
unnecessary dot-files and dot-directories.
• Bundling a running instance requires your private key and X.509 certificate. Put these and other
credentials in a location that is not bundled (such as the instance store).

Paid AMIs
A paid AMI is an AMI that you can purchase from a developer.

121
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Sell your AMI

Amazon EC2 integrates with AWS Marketplace, enabling developers to charge other Amazon EC2 users
for the use of their AMIs or to provide support for instances.

The AWS Marketplace is an online store where you can buy software that runs on AWS, including AMIs
that you can use to launch your EC2 instance. The AWS Marketplace AMIs are organized into categories,
such as Developer Tools, to enable you to find products to suit your requirements. For more information
about AWS Marketplace, see the AWS Marketplace site.

Launching an instance from a paid AMI is the same as launching an instance from any other AMI. No
additional parameters are required. The instance is charged according to the rates set by the owner of
the AMI, as well as the standard usage fees for the related web services, for example, the hourly rate for
running an m1.small instance type in Amazon EC2. Additional taxes might also apply. The owner of the
paid AMI can confirm whether a specific instance was launched using that paid AMI.
Important
Amazon DevPay is no longer accepting new sellers or products. AWS Marketplace is now
the single, unified e-commerce platform for selling software and services through AWS. For
information about how to deploy and sell software from AWS Marketplace, see Selling on AWS
Marketplace. AWS Marketplace supports AMIs backed by Amazon EBS.

Contents
• Sell your AMI (p. 122)
• Find a paid AMI (p. 122)
• Purchase a paid AMI (p. 123)
• Get the product code for your instance (p. 124)
• Use paid support (p. 124)
• Bills for paid and supported AMIs (p. 125)
• Manage your AWS Marketplace subscriptions (p. 125)

Sell your AMI

You can sell your AMI using AWS Marketplace. AWS Marketplace offers an organized shopping
experience. Additionally, AWS Marketplace also supports AWS features such as Amazon EBS-backed
AMIs, Reserved Instances, and Spot Instances.

For information about how to sell your AMI on AWS Marketplace, see Selling on AWS Marketplace.

Find a paid AMI

There are several ways that you can find AMIs that are available for you to purchase. For example, you
can use AWS Marketplace, the Amazon EC2 console, or the command line. Alternatively, a developer
might let you know about a paid AMI themselves.

Find a paid AMI using the console

To find a paid AMI using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose AMIs.
3. Choose Public images for the first filter.
4. In the Search bar, choose Owner, then AWS Marketplace.
5. If you know the product code, choose Product Code, then type the product code.

122
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Purchase a paid AMI

Find a paid AMI using AWS Marketplace

To find a paid AMI using AWS Marketplace

1. Open AWS Marketplace.

2. Enter the name of the operating system in the search box, and click Go.
3. To scope the results further, use one of the categories or filters.
4. Each product is labeled with its product type: either AMI or Software as a Service.

Find a paid AMI using the AWS CLI

You can find a paid AMI using the following describe-images command (AWS CLI).

aws ec2 describe-images

--owners aws-marketplace

This command returns numerous details that describe each AMI, including the product code for a paid
AMI. The output from describe-images includes an entry for the product code like the following:

"ProductCodes": [
{
"ProductCodeId": "product_code",
"ProductCodeType": "marketplace"
}
],

If you know the product code, you can filter the results by product code. This example returns the most
recent AMI with the specified product code.

aws ec2 describe-images

--owners aws-marketplace \
--filters "Name=product-code,Values=product_code" \
--query "sort_by(Images, &CreationDate)[-1].[ImageId]"

Purchase a paid AMI

You must sign up for (purchase) a paid AMI before you can launch an instance using the AMI.

Typically a seller of a paid AMI presents you with information about the AMI, including its price and a
link where you can buy it. When you click the link, you're first asked to log into AWS, and then you can
purchase the AMI.

Purchase a paid AMI using the console

You can purchase a paid AMI by using the Amazon EC2 launch wizard. For more information, see Launch
an AWS Marketplace instance (p. 546).

Subscribe to a product using AWS Marketplace

To use the AWS Marketplace, you must have an AWS account. To launch instances from AWS Marketplace
products, you must be signed up to use the Amazon EC2 service, and you must be subscribed to the

123
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get the product code for your instance

product from which to launch the instance. There are two ways to subscribe to products in the AWS
Marketplace:

• AWS Marketplace website: You can launch preconfigured software quickly with the 1-Click
deployment feature.
• Amazon EC2 launch wizard: You can search for an AMI and launch an instance directly from the
wizard. For more information, see Launch an AWS Marketplace instance (p. 546).

Get the product code for your instance

You can retrieve the AWS Marketplace product code for your instance using its instance metadata. For
more information about retrieving metadata, see Instance metadata and user data (p. 704).

To retrieve a product code, use the following command:

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" –v http://169.254.169.254/latest/meta-
data/product-codes

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/product-codes

If the instance has a product code, Amazon EC2 returns it.

Use paid support

Amazon EC2 also enables developers to offer support for software (or derived AMIs). Developers can
create support products that you can sign up to use. During sign-up for the support product, the
developer gives you a product code, which you must then associate with your own AMI. This enables
the developer to confirm that your instance is eligible for support. It also ensures that when you run
instances of the product, you are charged according to the terms for the product specified by the
developer.
Important
You can't use a support product with Reserved Instances. You always pay the price that's
specified by the seller of the support product.

To associate a product code with your AMI, use one of the following commands, where ami_id is the ID of
the AMI and product_code is the product code:

• modify-image-attribute (AWS CLI)

aws ec2 modify-image-attribute --image-id ami_id --product-codes "product_code"

• Edit-EC2ImageAttribute (AWS Tools for Windows PowerShell)

PS C:\> Edit-EC2ImageAttribute -ImageId ami_id -ProductCode product_code

After you set the product code attribute, it cannot be changed or removed.

124
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Bills for paid and supported AMIs

Bills for paid and supported AMIs

At the end of each month, you receive an email with the amount your credit card has been charged for
using any paid or supported AMIs during the month. This bill is separate from your regular Amazon EC2
bill. For more information, see Paying For AWS Marketplace Products.

Manage your AWS Marketplace subscriptions

On the AWS Marketplace website, you can check your subscription details, view the vendor's usage
instructions, manage your subscriptions, and more.

To check your subscription details

1. Log in to the AWS Marketplace.

2. Choose Your Marketplace Account.
3. Choose Manage your software subscriptions.
4. All your current subscriptions are listed. Choose Usage Instructions to view specific instructions for
using the product, for example, a user name for connecting to your running instance.

To cancel an AWS Marketplace subscription

1. Ensure that you have terminated any instances running from the subscription.

a. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

b. In the navigation pane, choose Instances.
c. Select the instance, and choose Actions, Instance State, Terminate.
d. Choose Yes, Terminate when prompted for confirmation.
2. Log in to the AWS Marketplace, and choose Your Marketplace Account, then Manage your software
subscriptions.
3. Choose Cancel subscription. You are prompted to confirm your cancellation.
Note
After you've canceled your subscription, you are no longer able to launch any instances
from that AMI. To use that AMI again, you need to resubscribe to it, either on the AWS
Marketplace website, or through the launch wizard in the Amazon EC2 console.

Create an Amazon EBS-backed Linux AMI

To create an Amazon EBS-backed Linux AMI, start from an instance that you've launched from an existing
Amazon EBS-backed Linux AMI. This can be an AMI you have obtained from the AWS Marketplace, an
AMI you have created using the AWS Server Migration Service or VM Import/Export, or any other AMI
you can access. After you customize the instance to suit your needs, create and register a new AMI, which
you can use to launch new instances with these customizations.

The procedures described below work for Amazon EC2 instances backed by encrypted Amazon Elastic
Block Store (Amazon EBS) volumes (including the root volume) as well as for unencrypted volumes.

The AMI creation process is different for instance store-backed AMIs. For more information about the
differences between Amazon EBS-backed and instance store-backed instances, and how to determine the
root device type for your instance, see Storage for the root device (p. 102). For more information about
creating an instance store-backed Linux AMI, see Create an instance store-backed Linux AMI (p. 130).

For more information about creating an Amazon EBS-backed Windows AMI, see Creating an Amazon
EBS-Backed Windows AMI in the Amazon EC2 User Guide for Windows Instances.

125
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Overview of creating Amazon EBS-backed AMIs

Overview of creating Amazon EBS-backed AMIs

First, launch an instance from an AMI that's similar to the AMI that you'd like to create. You can connect
to your instance and customize it. When the instance is configured correctly, ensure data integrity by
stopping the instance before you create an AMI, then create the image. When you create an Amazon
EBS-backed AMI, we automatically register it for you.

Amazon EC2 powers down the instance before creating the AMI to ensure that everything on the
instance is stopped and in a consistent state during the creation process. If you're confident that your
instance is in a consistent state appropriate for AMI creation, you can tell Amazon EC2 not to power
down and reboot the instance. Some file systems, such as XFS, can freeze and unfreeze activity, making it
safe to create the image without rebooting the instance.

During the AMI-creation process, Amazon EC2 creates snapshots of your instance's root volume and any
other EBS volumes attached to your instance. You're charged for the snapshots until you deregister the
AMI and delete the snapshots. For more information, see Deregister your Linux AMI (p. 175). If any
volumes attached to the instance are encrypted, the new AMI only launches successfully on instances
that support Amazon EBS encryption. For more information, see Amazon EBS encryption (p. 1176).

Depending on the size of the volumes, it can take several minutes for the AMI-creation process to
complete (sometimes up to 24 hours). You may find it more efficient to create snapshots of your volumes
before creating your AMI. This way, only small, incremental snapshots need to be created when the AMI
is created, and the process completes more quickly (the total time for snapshot creation remains the
same). For more information, see Create Amazon EBS snapshots (p. 1129).

After the process completes, you have a new AMI and snapshot created from the root volume of the
instance. When you launch an instance using the new AMI, we create a new EBS volume for its root
volume using the snapshot.

If you add instance-store volumes or EBS volumes to your instance in addition to the root device volume,
the block device mapping for the new AMI contains information for these volumes, and the block device
mappings for instances that you launch from the new AMI automatically contain information for these
volumes. The instance-store volumes specified in the block device mapping for the new instance are new
and don't contain any data from the instance store volumes of the instance you used to create the AMI.
The data on EBS volumes persists. For more information, see Block device mapping (p. 1294).
Note
When you create a new instance from an EBS-backed AMI, you should initialize both its root
volume and any additional EBS storage before putting it into production. For more information,
see Initializing Amazon EBS volumes.

Create a Linux AMI from an instance

You can create an AMI using the AWS Management Console or the command line. The following diagram
summarizes the process for creating an Amazon EBS-backed AMI from a running EC2 instance. Start with
an existing AMI, launch an instance, customize it, create a new AMI from it, and finally launch an instance
of your new AMI. The steps in the following diagram match the steps in the procedure below.

126
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create a Linux AMI from an instance

New console

To create an AMI from an instance using the console

1. Select an appropriate EBS-backed AMI to serve as a starting point for your new AMI, and
configure it as needed before launch. For more information, see Launch an instance using the
Launch Instance Wizard (p. 526).
2. Choose Launch to launch an instance of the EBS-backed AMI that you've selected. Accept the
default values as you step through the wizard. For more information, see Launch an instance
using the Launch Instance Wizard (p. 526).
3. While the instance is running, connect to it. You can perform any of the following actions on
your instance to customize it for your needs:

• Install software and applications

• Copy data
• Reduce start time by deleting temporary files, defragmenting your hard drive, and zeroing out
free space
• Attach additional EBS volumes
4. (Optional) Create snapshots of all the volumes attached to your instance. For more information
about creating snapshots, see Create Amazon EBS snapshots (p. 1129).
5. In the navigation pane, choose Instances, select your instance, and then choose Actions, Image
and templates, Create image.
Tip
If this option is disabled, your instance isn't an Amazon EBS-backed instance.
6. On the Create image page, specify the following information, and then choose Create image.

• Image name – A unique name for the image.

• Image description – An optional description of the image, up to 255 characters.
• No reboot – This option is not selected by default. Amazon EC2 shuts down the instance,
takes snapshots of any attached volumes, creates and registers the AMI, and then reboots the
instance. Select No reboot to avoid having your instance shut down.
Warning
If you select No reboot, we can't guarantee the file system integrity of the created
image.
• Instance volumes – The fields in this section enable you to modify the root volume, and add
additional Amazon EBS and instance store volumes.
• The root volume is defined in the first row. To change the size of the root volume, for Size,
enter the required value.
• If you select Delete on termination, when you terminate the instance created from this
AMI, the EBS volume is deleted. If you clear Delete on termination, when you terminate the
instance, the EBS volume is not deleted. For more information, see Preserve Amazon EBS
volumes on instance termination (p. 646).
• To add an EBS volume, choose Add volume (which adds a new row). For Volume type,
choose EBS, and fill in the fields in the row. When you launch an instance from your new
AMI, additional volumes are automatically attached to the instance. Empty volumes must
be formatted and mounted. Volumes based on a snapshot must be mounted.
• To add an instance store volume, see Add instance store volumes to an AMI (p. 1277). When
you launch an instance from your new AMI, additional volumes are automatically initialized
and mounted. These volumes do not contain data from the instance store volumes of the
running instance on which you based your AMI.
• Tags – You can tag the AMI and the snapshots with the same tags, or you can tag them with
different tags.
127
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create a Linux AMI from an instance

• To tag the AMI and the snapshots with the same tags, choose Tag image and snapshots
together. The same tags are applied to the AMI and every snapshot that is created.
• To tag the AMI and the snapshots with different tags, choose Tag image and snapshots
separately. Different tags are applied to the AMI and the snapshots that are created.
However, all the snapshots get the same tags; you can't tag each snapshot with a different
tag.

To add a tag, choose Add tag, and enter the key and value for the tag. Repeat for each tag.
7. To view the status of your AMI while it is being created, in the navigation pane, choose AMIs.
Initially, the status is pending but should change to available after a few minutes.

(Optional) To view the snapshot that was created for the new AMI, choose Snapshots. When you
launch an instance from this AMI, we use this snapshot to create its root device volume.
8. Launch an instance from your new AMI. For more information, see Launch an instance using the
Launch Instance Wizard (p. 526).
9. The new running instance contains all of the customizations that you applied in previous steps.

Old console

To create an AMI from an instance using the console

1. Select an appropriate EBS-backed AMI to serve as a starting point for your new AMI, and
configure it as needed before launch. For more information, see Launch an instance using the
Launch Instance Wizard (p. 526).
2. Choose Launch to launch an instance of the EBS-backed AMI that you've selected. Accept the
default values as you step through the wizard. For more information, see Launch an instance
using the Launch Instance Wizard (p. 526).
3. While the instance is running, connect to it. You can perform any of the following actions on
your instance to customize it for your needs:

• Install software and applications

• Copy data
• Reduce start time by deleting temporary files, defragmenting your hard drive, and zeroing out
free space
• Attach additional EBS volumes
4. (Optional) Create snapshots of all the volumes attached to your instance. For more information
about creating snapshots, see Create Amazon EBS snapshots (p. 1129).
5. In the navigation pane, choose Instances, select your instance, and then choose Actions, Image,
Create Image.
Tip
If this option is disabled, your instance isn't an Amazon EBS-backed instance.
6. In the Create Image dialog box, specify the following information, and then choose Create
Image.

• Image name – A unique name for the image.

• Image description – An optional description of the image, up to 255 characters.
• No reboot – This option is not selected by default. Amazon EC2 shuts down the instance,
takes snapshots of any attached volumes, creates and registers the AMI, and then reboots the
instance. Select No reboot to avoid having your instance shut down.
Warning
If you select No reboot, we can't guarantee the file system integrity of the created
image.

128
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create a Linux AMI from a snapshot

• Instance Volumes – The fields in this section enable you to modify the root volume, and add
additional Amazon EBS and instance store volumes. For information about each field, pause
on the i icon next to each field to display field tooltips. Some important points are listed
below.
• To change the size of the root volume, locate Root in the Volume Type column, and for
Size (GiB), type the required value.
• If you select Delete on Termination, when you terminate the instance created from this
AMI, the EBS volume is deleted. If you clear Delete on Termination, when you terminate
the instance, the EBS volume is not deleted. For more information, see Preserve Amazon
EBS volumes on instance termination (p. 646).
• To add an EBS volume, choose Add New Volume (which adds a new row). For Volume Type,
choose EBS, and fill in the fields in the row. When you launch an instance from your new
AMI, additional volumes are automatically attached to the instance. Empty volumes must
be formatted and mounted. Volumes based on a snapshot must be mounted.
• To add an instance store volume, see Add instance store volumes to an AMI (p. 1277). When
you launch an instance from your new AMI, additional volumes are automatically initialized
and mounted. These volumes do not contain data from the instance store volumes of the
running instance on which you based your AMI.
7. To view the status of your AMI while it is being created, in the navigation pane, choose AMIs.
Initially, the status is pending but should change to available after a few minutes.

(Optional) To view the snapshot that was created for the new AMI, choose Snapshots. When you
launch an instance from this AMI, we use this snapshot to create its root device volume.
8. Launch an instance from your new AMI. For more information, see Launch an instance using the
Launch Instance Wizard (p. 526).
9. The new running instance contains all of the customizations that you applied in previous steps.

To create an AMI from an instance using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• create-image (AWS CLI)

• New-EC2Image (AWS Tools for Windows PowerShell)

Create a Linux AMI from a snapshot

If you have a snapshot of the root device volume of an instance, you can create an AMI from this
snapshot using the AWS Management Console or the command line.

To create an AMI from a snapshot using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, under Elastic Block Store, choose Snapshots.
3. Choose the snapshot and choose Actions, Create Image.
4. In the Create Image from EBS Snapshot dialog box, complete the fields to create your AMI, then
choose Create. If you're re-creating a parent instance, then choose the same options as the parent
instance.

• Architecture: Choose i386 for 32-bit or x86_64 for 64-bit.

• Root device name: Enter the appropriate name for the root volume. For more information, see
Name devices on Linux instances (p. 1292).

129
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch an instance from an AMI you created

• Virtualization type: Choose whether instances launched from this AMI use paravirtual (PV) or
hardware virtual machine (HVM) virtualization. For more information, see Linux AMI virtualization
types (p. 104).
• (PV virtualization type only) Kernel ID and RAM disk ID: Choose the AKI and ARI from the lists. If
you choose the default AKI or don't choose an AKI, you must specify an AKI every time you launch
an instance using this AMI. In addition, your instance may fail the health checks if the default AKI
is incompatible with the instance.
• (Optional) Block Device Mappings: Add volumes or expand the default size of the root volume for
the AMI. For more information about resizing the file system on your instance for a larger volume,
see Extend a Linux file system after resizing a volume (p. 1172).

To create an AMI from a snapshot using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• register-image (AWS CLI)

• Register-EC2Image (AWS Tools for Windows PowerShell)

Launch an instance from an AMI you created

You can launch an instance from an AMI that you created from an instance or snapshot.

To launch an instance from your AMI

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, under Images, choose AMIs.
3. Set the filter to Owned by me and select your AMI.
4. Choose Actions, Launch.
5. Follow the wizard to launch your instance. For more information about each step in the wizard, see
Launch an instance using the Launch Instance Wizard (p. 526).

Automate the EBS-backed AMI lifecycle

You can use Amazon Data Lifecycle Manager to automate the creation, retention, and deletion of
Amazon EBS-backed AMIs and their backing snapshots. For more information, see Amazon Data Lifecycle
Manager (p. 1186).

Create an instance store-backed Linux AMI

The AMI that you specify when you launch your instance determines the type of root device volume.

To create an instance store-backed Linux AMI, start from an instance that you've launched from an
existing instance store-backed Linux AMI. After you've customized the instance to suit your needs,
bundle the volume and register a new AMI, which you can use to launch new instances with these
customizations.
Important
Only the following instance types support an instance store volume as the root device: C3, D2,
G2, I2, M3, and R3.

130
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Overview of the creation process
for instance store-backed AMIs
The AMI creation process is different for Amazon EBS-backed AMIs. For more information about the
differences between Amazon EBS-backed and instance store-backed instances, and how to determine
the root device type for your instance, see Storage for the root device (p. 102). If you need to create an
Amazon EBS-backed Linux AMI, see Create an Amazon EBS-backed Linux AMI (p. 125).

Overview of the creation process for instance store-

backed AMIs
The following diagram summarizes the process of creating an AMI from an instance store-backed
instance.

First, launch an instance from an AMI that's similar to the AMI that you'd like to create. You can connect
to your instance and customize it. When the instance is set up the way you want it, you can bundle it.
It takes several minutes for the bundling process to complete. After the process completes, you have a
bundle, which consists of an image manifest (image.manifest.xml) and files (image.part.xx) that
contain a template for the root volume. Next you upload the bundle to your Amazon S3 bucket and then
register your AMI.

When you launch an instance using the new AMI, we create the root volume for the instance using
the bundle that you uploaded to Amazon S3. The storage space used by the bundle in Amazon S3
incurs charges to your account until you delete it. For more information, see Deregister your Linux
AMI (p. 175).

If you add instance store volumes to your instance in addition to the root device volume, the block device
mapping for the new AMI contains information for these volumes, and the block device mappings for
instances that you launch from the new AMI automatically contain information for these volumes. For
more information, see Block device mapping (p. 1294).

Prerequisites
Before you can create an AMI, you must complete the following tasks:

• Install the AMI tools. For more information, see Set up the AMI tools (p. 132).
• Install the AWS CLI. For more information, see Getting Set Up with the AWS Command Line Interface.
• Ensure that you have an Amazon S3 bucket for the bundle. To create an Amazon S3 bucket, open the
Amazon S3 console and click Create Bucket. Alternatively, you can use the AWS CLI mb command.
• Ensure that you have your AWS account ID. For more information, see AWS Account Identifiers in the
AWS General Reference.
• Ensure that you have your access key ID and secret access key. For more information, see Access Keys in
the AWS General Reference.

131
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Set up the AMI tools

• Ensure that you have an X.509 certificate and corresponding private key.
• If you need to create an X.509 certificate, see Manage signing certificates (p. 134). The X.509
certificate and private key are used to encrypt and decrypt your AMI.
• [China (Beijing)] Use the $EC2_AMITOOL_HOME/etc/ec2/amitools/cert-ec2-cn-
north-1.pem certificate.
• [AWS GovCloud (US-West)] Use the $EC2_AMITOOL_HOME/etc/ec2/amitools/cert-ec2-
gov.pem certificate.
• Connect to your instance and customize it. For example, you can install software and applications,
copy data, delete temporary files, and modify the Linux configuration.

Tasks

• Set up the AMI tools (p. 132)

• Create an AMI from an instance store-backed Amazon Linux instance (p. 135)
• Create an AMI from an instance store-backed Ubuntu instance (p. 138)
• Convert your instance store-backed AMI to an Amazon EBS-backed AMI (p. 142)

Set up the AMI tools

You can use the AMI tools to create and manage instance store-backed Linux AMIs. To use the tools, you
must install them on your Linux instance. The AMI tools are available as both an RPM and as a .zip file for
Linux distributions that don't support RPM.

To set up the AMI tools using the RPM

1. Install Ruby using the package manager for your Linux distribution, such as yum. For example:

[ec2-user ~]$ sudo yum install -y ruby

2. Download the RPM file using a tool such as wget or curl. For example:

[ec2-user ~]$ wget https://s3.amazonaws.com/ec2-downloads/ec2-ami-tools.noarch.rpm

3. Verify the RPM file's signature using the following command:

[ec2-user ~]$ rpm -K ec2-ami-tools.noarch.rpm

The command above should indicate that the file's SHA1 and MD5 hashes are OK. If the command
indicates that the hashes are NOT OK, use the following command to view the file's Header SHA1
and MD5 hashes:

[ec2-user ~]$ rpm -Kv ec2-ami-tools.noarch.rpm

Then, compare your file's Header SHA1 and MD5 hashes with the following verified AMI tools hashes
to confirm the file's authenticity:

• Header SHA1: a1f662d6f25f69871104e6a62187fa4df508f880

• MD5: 9faff05258064e2f7909b66142de6782

If your file's Header SHA1 and MD5 hashes match the verified AMI tools hashes, continue to the next
step.

132
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Set up the AMI tools

4. Install the RPM using the following command:

[ec2-user ~]$ sudo yum install ec2-ami-tools.noarch.rpm

5. Verify your AMI tools installation using the ec2-ami-tools-version (p. 145) command.

[ec2-user ~]$ ec2-ami-tools-version

Note
If you receive a load error such as "cannot load such file -- ec2/amitools/version
(LoadError)", complete the next step to add the location of your AMI tools installation to
your RUBYLIB path.
6. (Optional) If you received an error in the previous step, add the location of your AMI tools
installation to your RUBYLIB path.

a. Run the following command to determine the paths to add.

[ec2-user ~]$ rpm -qil ec2-ami-tools | grep ec2/amitools/version

/usr/lib/ruby/site_ruby/ec2/amitools/version.rb
/usr/lib64/ruby/site_ruby/ec2/amitools/version.rb

In the above example, the missing file from the previous load error is located at /usr/lib/
ruby/site_ruby and /usr/lib64/ruby/site_ruby.
b. Add the locations from the previous step to your RUBYLIB path.

[ec2-user ~]$ export RUBYLIB=$RUBYLIB:/usr/lib/ruby/site_ruby:/usr/lib64/ruby/

site_ruby

c. Verify your AMI tools installation using the ec2-ami-tools-version (p. 145) command.

[ec2-user ~]$ ec2-ami-tools-version

To set up the AMI tools using the .zip file

1. Install Ruby and unzip using the package manager for your Linux distribution, such as apt-get. For
example:

[ec2-user ~]$ sudo apt-get update -y && sudo apt-get install -y ruby unzip

2. Download the .zip file using a tool such as wget or curl. For example:

[ec2-user ~]$ wget https://s3.amazonaws.com/ec2-downloads/ec2-ami-tools.zip

3. Unzip the files into a suitable installation directory, such as /usr/local/ec2.

[ec2-user ~]$ sudo mkdir -p /usr/local/ec2

$ sudo unzip ec2-ami-tools.zip -d /usr/local/ec2

Notice that the .zip file contains a folder ec2-ami-tools-x.x.x, where x.x.x is the version number of
the tools (for example, ec2-ami-tools-1.5.7).
4. Set the EC2_AMITOOL_HOME environment variable to the installation directory for the tools. For
example:

133
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Set up the AMI tools

[ec2-user ~]$ export EC2_AMITOOL_HOME=/usr/local/ec2/ec2-ami-tools-x.x.x

5. Add the tools to your PATH environment variable. For example:

[ec2-user ~]$ export PATH=$EC2_AMITOOL_HOME/bin:$PATH

6. You can verify your AMI tools installation using the ec2-ami-tools-version (p. 145) command.

[ec2-user ~]$ ec2-ami-tools-version

Manage signing certificates

Certain commands in the AMI tools require a signing certificate (also known as X.509 certificate). You
must create the certificate and then upload it to AWS. For example, you can use a third-party tool such
as OpenSSL to create the certificate.

To create a signing certificate

1. Install and configure OpenSSL.

2. Create a private key using the openssl genrsa command and save the output to a .pem file. We
recommend that you create a 2048- or 4096-bit RSA key.

openssl genrsa 2048 > private-key.pem

3. Generate a certificate using the openssl req command.

openssl req -new -x509 -nodes -sha256 -days 365 -key private-key.pem -outform PEM -
out certificate.pem

To upload the certificate to AWS, use the upload-signing-certificate command.

aws iam upload-signing-certificate --user-name user-name --certificate-body file://path/to/

certificate.pem

To list the certificates for a user, use the list-signing-certificates command:

aws iam list-signing-certificates --user-name user-name

To disable or re-enable a signing certificate for a user, use the update-signing-certificate command. The
following command disables the certificate:

aws iam update-signing-certificate --certificate-id OFHPLP4ZULTHYPMSYEX7O4BEXAMPLE --

status Inactive --user-name user-name

To delete a certificate, use the delete-signing-certificate command:

aws iam delete-signing-certificate --user-name user-name --certificate-

id OFHPLP4ZULTHYPMSYEX7O4BEXAMPLE

134
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create an AMI from an instance store-backed instance

Create an AMI from an instance store-backed

instance
The following procedures are for creating an instance store-backed AMI from an instance store-backed
instance. Before you begin, ensure that you've read the Prerequisites (p. 131).

Topics
• Create an AMI from an instance store-backed Amazon Linux instance (p. 135)
• Create an AMI from an instance store-backed Ubuntu instance (p. 138)

Create an AMI from an instance store-backed Amazon Linux

instance
This section describes the creation of an AMI from an Amazon Linux instance. The following procedures
may not work for instances running other Linux distributions. For Ubuntu-specific procedures, see Create
an AMI from an instance store-backed Ubuntu instance (p. 138).

To prepare to use the AMI tools (HVM instances only)

1. The AMI tools require GRUB Legacy to boot properly. Use the following command to install GRUB:

[ec2-user ~]$ sudo yum install -y grub

2. Install the partition management packages with the following command:

[ec2-user ~]$ sudo yum install -y gdisk kpartx parted

To create an AMI from an instance store-backed Amazon Linux instance

This procedure assumes that you have satisfied the prerequisites in Prerequisites (p. 131).

1. Upload your credentials to your instance. We use these credentials to ensure that only you and
Amazon EC2 can access your AMI.

a. Create a temporary directory on your instance for your credentials as follows:

[ec2-user ~]$ mkdir /tmp/cert

This enables you to exclude your credentials from the created image.
b. Copy your X.509 certificate and corresponding private key from your computer to the /tmp/
cert directory on your instance using a secure copy tool such as scp (p. 602). The -i my-
private-key.pem option in the following scp command is the private key you use to connect
to your instance with SSH, not the X.509 private key. For example:

you@your_computer:~ $ scp -i my-private-key.pem /

path/to/pk-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem /
path/to/cert-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem ec2-
[email protected]:/tmp/cert/
pk-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem 100% 717 0.7KB/s 00:00
cert-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem 100% 685 0.7KB/s 00:00

135
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create an AMI from an instance store-backed instance

Alternatively, because these are plain text files, you can open the certificate and key in a text editor
and copy their contents into new files in /tmp/cert.
2. Prepare the bundle to upload to Amazon S3 by running the ec2-bundle-vol (p. 148) command
from inside your instance. Be sure to specify the -e option to exclude the directory where your
credentials are stored. By default, the bundle process excludes files that might contain sensitive
information. These files include *.sw, *.swo, *.swp, *.pem, *.priv, *id_rsa*, *id_dsa*
*.gpg, *.jks, */.ssh/authorized_keys, and */.bash_history. To include all of these files,
use the --no-filter option. To include some of these files, use the --include option.
Important
By default, the AMI bundling process creates a compressed, encrypted collection of files
in the /tmp directory that represents your root volume. If you do not have enough free
disk space in /tmp to store the bundle, you need to specify a different location for the
bundle to be stored with the -d /path/to/bundle/storage option. Some instances
have ephemeral storage mounted at /mnt or /media/ephemeral0 that you can use, or
you can also create (p. 1104), attach (p. 1106), and mount (p. 1111) a new Amazon Elastic
Block Store (Amazon EBS) volume to store the bundle.

a. You must run the ec2-bundle-vol command as root. For most commands, you can use sudo to
gain elevated permissions, but in this case, you should run sudo -E su to keep your environment
variables.

[ec2-user ~]$ sudo -E su

Note that bash prompt now identifies you as the root user, and that the dollar sign has been
replaced by a hash tag, signalling that you are in a root shell:

[root ec2-user]#

b. To create the AMI bundle, run the ec2-bundle-vol (p. 148) command as follows:

[root ec2-user]# ec2-bundle-vol -k /tmp/cert/pk-

HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem -c /tmp/cert/cert-
HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem -u 123456789012 -r x86_64 -e /tmp/cert --
partition gpt

Note
For the China (Beijing) and AWS GovCloud (US-West) Regions, use the --ec2cert
parameter and specify the certificates as per the prerequisites (p. 131).

It can take a few minutes to create the image. When this command completes, your /tmp
(or non-default) directory contains the bundle (image.manifest.xml, plus multiple
image.part.xx files).
c. Exit from the root shell.

[root ec2-user]# exit

3. (Optional) To add more instance store volumes, edit the block device mappings in the
image.manifest.xml file for your AMI. For more information, see Block device mapping (p. 1294).

a. Create a backup of your image.manifest.xml file.

[ec2-user ~]$ sudo cp /tmp/image.manifest.xml /tmp/image.manifest.xml.bak

b. Reformat the image.manifest.xml file so that it is easier to read and edit.

136
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create an AMI from an instance store-backed instance

[ec2-user ~]$ sudo xmllint --format /tmp/image.manifest.xml.bak > sudo /tmp/

image.manifest.xml

c. Edit the block device mappings in image.manifest.xml with a text editor. The example below
shows a new entry for the ephemeral1 instance store volume.
Note
For a list of excluded files, see ec2-bundle-vol (p. 148).

<block_device_mapping>
<mapping>
<virtual>ami</virtual>
<device>sda</device>
</mapping>
<mapping>
<virtual>ephemeral0</virtual>
<device>sdb</device>
</mapping>
<mapping>
<virtual>ephemeral1</virtual>
<device>sdc</device>
</mapping>
<mapping>
<virtual>root</virtual>
<device>/dev/sda1</device>
</mapping>
</block_device_mapping>

d. Save the image.manifest.xml file and exit your text editor.

4. To upload your bundle to Amazon S3, run the ec2-upload-bundle (p. 158) command as follows.

[ec2-user ~]$ ec2-upload-bundle -b my-s3-bucket/bundle_folder/bundle_name -m /tmp/

image.manifest.xml -a your_access_key_id -s your_secret_access_key

Important
To register your AMI in a Region other than US East (N. Virginia), you must specify both the
target Region with the --region option and a bucket path that already exists in the target
Region or a unique bucket path that can be created in the target Region.
5. (Optional) After the bundle is uploaded to Amazon S3, you can remove the bundle from the /tmp
directory on the instance using the following rm command:

[ec2-user ~]$ sudo rm /tmp/image.manifest.xml /tmp/image.part.* /tmp/image

Important
If you specified a path with the -d /path/to/bundle/storage option in Step
2 (p. 136), use that path instead of /tmp.
6. To register your AMI, run the register-image command as follows.

[ec2-user ~]$ aws ec2 register-image --image-location my-s3-

bucket/bundle_folder/bundle_name/image.manifest.xml --name AMI_name --virtualization-
type hvm

Important
If you previously specified a Region for the ec2-upload-bundle (p. 158) command, specify
that Region again for this command.

137
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create an AMI from an instance store-backed instance

Create an AMI from an instance store-backed Ubuntu instance

This section describes the creation of an AMI from an Ubuntu Linux instance with an instance store
volume as the root volume. The following procedures may not work for instances running other Linux
distributions. For procedures specific to Amazon Linux, see Create an AMI from an instance store-backed
Amazon Linux instance (p. 135).

To prepare to use the AMI tools (HVM instances only)

The AMI tools require GRUB Legacy to boot properly. However, Ubuntu is configured to use GRUB 2. You
must check to see that your instance uses GRUB Legacy, and if not, you need to install and configure it.

HVM instances also require partitioning tools to be installed for the AMI tools to work properly.

1. GRUB Legacy (version 0.9x or less) must be installed on your instance. Check to see if GRUB Legacy
is present and install it if necessary.

a. Check the version of your GRUB installation.

ubuntu:~$ grub-install --version

grub-install (GRUB) 1.99-21ubuntu3.10

In this example, the GRUB version is greater than 0.9x, so you must install GRUB Legacy.
Proceed to Step 1.b (p. 138). If GRUB Legacy is already present, you can skip to Step
2 (p. 138).
b. Install the grub package using the following command.

ubuntu:~$ sudo apt-get install -y grub

2. Install the following partition management packages using the package manager for your
distribution.

• gdisk (some distributions may call this package gptfdisk instead)

• kpartx
• parted

Use the following command.

ubuntu:~$ sudo apt-get install -y gdisk kpartx parted

3. Check the kernel parameters for your instance.

ubuntu:~$ cat /proc/cmdline

BOOT_IMAGE=/boot/vmlinuz-3.2.0-54-virtual root=UUID=4f392932-ed93-4f8f-
aee7-72bc5bb6ca9d ro console=ttyS0 xen_emul_unplug=unnecessary

Note the options following the kernel and root device parameters: ro, console=ttyS0, and
xen_emul_unplug=unnecessary. Your options may differ.
4. Check the kernel entries in /boot/grub/menu.lst.

ubuntu:~$ grep ^kernel /boot/grub/menu.lst

kernel /boot/vmlinuz-3.2.0-54-virtual root=LABEL=cloudimg-rootfs ro console=hvc0
kernel /boot/vmlinuz-3.2.0-54-virtual root=LABEL=cloudimg-rootfs ro single
kernel /boot/memtest86+.bin

138
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create an AMI from an instance store-backed instance

Note that the console parameter is pointing to hvc0 instead of ttyS0 and that the
xen_emul_unplug=unnecessary parameter is missing. Again, your options may differ.
5. Edit the /boot/grub/menu.lst file with your favorite text editor (such as vim or nano) to change
the console and add the parameters you identified earlier to the boot entries.

title Ubuntu 12.04.3 LTS, kernel 3.2.0-54-virtual

root (hd0)
kernel /boot/vmlinuz-3.2.0-54-virtual root=LABEL=cloudimg-rootfs
ro console=ttyS0 xen_emul_unplug=unnecessary
initrd /boot/initrd.img-3.2.0-54-virtual

title Ubuntu 12.04.3 LTS, kernel 3.2.0-54-virtual (recovery mode)

root (hd0)
kernel /boot/vmlinuz-3.2.0-54-virtual root=LABEL=cloudimg-rootfs ro
single console=ttyS0 xen_emul_unplug=unnecessary
initrd /boot/initrd.img-3.2.0-54-virtual

title Ubuntu 12.04.3 LTS, memtest86+

root (hd0)
kernel /boot/memtest86+.bin

6. Verify that your kernel entries now contain the correct parameters.

ubuntu:~$ grep ^kernel /boot/grub/menu.lst

kernel /boot/vmlinuz-3.2.0-54-virtual root=LABEL=cloudimg-rootfs ro console=ttyS0
xen_emul_unplug=unnecessary
kernel /boot/vmlinuz-3.2.0-54-virtual root=LABEL=cloudimg-rootfs ro single
console=ttyS0 xen_emul_unplug=unnecessary
kernel /boot/memtest86+.bin

7. [For Ubuntu 14.04 and later only] Starting with Ubuntu 14.04, instance store backed Ubuntu AMIs
use a GPT partition table and a separate EFI partition mounted at /boot/efi. The ec2-bundle-vol
command will not bundle this boot partition, so you need to comment out the /etc/fstab entry
for the EFI partition as shown in the following example.

LABEL=cloudimg-rootfs / ext4 defaults 0 0

#LABEL=UEFI /boot/efi vfat defaults 0 0
/dev/xvdb /mnt auto defaults,nobootwait,comment=cloudconfig 0 2

To create an AMI from an instance store-backed Ubuntu instance

This procedure assumes that you have satisfied the prerequisites in Prerequisites (p. 131).

1. Upload your credentials to your instance. We use these credentials to ensure that only you and
Amazon EC2 can access your AMI.

a. Create a temporary directory on your instance for your credentials as follows:

ubuntu:~$ mkdir /tmp/cert

This enables you to exclude your credentials from the created image.
b. Copy your X.509 certificate and private key from your computer to the /tmp/cert directory on
your instance, using a secure copy tool such as scp (p. 602). The -i my-private-key.pem
option in the following scp command is the private key you use to connect to your instance with
SSH, not the X.509 private key. For example:

139
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create an AMI from an instance store-backed instance

you@your_computer:~ $ scp -i my-private-key.pem /

path/to/pk-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem /
path/to/cert-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem ec2-
[email protected]:/tmp/cert/
pk-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem 100% 717 0.7KB/s 00:00
cert-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem 100% 685 0.7KB/s 00:00

Alternatively, because these are plain text files, you can open the certificate and key in a text editor
and copy their contents into new files in /tmp/cert.
2. Prepare the bundle to upload to Amazon S3 by running the ec2-bundle-vol (p. 148) command
from your instance. Be sure to specify the -e option to exclude the directory where your credentials
are stored. By default, the bundle process excludes files that might contain sensitive information.
These files include *.sw, *.swo, *.swp, *.pem, *.priv, *id_rsa*, *id_dsa* *.gpg, *.jks,
*/.ssh/authorized_keys, and */.bash_history. To include all of these files, use the --no-
filter option. To include some of these files, use the --include option.
Important
By default, the AMI bundling process creates a compressed, encrypted collection of files
in the /tmp directory that represents your root volume. If you do not have enough free
disk space in /tmp to store the bundle, you need to specify a different location for the
bundle to be stored with the -d /path/to/bundle/storage option. Some instances
have ephemeral storage mounted at /mnt or /media/ephemeral0 that you can use, or
you can also create (p. 1104), attach (p. 1106), and mount (p. 1111) a new Amazon Elastic
Block Store (Amazon EBS) volume to store the bundle.

a. You must run the ec2-bundle-vol command needs as root. For most commands, you can use
sudo to gain elevated permissions, but in this case, you should run sudo -E su to keep your
environment variables.

ubuntu:~$ sudo -E su

Note that bash prompt now identifies you as the root user, and that the dollar sign has been
replaced by a hash tag, signalling that you are in a root shell:

root@ubuntu:#

b. To create the AMI bundle, run the ec2-bundle-vol (p. 148) command as follows.

root@ubuntu:# ec2-bundle-vol -k /tmp/cert/pk-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem

-c /tmp/cert/cert-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem -u your_aws_account_id -r
x86_64 -e /tmp/cert --partition gpt

Important
For Ubuntu 14.04 and later HVM instances, add the --partition mbr flag to bundle
the boot instructions properly; otherwise, your newly-created AMI will not boot.

It can take a few minutes to create the image. When this command completes, your tmp
directory contains the bundle (image.manifest.xml, plus multiple image.part.xx files).
c. Exit from the root shell.

root@ubuntu:# exit

3. (Optional) To add more instance store volumes, edit the block device mappings in the
image.manifest.xml file for your AMI. For more information, see Block device mapping (p. 1294).

140
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create an AMI from an instance store-backed instance

a. Create a backup of your image.manifest.xml file.

ubuntu:~$ sudo cp /tmp/image.manifest.xml /tmp/image.manifest.xml.bak

b. Reformat the image.manifest.xml file so that it is easier to read and edit.

ubuntu:~$ sudo xmllint --format /tmp/image.manifest.xml.bak > /tmp/

image.manifest.xml

c. Edit the block device mappings in image.manifest.xml with a text editor. The example below
shows a new entry for the ephemeral1 instance store volume.

<block_device_mapping>
<mapping>
<virtual>ami</virtual>
<device>sda</device>
</mapping>
<mapping>
<virtual>ephemeral0</virtual>
<device>sdb</device>
</mapping>
<mapping>
<virtual>ephemeral1</virtual>
<device>sdc</device>
</mapping>
<mapping>
<virtual>root</virtual>
<device>/dev/sda1</device>
</mapping>
</block_device_mapping>

d. Save the image.manifest.xml file and exit your text editor.

4. To upload your bundle to Amazon S3, run the ec2-upload-bundle (p. 158) command as follows.

ubuntu:~$ ec2-upload-bundle -b my-s3-bucket/bundle_folder/bundle_name -m /tmp/

image.manifest.xml -a your_access_key_id -s your_secret_access_key

Important
If you intend to register your AMI in a Region other than US East (N. Virginia), you must
specify both the target Region with the --region option and a bucket path that already
exists in the target Region or a unique bucket path that can be created in the target Region.
5. (Optional) After the bundle is uploaded to Amazon S3, you can remove the bundle from the /tmp
directory on the instance using the following rm command:

ubuntu:~$ sudo rm /tmp/image.manifest.xml /tmp/image.part.* /tmp/image

Important
If you specified a path with the -d /path/to/bundle/storage option in Step
2 (p. 140), use that same path below, instead of /tmp.
6. To register your AMI, run the register-image AWS CLI command as follows.

ubuntu:~$ aws ec2 register-image --image-location my-s3-

bucket/bundle_folder/bundle_name/image.manifest.xml --name AMI_name --virtualization-
type hvm

141
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Convert to an Amazon EBS-Backed AMI

Important
If you previously specified a Region for the ec2-upload-bundle (p. 158) command, specify
that Region again for this command.
7. [Ubuntu 14.04 and later] Uncomment the EFI entry in /etc/fstab; otherwise, your running
instance will not be able to reboot.

Convert your instance store-backed AMI to an

Amazon EBS-backed AMI
You can convert an instance store-backed Linux AMI that you own to an Amazon EBS-backed Linux AMI.
Important
You can't convert an instance store-backed Windows AMI to an Amazon EBS-backed Windows
AMI and you cannot convert an AMI that you do not own.

To convert an instance store-backed AMI to an Amazon EBS-backed AMI

1. Launch an Amazon Linux instance from an Amazon EBS-backed AMI. For more information, see
Launch an instance using the Launch Instance Wizard (p. 526). Amazon Linux instances have the
AWS CLI and AMI tools pre-installed.
2. Upload the X.509 private key that you used to bundle your instance store-backed AMI to your
instance. We use this key to ensure that only you and Amazon EC2 can access your AMI.

a. Create a temporary directory on your instance for your X.509 private key as follows:

[ec2-user ~]$ mkdir /tmp/cert

b. Copy your X.509 private key from your computer to the /tmp/cert directory on your instance,
using a secure copy tool such as scp (p. 602). The my-private-key parameter in the
following command is the private key you use to connect to your instance with SSH. For
example:

you@your_computer:~ $ scp -i my-private-key.pem /

path/to/pk-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem ec2-
[email protected]:/tmp/cert/
pk-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem 100% 717 0.7KB/s 00:00

3. Set environment variables for your AWS access key and secret key.

[ec2-user ~]$ export AWS_ACCESS_KEY_ID=your_access_key_id

[ec2-user ~]$ export AWS_SECRET_ACCESS_KEY=your_secret_access_key

4. Prepare an Amazon Elastic Block Store (Amazon EBS) volume for your new AMI.

a. Create an empty EBS volume in the same Availability Zone as your instance using the create-
volume command. Note the volume ID in the command output.
Important
This EBS volume must be the same size or larger than the original instance store root
volume.

[ec2-user ~]$ aws ec2 create-volume --size 10 --region us-west-2 --availability-

zone us-west-2b

b. Attach the volume to your Amazon EBS-backed instance using the attach-volume command.

142
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Convert to an Amazon EBS-Backed AMI

[ec2-user ~]$ aws ec2 attach-volume --volume-id volume_id --instance-id instance_id

--device /dev/sdb --region us-west-2

5. Create a folder for your bundle.

[ec2-user ~]$ mkdir /tmp/bundle

6. Download the bundle for your instance store-based AMI to /tmp/bundle using the ec2-download-
bundle (p. 153) command.

[ec2-user ~]$ ec2-download-bundle -b my-s3-bucket/bundle_folder/bundle_name -m

image.manifest.xml -a $AWS_ACCESS_KEY_ID -s $AWS_SECRET_ACCESS_KEY --privatekey /path/
to/pk-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem -d /tmp/bundle

7. Reconstitute the image file from the bundle using the ec2-unbundle (p. 157) command.

a. Change directories to the bundle folder.

[ec2-user ~]$ cd /tmp/bundle/

b. Run the ec2-unbundle (p. 157) command.

[ec2-user bundle]$ ec2-unbundle -m image.manifest.xml --privatekey /path/to/pk-

HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem

8. Copy the files from the unbundled image to the new EBS volume.

[ec2-user bundle]$ sudo dd if=/tmp/bundle/image of=/dev/sdb bs=1M

9. Probe the volume for any new partitions that were unbundled.

[ec2-user bundle]$ sudo partprobe /dev/sdb1

10. List the block devices to find the device name to mount.

[ec2-user bundle]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
/dev/sda 202:0 0 8G 0 disk
##/dev/sda1 202:1 0 8G 0 part /
/dev/sdb 202:80 0 10G 0 disk
##/dev/sdb1 202:81 0 10G 0 part

In this example, the partition to mount is /dev/sdb1, but your device name will likely be different.
If your volume is not partitioned, then the device to mount will be similar to /dev/sdb (without a
device partition trailing digit).
11. Create a mount point for the new EBS volume and mount the volume.

[ec2-user bundle]$ sudo mkdir /mnt/ebs

[ec2-user bundle]$ sudo mount /dev/sdb1 /mnt/ebs

12. Open the /etc/fstab file on the EBS volume with your favorite text editor (such as vim or nano)
and remove any entries for instance store (ephemeral) volumes. Because the EBS volume is mounted
on /mnt/ebs, the fstab file is located at /mnt/ebs/etc/fstab.

[ec2-user bundle]$ sudo nano /mnt/ebs/etc/fstab

#

143
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Convert to an Amazon EBS-Backed AMI

LABEL=/ / ext4 defaults,noatime 1 1

tmpfs /dev/shm tmpfs defaults 0 0
devpts /dev/pts devpts gid=5,mode=620 0 0
sysfs /sys sysfs defaults 0 0
proc /proc proc defaults 0 0
/dev/sdb /media/ephemeral0 auto defaults,comment=cloudconfig 0
2

In this example, the last line should be removed.

13. Unmount the volume and detach it from the instance.

[ec2-user bundle]$ sudo umount /mnt/ebs

[ec2-user bundle]$ aws ec2 detach-volume --volume-id volume_id --region us-west-2

14. Create an AMI from the new EBS volume as follows.

a. Create a snapshot of the new EBS volume.

[ec2-user bundle]$ aws ec2 create-snapshot --region us-west-2 --description

"your_snapshot_description" --volume-id volume_id

b. Check to see that your snapshot is complete.

[ec2-user bundle]$ aws ec2 describe-snapshots --region us-west-2 --snapshot-

id snapshot_id

c. Identify the processor architecture, virtualization type, and the kernel image (aki) used on the
original AMI with the describe-images command. You need the AMI ID of the original instance
store-backed AMI for this step.

[ec2-user bundle]$ aws ec2 describe-images --region us-west-2 --image-id ami-id --

output text
IMAGES x86_64 amazon/amzn-ami-pv-2013.09.2.x86_64-s3 ami-8ef297be amazon available
public machine aki-fc8f11cc instance-store paravirtual xen

In this example, the architecture is x86_64 and the kernel image ID is aki-fc8f11cc. Use
these values in the following step. If the output of the above command also lists an ari ID, take
note of that as well.
d. Register your new AMI with the snapshot ID of your new EBS volume and the values from the
previous step. If the previous command output listed an ari ID, include that in the following
command with --ramdisk-id ari_id.

[ec2-user bundle]$ aws ec2 register-image --region us-west-2 --

name your_new_ami_name --block-device-mappings DeviceName=device-
name,Ebs={SnapshotId=snapshot_id} --virtualization-type paravirtual --architecture
x86_64 --kernel-id aki-fc8f11cc --root-device-name device-name

15. (Optional) After you have tested that you can launch an instance from your new AMI, you can delete
the EBS volume that you created for this procedure.

aws ec2 delete-volume --volume-id volume_id

144
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

AMI tools reference

You can use the AMI tools commands to create and manage instance store-backed Linux AMIs. To set up
the tools, see Set up the AMI tools (p. 132).

For information about your access keys, see Best Practices for Managing AWS Access Keys.

Commands
• ec2-ami-tools-version (p. 145)
• ec2-bundle-image (p. 145)
• ec2-bundle-vol (p. 148)
• ec2-delete-bundle (p. 151)
• ec2-download-bundle (p. 153)
• ec2-migrate-manifest (p. 155)
• ec2-unbundle (p. 157)
• ec2-upload-bundle (p. 158)
• Common options for AMI tools (p. 161)

ec2-ami-tools-version
Description
Describes the version of the AMI tools.

Syntax
ec2-ami-tools-version

Output
The version information.

Example
This example command displays the version information for the AMI tools that you're using.

[ec2-user ~]$ ec2-ami-tools-version

1.5.2 20071010

ec2-bundle-image
Description
Creates an instance store-backed Linux AMI from an operating system image created in a loopback file.

Syntax
ec2-bundle-image -c path -k path -u account -i path [-d path] [--ec2cert path]
[-r architecture] [--productcodes code1,code2,...] [-B mapping] [-p prefix]

Options
-c, --cert path

The user's PEM encoded RSA public key certificate file.

145
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

Required: Yes
-k, --privatekey path

The path to a PEM-encoded RSA key file. You'll need to specify this key to unbundle this bundle, so
keep it in a safe place. Note that the key doesn't have to be registered to your AWS account.

Required: Yes
-u, --user account

The user's AWS account ID, without dashes.

Required: Yes
-i, --image path

The path to the image to bundle.

Required: Yes
-d, --destination path

The directory in which to create the bundle.

Default: /tmp

Required: No
--ec2cert path

The path to the Amazon EC2 X.509 public key certificate used to encrypt the image manifest.

The us-gov-west-1 and cn-north-1 Regions use a non-default public key certificate and the
path to that certificate must be specified with this option. The path to the certificate varies based
on the installation method of the AMI tools. For Amazon Linux, the certificates are located at /opt/
aws/amitools/ec2/etc/ec2/amitools/. If you installed the AMI tools from the RPM or ZIP file
in Set up the AMI tools (p. 132), the certificates are located at $EC2_AMITOOL_HOME/etc/ec2/
amitools/.

Required: Only for the us-gov-west-1 and cn-north-1 Regions.

-r, --arch architecture

Image architecture. If you don't provide the architecture on the command line, you'll be prompted
for it when bundling starts.

Valid values: i386 | x86_64

Required: No
--productcodes code1,code2,...

Product codes to attach to the image at registration time, separated by commas.

Required: No
-B, --block-device-mapping mapping

Defines how block devices are exposed to an instance of this AMI if its instance type supports the
specified device.

Specify a comma-separated list of key-value pairs, where each key is a virtual name and each value is
the corresponding device name. Virtual names include the following:

146
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

• ami—The root file system device, as seen by the instance

• root—The root file system device, as seen by the kernel
• swap—The swap device, as seen by the instance
• ephemeralN—The Nth instance store volume

Required: No
-p, --prefix prefix

The filename prefix for bundled AMI files.

Default: The name of the image file. For example, if the image path is /var/spool/my-image/
version-2/debian.img, then the default prefix is debian.img.

Required: No
--kernel kernel_id

Deprecated. Use register-image to set the kernel.

Required: No
--ramdisk ramdisk_id

Deprecated. Use register-image to set the RAM disk if required.

Required: No

Output
Status messages describing the stages and status of the bundling process.

Example
This example creates a bundled AMI from an operating system image that was created in a loopback file.

[ec2-user ~]$ ec2-bundle-image -k pk-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem -c cert-

HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem -u 111122223333 -i image.img -d bundled/ -r x86_64
Please specify a value for arch [i386]:
Bundling image file...
Splitting bundled/image.gz.crypt...
Created image.part.00
Created image.part.01
Created image.part.02
Created image.part.03
Created image.part.04
Created image.part.05
Created image.part.06
Created image.part.07
Created image.part.08
Created image.part.09
Created image.part.10
Created image.part.11
Created image.part.12
Created image.part.13
Created image.part.14
Generating digests for each part...
Digests generated.
Creating bundle manifest...
ec2-bundle-image complete.

147
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

ec2-bundle-vol
Description
Creates an instance store-backed Linux AMI by compressing, encrypting, and signing a copy of the root
device volume for the instance.

Amazon EC2 attempts to inherit product codes, kernel settings, RAM disk settings, and block device
mappings from the instance.

By default, the bundle process excludes files that might contain sensitive information. These files
include *.sw, *.swo, *.swp, *.pem, *.priv, *id_rsa*, *id_dsa* *.gpg, *.jks, */.ssh/
authorized_keys, and */.bash_history. To include all of these files, use the --no-filter option.
To include some of these files, use the --include option.

For more information, see Create an instance store-backed Linux AMI (p. 130).

Syntax
ec2-bundle-vol -c path -k path -u account [-d path] [--ec2cert path] [-
r architecture] [--productcodes code1,code2,...] [-B mapping] [--all] [-e
directory1,directory2,...] [-i file1,file2,...] [--no-filter] [-p prefix] [-
s size] [--[no-]inherit] [-v volume] [-P type] [-S script] [--fstab path] [--
generate-fstab] [--grub-config path]

Options
-c, --cert path

The user's PEM encoded RSA public key certificate file.

Required: Yes
-k, --privatekey path

The path to the user's PEM-encoded RSA key file.

Required: Yes
-u, --user account

The user's AWS account ID, without dashes.

Required: Yes
-d, --destination destination

The directory in which to create the bundle.

Default: /tmp

Required: No
--ec2cert path

The path to the Amazon EC2 X.509 public key certificate used to encrypt the image manifest.

The us-gov-west-1 and cn-north-1 Regions use a non-default public key certificate and the
path to that certificate must be specified with this option. The path to the certificate varies based
on the installation method of the AMI tools. For Amazon Linux, the certificates are located at /opt/
aws/amitools/ec2/etc/ec2/amitools/. If you installed the AMI tools from the RPM or ZIP file

148
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

in Set up the AMI tools (p. 132), the certificates are located at $EC2_AMITOOL_HOME/etc/ec2/
amitools/.

Required: Only for the us-gov-west-1 and cn-north-1 Regions.

-r, --arch architecture

The image architecture. If you don't provide this on the command line, you'll be prompted to provide
it when the bundling starts.

Valid values: i386 | x86_64

Required: No
--productcodes code1,code2,...

Product codes to attach to the image at registration time, separated by commas.

Required: No
-B, --block-device-mapping mapping

Defines how block devices are exposed to an instance of this AMI if its instance type supports the
specified device.

Specify a comma-separated list of key-value pairs, where each key is a virtual name and each value is
the corresponding device name. Virtual names include the following:
• ami—The root file system device, as seen by the instance
• root—The root file system device, as seen by the kernel
• swap—The swap device, as seen by the instance
• ephemeralN—The Nth instance store volume

Required: No
-a, --all

Bundle all directories, including those on remotely mounted file systems.

Required: No
-e, --exclude directory1,directory2,...

A list of absolute directory paths and files to exclude from the bundle operation. This parameter
overrides the --all option. When exclude is specified, the directories and subdirectories listed with
the parameter will not be bundled with the volume.

Required: No
-i, --include file1,file2,...

A list of files to include in the bundle operation. The specified files would otherwise be excluded
from the AMI because they might contain sensitive information.

Required: No
--no-filter

If specified, we won't exclude files from the AMI because they might contain sensitive information.

Required: No
-p, --prefix prefix

The file name prefix for bundled AMI files.

149
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

Default: image

Required: No
-s, --size size

The size, in MB (1024 * 1024 bytes), of the image file to create. The maximum size is 10240 MB.

Default: 10240

Required: No
--[no-]inherit

Indicates whether the image should inherit the instance's metadata (the default is to inherit).
Bundling fails if you enable --inherit but the instance metadata is not accessible.

Required: No
-v, --volume volume

The absolute path to the mounted volume from which to create the bundle.

Default: The root directory (/)

Required: No
-P, --partition type

Indicates whether the disk image should use a partition table. If you don't specify a partition table
type, the default is the type used on the parent block device of the volume, if applicable, otherwise
the default is gpt.

Valid values: mbr | gpt | none

Required: No
-S, --script script

A customization script to be run right before bundling. The script must expect a single argument, the
mount point of the volume.

Required: No
--fstab path

The path to the fstab to bundle into the image. If this is not specified, Amazon EC2 bundles /etc/
fstab.

Required: No
--generate-fstab

Bundles the volume using an Amazon EC2-provided fstab.

Required: No
--grub-config

The path to an alternate grub configuration file to bundle into the image. By default, ec2-bundle-
vol expects either /boot/grub/menu.lst or /boot/grub/grub.conf to exist on the cloned
image. This option allows you to specify a path to an alternative grub configuration file, which will
then be copied over the defaults (if present).

Required: No

150
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

--kernel kernel_id

Deprecated. Use register-image to set the kernel.

Required: No
--ramdiskramdisk_id

Deprecated. Use register-image to set the RAM disk if required.

Required: No

Output
Status messages describing the stages and status of the bundling.

Example
This example creates a bundled AMI by compressing, encrypting and signing a snapshot of the local
machine's root file system.

[ec2-user ~]$ ec2-bundle-vol -d /mnt -k pk-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem -c cert-

HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem -u 111122223333 -r x86_64
Copying / into the image file /mnt/image...
Excluding:
sys
dev/shm
proc
dev/pts
proc/sys/fs/binfmt_misc
dev
media
mnt
proc
sys
tmp/image
mnt/img-mnt
1+0 records in
1+0 records out
mke2fs 1.38 (30-Jun-2005)
warning: 256 blocks unused.

Splitting /mnt/image.gz.crypt...
Created image.part.00
Created image.part.01
Created image.part.02
Created image.part.03
...
Created image.part.22
Created image.part.23
Generating digests for each part...
Digests generated.
Creating bundle manifest...
Bundle Volume complete.

ec2-delete-bundle
Description
Deletes the specified bundle from Amazon S3 storage. After you delete a bundle, you can't launch
instances from the corresponding AMI.

151
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

Syntax
ec2-delete-bundle -b bucket -a access_key_id -s secret_access_key [-t token]
[--url url] [--region region] [--sigv version] [-m path] [-p prefix] [--clear]
[--retry] [-y]

Options
-b, --bucket bucket

The name of the Amazon S3 bucket containing the bundled AMI, followed by an optional '/'-
delimited path prefix

Required: Yes
-a, --access-key access_key_id

The AWS access key ID.

Required: Yes
-s, --secret-key secret_access_key

The AWS secret access key.

Required: Yes
-t, --delegation-token token

The delegation token to pass along to the AWS request. For more information, see the Using
Temporary Security Credentials.

Required: Only when you are using temporary security credentials.

Default: The value of the AWS_DELEGATION_TOKEN environment variable (if set).

--regionregion

The Region to use in the request signature.

Default: us-east-1

Required: Required if using signature version 4

--sigvversion

The signature version to use when signing the request.

Valid values: 2 | 4

Default: 4

Required: No
-m, --manifestpath

The path to the manifest file.

Required: You must specify --prefix or --manifest.

-p, --prefix prefix

The bundled AMI filename prefix. Provide the entire prefix. For example, if the prefix is image.img,
use -p image.img and not -p image.

Required: You must specify --prefix or --manifest.

152
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

--clear

Deletes the Amazon S3 bucket if it's empty after deleting the specified bundle.

Required: No
--retry

Automatically retries on all Amazon S3 errors, up to five times per operation.

Required: No
-y, --yes

Automatically assumes the answer to all prompts is yes.

Required: No

Output
Amazon EC2 displays status messages indicating the stages and status of the delete process.

Example
This example deletes a bundle from Amazon S3.

[ec2-user ~]$ ec2-delete-bundle -b DOC-EXAMPLE-BUCKET1 -a your_access_key_id -

s your_secret_access_key
Deleting files:
DOC-EXAMPLE-BUCKET1/
image.manifest.xml
DOC-EXAMPLE-BUCKET1/
image.part.00
DOC-EXAMPLE-BUCKET1/
image.part.01
DOC-EXAMPLE-BUCKET1/
image.part.02
DOC-EXAMPLE-BUCKET1/
image.part.03
DOC-EXAMPLE-BUCKET1/
image.part.04
DOC-EXAMPLE-BUCKET1/
image.part.05
DOC-EXAMPLE-BUCKET1/image.part.06
Continue? [y/n]
y
Deleted DOC-EXAMPLE-BUCKET1/image.manifest.xml
Deleted DOC-EXAMPLE-BUCKET1/image.part.00
Deleted DOC-EXAMPLE-BUCKET1/image.part.01
Deleted DOC-EXAMPLE-BUCKET1/image.part.02
Deleted DOC-EXAMPLE-BUCKET1/image.part.03
Deleted DOC-EXAMPLE-BUCKET1/image.part.04
Deleted DOC-EXAMPLE-BUCKET1/image.part.05
Deleted DOC-EXAMPLE-BUCKET1/image.part.06
ec2-delete-bundle complete.

ec2-download-bundle
Description
Downloads the specified instance store-backed Linux AMIs from Amazon S3 storage.

153
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

Syntax
ec2-download-bundle -b bucket -a access_key_id -s secret_access_key -k path
[--url url] [--region region] [--sigv version] [-m file] [-p prefix] [-d
directory] [--retry]

Options
-b, --bucket bucket

The name of the Amazon S3 bucket where the bundle is located, followed by an optional '/'-
delimited path prefix.

Required: Yes
-a, --access-key access_key_id

The AWS access key ID.

Required: Yes
-s, --secret-key secret_access_key

The AWS secret access key.

Required: Yes
-k, --privatekey path

The private key used to decrypt the manifest.

Required: Yes
--url url

The Amazon S3 service URL.

Default: https://s3.amazonaws.com/

Required: No
--region region

The Region to use in the request signature.

Default: us-east-1

Required: Required if using signature version 4

--sigv version

The signature version to use when signing the request.

Valid values: 2 | 4

Default: 4

Required: No
-m, --manifest file

The name of the manifest file (without the path). We recommend that you specify either the
manifest (-m) or a prefix (-p).

Required: No
-p, --prefix prefix

The filename prefix for the bundled AMI files.

154
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

Default: image

Required: No
-d, --directory directory

The directory where the downloaded bundle is saved. The directory must exist.

Default: The current working directory.

Required: No
--retry

Automatically retries on all Amazon S3 errors, up to five times per operation.

Required: No

Output
Status messages indicating the various stages of the download process are displayed.

Example
This example creates the bundled directory (using the Linux mkdir command) and downloads the
bundle from the DOC-EXAMPLE-BUCKET1 Amazon S3 bucket.

[ec2-user ~]$ mkdir bundled

[ec2-user ~]$ ec2-download-bundle -b DOC-EXAMPLE-BUCKET1/bundles/bundle_name
-m image.manifest.xml -a your_access_key_id -s your_secret_access_key -k pk-
HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem -d mybundle
Downloading manifest image.manifest.xml from DOC-EXAMPLE-BUCKET1 to mybundle/
image.manifest.xml ...
Downloading part image.part.00 from DOC-EXAMPLE-BUCKET1/bundles/bundle_name to mybundle/
image.part.00 ...
Downloaded image.part.00 from DOC-EXAMPLE-BUCKET1
Downloading part image.part.01 from DOC-EXAMPLE-BUCKET1/bundles/bundle_name to mybundle/
image.part.01 ...
Downloaded image.part.01 from DOC-EXAMPLE-BUCKET1
Downloading part image.part.02 from DOC-EXAMPLE-BUCKET1/bundles/bundle_name to mybundle/
image.part.02 ...
Downloaded image.part.02 from DOC-EXAMPLE-BUCKET1
Downloading part image.part.03 from DOC-EXAMPLE-BUCKET1/bundles/bundle_name to mybundle/
image.part.03 ...
Downloaded image.part.03 from DOC-EXAMPLE-BUCKET1
Downloading part image.part.04 from DOC-EXAMPLE-BUCKET1/bundles/bundle_name to mybundle/
image.part.04 ...
Downloaded image.part.04 from DOC-EXAMPLE-BUCKET1
Downloading part image.part.05 from DOC-EXAMPLE-BUCKET1/bundles/bundle_name to mybundle/
image.part.05 ...
Downloaded image.part.05 from DOC-EXAMPLE-BUCKET1
Downloading part image.part.06 from DOC-EXAMPLE-BUCKET1/bundles/bundle_name to mybundle/
image.part.06 ...
Downloaded image.part.06 from DOC-EXAMPLE-BUCKET1

ec2-migrate-manifest
Description
Modifies an instance store-backed Linux AMI (for example, its certificate, kernel, and RAM disk) so that it
supports a different Region.

155
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

Syntax
ec2-migrate-manifest -c path -k path -m path {(-a access_key_id -s
secret_access_key --region region) | (--no-mapping)} [--ec2cert ec2_cert_path]
[--kernel kernel-id] [--ramdisk ramdisk_id]

Options
-c, --cert path

The user's PEM encoded RSA public key certificate file.

Required: Yes
-k, --privatekey path

The path to the user's PEM-encoded RSA key file.

Required: Yes
--manifest path

The path to the manifest file.

Required: Yes
-a, --access-key access_key_id

The AWS access key ID.

Required: Required if using automatic mapping.

-s, --secret-key secret_access_key

The AWS secret access key.

Required: Required if using automatic mapping.

--region region

The Region to look up in the mapping file.

Required: Required if using automatic mapping.

--no-mapping

Disables automatic mapping of kernels and RAM disks.

During migration, Amazon EC2 replaces the kernel and RAM disk in the manifest file with a kernel
and RAM disk designed for the destination region. Unless the --no-mapping parameter is given,
ec2-migrate-bundle might use the DescribeRegions and DescribeImages operations to
perform automated mappings.

Required: Required if you're not providing the -a, -s, and --region options used for automatic
mapping.
--ec2cert path

The path to the Amazon EC2 X.509 public key certificate used to encrypt the image manifest.

The us-gov-west-1 and cn-north-1 Regions use a non-default public key certificate and the
path to that certificate must be specified with this option. The path to the certificate varies based
on the installation method of the AMI tools. For Amazon Linux, the certificates are located at /
opt/aws/amitools/ec2/etc/ec2/amitools/. If you installed the AMI tools from the ZIP file
in Set up the AMI tools (p. 132), the certificates are located at $EC2_AMITOOL_HOME/etc/ec2/
amitools/.

156
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

Required: Only for the us-gov-west-1 and cn-north-1 Regions.

--kernel kernel_id

The ID of the kernel to select.

Important
We recommend that you use PV-GRUB instead of kernels and RAM disks. For more
information, see Enabling Your Own Linux Kernels (p. 196).

Required: No
--ramdisk ramdisk_id

The ID of the RAM disk to select.

Important
We recommend that you use PV-GRUB instead of kernels and RAM disks. For more
information, see Enabling Your Own Linux Kernels (p. 196).

Required: No

Output
Status messages describing the stages and status of the bundling process.

Example
This example copies the AMI specified in the my-ami.manifest.xml manifest from the US to the EU.

[ec2-user ~]$ ec2-migrate-manifest --manifest my-ami.manifest.xml --cert cert-

HKZYKTAIG2ECMXYIBH3HXV4ZBZQ55CLO.pem --privatekey pk-HKZYKTAIG2ECMXYIBH3HXV4ZBZQ55CLO.pem
--region eu-west-1

Backing up manifest...
Successfully migrated my-ami.manifest.xml It is now suitable for use in eu-west-1.

ec2-unbundle
Description
Re-creates the bundle from an instance store-backed Linux AMI.

Syntax
ec2-unbundle -k path -m path [-s source_directory] [-d destination_directory]

Options
-k, --privatekey path

The path to your PEM-encoded RSA key file.

Required: Yes
-m, --manifest path

The path to the manifest file.

Required: Yes

157
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

-s, --source source_directory

The directory containing the bundle.

Default: The current directory.

Required: No
-d, --destination destination_directory

The directory in which to unbundle the AMI. The destination directory must exist.

Default: The current directory.

Required: No

Example
This Linux and UNIX example unbundles the AMI specified in the image.manifest.xml file.

[ec2-user ~]$ mkdir unbundled

$ ec2-unbundle -m mybundle/image.manifest.xml -k pk-HKZYKTAIG2ECMXYIBH3HXV4ZBEXAMPLE.pem -s
mybundle -d unbundled
$ ls -l unbundled
total 1025008
-rw-r--r-- 1 root root 1048578048 Aug 25 23:46 image.img

Output
Status messages indicating the various stages of the unbundling process are displayed.

ec2-upload-bundle
Description
Uploads the bundle for an instance store-backed Linux AMI to Amazon S3 and sets the appropriate ACLs
on the uploaded objects. For more information, see Create an instance store-backed Linux AMI (p. 130).

Syntax
ec2-upload-bundle -b bucket -a access_key_id -s secret_access_key [-t token] -m
path [--url url] [--region region] [--sigv version] [--acl acl] [-d directory]
[--part part] [--retry] [--skipmanifest]

Options
-b, --bucket bucket

The name of the Amazon S3 bucket in which to store the bundle, followed by an optional '/'-
delimited path prefix. If the bucket doesn't exist, it's created if the bucket name is available.

Required: Yes
-a, --access-key access_key_id

Your AWS access key ID.

Required: Yes
-s, --secret-key secret_access_key

Your AWS secret access key.

158
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

Required: Yes
-t, --delegation-token token

The delegation token to pass along to the AWS request. For more information, see the Using
Temporary Security Credentials.

Required: Only when you are using temporary security credentials.

Default: The value of the AWS_DELEGATION_TOKEN environment variable (if set).

-m, --manifest path

The path to the manifest file. The manifest file is created during the bundling process and can be
found in the directory containing the bundle.

Required: Yes
--url url

Deprecated. Use the --region option instead unless your bucket is constrained to the EU location
(and not eu-west-1). The --location flag is the only way to target that specific location restraint.

The Amazon S3 endpoint service URL.

Default: https://s3.amazonaws.com/

Required: No
--region region

The Region to use in the request signature for the destination S3 bucket.
• If the bucket doesn't exist and you don't specify a Region, the tool creates the bucket without a
location constraint (in us-east-1).
• If the bucket doesn't exist and you specify a Region, the tool creates the bucket in the specified
Region.
• If the bucket exists and you don't specify a Region, the tool uses the bucket's location.
• If the bucket exists and you specify us-east-1 as the Region, the tool uses the bucket's actual
location without any error message, any existing matching files are over-written.
• If the bucket exists and you specify a Region (other than us-east-1) that doesn't match the
bucket's actual location, the tool exits with an error.

If your bucket is constrained to the EU location (and not eu-west-1), use the --location flag
instead. The --location flag is the only way to target that specific location restraint.

Default: us-east-1

Required: Required if using signature version 4

--sigv version

The signature version to use when signing the request.

Valid values: 2 | 4

Default: 4

Required: No
--acl acl

The access control list policy of the bundled image.

Valid values: public-read | aws-exec-read

Default: aws-exec-read

159
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI tools reference

Required: No
-d, --directory directory

The directory containing the bundled AMI parts.

Default: The directory containing the manifest file (see the -m option).

Required: No
--part part

Starts uploading the specified part and all subsequent parts. For example, --part 04.

Required: No
--retry

Automatically retries on all Amazon S3 errors, up to five times per operation.

Required: No
--skipmanifest

Does not upload the manifest.

Required: No
--location location

Deprecated. Use the --region option instead, unless your bucket is constrained to the EU location
(and not eu-west-1). The --location flag is the only way to target that specific location restraint.

The location constraint of the destination Amazon S3 bucket. If the bucket exists and you specify a
location that doesn't match the bucket's actual location, the tool exits with an error. If the bucket
exists and you don't specify a location, the tool uses the bucket's location. If the bucket doesn't exist
and you specify a location, the tool creates the bucket in the specified location. If the bucket doesn't
exist and you don't specify a location, the tool creates the bucket without a location constraint (in
us-east-1).

Default: If --region is specified, the location is set to that specified Region. If --region is not
specified, the location defaults to us-east-1.

Required: No

Output
Amazon EC2 displays status messages that indicate the stages and status of the upload process.

Example
This example uploads the bundle specified by the image.manifest.xml manifest.

[ec2-user ~]$ ec2-upload-bundle -b DOC-EXAMPLE-BUCKET1/bundles/bundle_name -m

image.manifest.xml -a your_access_key_id -s your_secret_access_key
Creating bucket...
Uploading bundled image parts to the S3 bucket DOC-EXAMPLE-BUCKET1 ...
Uploaded image.part.00
Uploaded image.part.01
Uploaded image.part.02
Uploaded image.part.03
Uploaded image.part.04
Uploaded image.part.05
Uploaded image.part.06
Uploaded image.part.07

160
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Use encryption with EBS-backed AMIs

Uploaded image.part.08
Uploaded image.part.09
Uploaded image.part.10
Uploaded image.part.11
Uploaded image.part.12
Uploaded image.part.13
Uploaded image.part.14
Uploading manifest ...
Uploaded manifest.
Bundle upload completed.

Common options for AMI tools

Most of the AMI tools accept the following optional parameters.

--help, -h

Displays the help message.

--version

Displays the version and copyright notice.

--manual

Displays the manual entry.

--batch

Runs in batch mode, suppressing interactive prompts.

--debug

Displays information that can be useful when troubleshooting problems.

Use encryption with EBS-backed AMIs

AMIs that are backed by Amazon EBS snapshots can take advantage of Amazon EBS encryption.
Snapshots of both data and root volumes can be encrypted and attached to an AMI. You can launch
instances and copy images with full EBS encryption support included. Encryption parameters for these
operations are supported in all Regions where AWS KMS is available.

EC2 instances with encrypted EBS volumes are launched from AMIs in the same way as other instances.
In addition, when you launch an instance from an AMI backed by unencrypted EBS snapshots, you can
encrypt some or all of the volumes during launch.

Like EBS volumes, snapshots in AMIs can be encrypted by either your default AWS Key Management
Service customer master key (CMK), or to a customer managed key that you specify. You must in all cases
have permission to use the selected key.

AMIs with encrypted snapshots can be shared across AWS accounts. For more information, see Shared
AMIs.

Encryption with EBS-backed AMIs topics

• Instance-launching scenarios (p. 161)
• Image-copying scenarios (p. 164)

Instance-launching scenarios

161
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance-launching scenarios

Amazon EC2 instances are launched from AMIs using the RunInstances action with parameters
supplied through block device mapping, either by means of the AWS Management Console or directly
using the Amazon EC2 API or CLI. For more information about block device mapping, see Block device
mapping. For examples of controlling block device mapping from the AWS CLI, see Launch, List, and
Terminate EC2 Instances.

By default, without explicit encryption parameters, a RunInstances action maintains the existing
encryption state of an AMI's source snapshots while restoring EBS volumes from them. If Encryption by
default (p. 1178) is enabled, all volumes created from the AMI (whether from encrypted or unencrypted
snapshots) will be encrypted. If encryption by default is not enabled, then the instance maintains the
encryption state of the AMI.

You can also launch an instance and simultaneously apply a new encryption state to the resulting
volumes by supplying encryption parameters. Consequently, the following behaviors are observed:

Launch with no encryption parameters

• An unencrypted snapshot is restored to an unencrypted volume, unless encryption by default is

enabled, in which case all the newly created volumes will be encrypted.
• An encrypted snapshot that you own is restored to a volume that is encrypted to the same CMK.
• An encrypted snapshot that you do not own (for example, the AMI is shared with you) is restored to a
volume that is encrypted by your AWS account's default CMK.

The default behaviors can be overridden by supplying encryption parameters. The available parameters
are Encrypted and KmsKeyId. Setting only the Encrypted parameter results in the following:

Instance launch behaviors with Encrypted set, but no KmsKeyId specified

• An unencrypted snapshot is restored to an EBS volume that is encrypted by your AWS account's
default CMK.
• An encrypted snapshot that you own is restored to an EBS volume encrypted by the same CMK. (In
other words, the Encrypted parameter has no effect.)
• An encrypted snapshot that you do not own (i.e., the AMI is shared with you) is restored to a volume
that is encrypted by your AWS account's default CMK. (In other words, the Encrypted parameter has
no effect.)

Setting both the Encrypted and KmsKeyId parameters allows you to specify a non-default CMK for an
encryption operation. The following behaviors result:

Instance with both Encrypted and KmsKeyId set

• An unencrypted snapshot is restored to an EBS volume encrypted by the specified CMK.

• An encrypted snapshot is restored to an EBS volume encrypted not to the original CMK, but instead to
the specified CMK.

Submitting a KmsKeyId without also setting the Encrypted parameter results in an error.

The following sections provide examples of launching instances from AMIs using non-default encryption
parameters. In each of these scenarios, parameters supplied to the RunInstances action result in a
change of encryption state during restoration of a volume from a snapshot.
Note
For detailed console procedures to launch an instance from an AMI, see Launch Your Instance.
For documentation of the RunInstances API, see RunInstances.
For documentation of the run-instances command in the AWS Command Line Interface, see
run-instances.

162
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance-launching scenarios

Encrypt a volume during launch

In this example, an AMI backed by an unencrypted snapshot is used to launch an EC2 instance with an
encrypted EBS volume.

The Encrypted parameter alone results in the volume for this instance being encrypted. Providing
a KmsKeyId parameter is optional. If no key ID is specified, the AWS account's default CMK is used to
encrypt the volume. To encrypt the volume to a different CMK that you own, supply the KmsKeyId
parameter.

Re-encrypt a volume during launch

In this example, an AMI backed by an encrypted snapshot is used to launch an EC2 instance with an EBS
volume encrypted by a new CMK.

163
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Image-copying scenarios

If you own the AMI and supply no encryption parameters, the resulting instance has a volume encrypted
by the same key as the snapshot. If the AMI is shared rather than owned by you, and you supply no
encryption parameters, the volume is encrypted by your default CMK. With encryption parameters
supplied as shown, the volume is encrypted by the specified CMK.

Change encryption state of multiple volumes during launch

In this more complex example, an AMI backed by multiple snapshots (each with its own encryption state)
is used to launch an EC2 instance with a newly encrypted volume and a re-encrypted volume.

In this scenario, the RunInstances action is supplied with encryption parameters for each of the source
snapshots. When all possible encryption parameters are specified, the resulting instance is the same
regardless of whether you own the AMI.

Image-copying scenarios
Amazon EC2 AMIs are copied using the CopyImage action, either through the AWS Management Console
or directly using the Amazon EC2 API or CLI.

By default, without explicit encryption parameters, a CopyImage action maintains the existing
encryption state of an AMI's source snapshots during copy. You can also copy an AMI and simultaneously
apply a new encryption state to its associated EBS snapshots by supplying encryption parameters.
Consequently, the following behaviors are observed:

Copy with no encryption parameters

164
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Image-copying scenarios

• An unencrypted snapshot is copied to another unencrypted snapshot, unless encryption by default is

enabled, in which case all the newly created snapshots will be encrypted.
• An encrypted snapshot that you own is copied to a snapshot encrypted with the same key.
• An encrypted snapshot that you do not own (that is, the AMI is shared with you) is copied to a
snapshot that is encrypted by your AWS account's default CMK.

All of these default behaviors can be overridden by supplying encryption parameters. The available
parameters are Encrypted and KmsKeyId. Setting only the Encrypted parameter results in the
following:

Copy-image behaviors with Encrypted set, but no KmsKeyId specified

• An unencrypted snapshot is copied to a snapshot encrypted by the AWS account's default CMK.
• An encrypted snapshot is copied to a snapshot encrypted by the same CMK. (In other words, the
Encrypted parameter has no effect.)
• An encrypted snapshot that you do not own (i.e., the AMI is shared with you) is copied to a volume that
is encrypted by your AWS account's default CMK. (In other words, the Encrypted parameter has no
effect.)

Setting both the Encrypted and KmsKeyId parameters allows you to specify a customer managed CMK
for an encryption operation. The following behaviors result:

Copy-image behaviors with both Encrypted and KmsKeyId set

• An unencrypted snapshot is copied to a snapshot encrypted by the specified CMK.

• An encrypted snapshot is copied to a snapshot encrypted not to the original CMK, but instead to the
specified CMK.

Submitting a KmsKeyId without also setting the Encrypted parameter results in an error.

The following section provides an example of copying an AMI using non-default encryption parameters,
resulting in a change of encryption state.
Note
For detailed console procedures to copy an AMI, see Copying an AMI.
For documentation of the CopyImage API, see CopyImage.
For documentation of the command copy-image in the AWS Command Line Interface, see
copy-image.

Encrypt an unencrypted image during copy

In this scenario, an AMI backed by an unencrypted root snapshot is copied to an AMI with an encrypted
root snapshot. The CopyImage action is invoked with two encryption parameters, including a CMK. As
a result, the encryption status of the root snapshot changes, so that the target AMI is backed by a root
snapshot containing the same data as the source snapshot, but encrypted using the specified key. You
incur storage costs for the snapshots in both AMIs, as well as charges for any instances you launch from
either AMI.
Note
Enabling encryption by default (p. 1178) has the same effect as setting the Encrypted
parameter to true for all snapshots in the AMI.

165
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Copy an AMI

Setting the Encrypted parameter encrypts the single snapshot for this instance. If you do not specify
the KmsKeyId parameter, the default CMK is used to encrypt the snapshot copy.
Note
You can also copy an image with multiple snapshots and configure the encryption state of each
individually.

Copy an AMI
You can copy an Amazon Machine Image (AMI) within or across AWS Regions using the AWS
Management Console, the AWS Command Line Interface or SDKs, or the Amazon EC2 API, all of which
support the CopyImage action. You can copy both Amazon EBS-backed AMIs and instance-store-backed
AMIs. You can copy AMIs with encrypted snapshots and also change encryption status during the copy
process.

Copying a source AMI results in an identical but distinct target AMI with its own unique identifier. In the
case of an Amazon EBS-backed AMI, each of its backing snapshots is, by default, copied to an identical
but distinct target snapshot. (The sole exceptions are when you choose to encrypt or re-encrypt the
snapshot.) You can change or deregister the source AMI with no effect on the target AMI. The reverse is
also true.

There are no charges for copying an AMI. However, standard storage and data transfer rates apply. If you
copy an EBS-backed AMI, you will incur charges for the storage of any additional EBS snapshots.

AWS does not copy launch permissions, user-defined tags, or Amazon S3 bucket permissions from the
source AMI to the new AMI. After the copy operation is complete, you can apply launch permissions,
user-defined tags, and Amazon S3 bucket permissions to the new AMI.

If you are using an AWS Marketplace AMI, or an AMI that was directly or indirectly derived from an AWS
Marketplace AMI, you cannot copy it across accounts. Instead, launch an EC2 instance using the AWS
Marketplace AMI and then create an AMI from the instance. For more information, see Create an Amazon
EBS-backed Linux AMI (p. 125).

166
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Permissions for copying an instance store-backed AMI

Topics
• Permissions for copying an instance store-backed AMI (p. 167)
• Cross-Region copying (p. 168)
• Cross-account copying (p. 169)
• Encryption and copying (p. 169)
• Copy an AMI (p. 170)
• Stop a pending AMI copy operation (p. 171)

Permissions for copying an instance store-backed AMI

If you use an IAM user to copy an instance store-backed AMI, the user must have the following Amazon
S3 permissions: s3:CreateBucket, s3:GetBucketAcl, s3:ListAllMyBuckets, s3:GetObject,
s3:PutObject, and s3:PutObjectAcl.

The following example policy allows the user to copy the AMI source in the specified bucket to the
specified Region.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "s3:ListAllMyBuckets",
"Resource": [
"arn:aws:s3:::*"
]
},
{
"Effect": "Allow",
"Action": "s3:GetObject",
"Resource": [
"arn:aws:s3:::ami-source-bucket/*"
]
},
{
"Effect": "Allow",
"Action": [
"s3:CreateBucket",
"s3:GetBucketAcl",
"s3:PutObjectAcl",
"s3:PutObject"
],
"Resource": [
"arn:aws:s3:::amis-for-123456789012-in-us-east-1*"
]
}
]
}

To find the Amazon Resource Name (ARN) of the AMI source bucket, open the Amazon EC2 console at
https://console.aws.amazon.com/ec2/, in the navigation pane choose AMIs, and locate the bucket name
in the Source column.
Note
The s3:CreateBucket permission is only needed the first time that the IAM user copies an
instance store-backed AMI to an individual Region. After that, the Amazon S3 bucket that is
already created in the Region is used to store all future AMIs that you copy to that Region.

167
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Cross-Region copying

Cross-Region copying
Copying an AMI across geographically diverse Regions provides the following benefits:

• Consistent global deployment: Copying an AMI from one Region to another enables you to launch
consistent instances in different Regions based on the same AMI.
• Scalability: You can more easily design and build global applications that meet the needs of your users,
regardless of their location.
• Performance: You can increase performance by distributing your application, as well as locating critical
components of your application in closer proximity to your users. You can also take advantage of
Region-specific features, such as instance types or other AWS services.
• High availability: You can design and deploy applications across AWS Regions, to increase availability.

The following diagram shows the relations among a source AMI and two copied AMIs in different
Regions, as well as the EC2 instances launched from each. When you launch an instance from an AMI, it
resides in the same Region where the AMI resides. If you make changes to the source AMI and want those
changes to be reflected in the AMIs in the target Regions, you must recopy the source AMI to the target
Regions.

When you first copy an instance store-backed AMI to a Region, we create an Amazon S3 bucket for the
AMIs copied to that Region. All instance store-backed AMIs that you copy to that Region are stored in this
bucket. The bucket names have the following format: amis-for-account-in-region-hash. For example:
amis-for-123456789012-in-us-east-2-yhjmxvp6.

Prerequisite

Prior to copying an AMI, you must ensure that the contents of the source AMI are updated to support
running in a different Region. For example, you should update any database connection strings or similar
application configuration data to point to the appropriate resources. Otherwise, instances launched from
the new AMI in the destination Region may still use the resources from the source Region, which can
impact performance and cost.

Limits

• Destination Regions are limited to 100 concurrent AMI copies.

• You cannot copy a paravirtual (PV) AMI to a Region that does not support PV AMIs. For more
information, see Linux AMI virtualization types (p. 104).

168
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Cross-account copying

Cross-account copying
You can share an AMI with another AWS account. Sharing an AMI does not affect the ownership of the
AMI. The owning account is charged for the storage in the Region. For more information, see Share an
AMI with specific AWS accounts (p. 115).

If you copy an AMI that has been shared with your account, you are the owner of the target AMI in your
account. The owner of the source AMI is charged standard Amazon EBS or Amazon S3 transfer fees, and
you are charged for the storage of the target AMI in the destination Region.

Resource Permissions

To copy an AMI that was shared with you from another account, the owner of the source AMI must grant
you read permissions for the storage that backs the AMI, either the associated EBS snapshot (for an
Amazon EBS-backed AMI) or an associated S3 bucket (for an instance store-backed AMI). If the shared
AMI has encrypted snapshots, the owner must share the key or keys with you as well.

Encryption and copying

The following table shows encryption support for various AMI-copying scenarios. While it is possible to
copy an unencrypted snapshot to yield an encrypted snapshot, you cannot copy an encrypted snapshot
to yield an unencrypted one.

Scenario Description Supported

1 Unencrypted-to-unencrypted Yes

2 Encrypted-to-encrypted Yes

3 Unencrypted-to-encrypted Yes

4 Encrypted-to-unencrypted No

Note
Encrypting during the CopyImage action applies only to Amazon EBS-backed AMIs. Because
an instance store-backed AMI does not rely on snapshots, you cannot use copying to change its
encryption status.

By default (i.e., without specifying encryption parameters), the backing snapshot of an AMI is copied
with its original encryption status. Copying an AMI backed by an unencrypted snapshot results in an
identical target snapshot that is also unencrypted. If the source AMI is backed by an encrypted snapshot,
copying it results in an identical target snapshot that is encrypted by the same customer master key
(CMK). Copying an AMI backed by multiple snapshots preserves, by default, the source encryption status
in each target snapshot.

If you specify encryption parameters while copying an AMI, you can encrypt or re-encrypt its backing
snapshots. The following example shows a non-default case that supplies encryption parameters to the
CopyImage action in order to change the target AMI's encryption state.

Copy an unencrypted source AMI to an encrypted target AMI

In this scenario, an AMI backed by an unencrypted root snapshot is copied to an AMI with an encrypted
root snapshot. The CopyImage action is invoked with two encryption parameters, including a CMK. As
a result, the encryption status of the root snapshot changes, so that the target AMI is backed by a root
snapshot containing the same data as the source snapshot, but encrypted using the specified key. You

169
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Copy an AMI

incur storage costs for the snapshots in both AMIs, as well as charges for any instances you launch from
either AMI.
Note
Enabling encryption by default (p. 1178) has the same effect as setting the Encrypted
parameter to true for all snapshots in the AMI.

Setting the Encrypted parameter encrypts the single snapshot for this instance. If you do not specify
the KmsKeyId parameter, the default CMK is used to encrypt the snapshot copy.

For more information about copying AMIs with encrypted snapshots, see Use encryption with EBS-
backed AMIs (p. 161).

Copy an AMI
You can copy an AMI as follows.

Prerequisite

Create or obtain an AMI backed by an Amazon EBS snapshot. Note that you can use the Amazon EC2
console to search a wide variety of AMIs provided by AWS. For more information, see Create an Amazon
EBS-backed Linux AMI (p. 125) and Finding an AMI.

To copy an AMI using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the console navigation bar, select the Region that contains the AMI. In the navigation pane,
choose Images, AMIs to display the list of AMIs available to you in the Region.
3. Select the AMI to copy and choose Actions, Copy AMI.
4. In the Copy AMI dialog box, specify the following information and then choose Copy AMI:

• Destination region: The Region in which to copy the AMI.

• Name: A name for the new AMI. You can include operating system information in the name, as we
do not provide this information when displaying details about the AMI.

170
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Stop a pending AMI copy operation

• Description: By default, the description includes information about the source AMI so that you can
distinguish a copy from its original. You can change this description as needed.
• Encryption: Select this field to encrypt the target snapshots, or to re-encrypt them using a
different key. If you have enabled encryption by default, the Encryption option is set and cannot
be unset from the AMI console.
• Master Key: The KMS key to used to encrypt the target snapshots.
5. We display a confirmation page to let you know that the copy operation has been initiated and to
provide you with the ID of the new AMI.

To check on the progress of the copy operation immediately, follow the provided link. To check on
the progress later, choose Done, and then when you are ready, use the navigation bar to switch to
the target Region (if applicable) and locate your AMI in the list of AMIs.

The initial status of the target AMI is pending and the operation is complete when the status is
available.

To copy an AMI using the AWS CLI

You can copy an AMI using the copy-image command. You must specify both the source and destination
Regions. You specify the source Region using the --source-region parameter. You can specify
the destination Region using either the --region parameter or an environment variable. For more
information, see Configuring the AWS Command Line Interface.

When you encrypt a target snapshot during copying, you must specify these additional parameters: --
encrypted and --kms-key-id.

To copy an AMI using the Tools for Windows PowerShell

You can copy an AMI using the Copy-EC2Image command. You must specify both the source and
destination Regions. You specify the source Region using the -SourceRegion parameter. You can
specify the destination Region using either the -Region parameter or the Set-AWSDefaultRegion
command. For more information, see Specifying AWS Regions.

When you encrypt a target snapshot during copying, you must specify these additional parameters: -
Encrypted and -KmsKeyId.

Stop a pending AMI copy operation

You can stop a pending AMI copy as follows.

To stop an AMI copy operation using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the destination Region from the Region selector.
3. In the navigation pane, choose AMIs.
4. Select the AMI to stop copying and choose Actions, Deregister.
5. When asked for confirmation, choose Continue.

To stop an AMI copy operation using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• deregister-image (AWS CLI)

• Unregister-EC2Image (AWS Tools for Windows PowerShell)

171
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Understanding AMI billing

Understanding AMI billing information

There are many Amazon Machine Images (AMIs) to choose from when launching your instances, and they
support a variety of operating system platforms and features. To understand how the AMI you choose
when launching your instance affects the bottom line on your AWS bill, you can research the associated
operating system platform and billing information. Do this before you launch any On-Demand or Spot
Instances, or purchase a Reserved Instance.

Here are two examples of how researching your AMI in advance can help you choose the AMI that best
suits your needs:

• For Spot Instances, you can use the AMI Platform details to confirm that the AMI is supported for Spot
Instances.
• When purchasing a Reserved Instance, you can make sure that you select the operating system
platform (Platform) that maps to the AMI Platform details.

For more information about instance pricing, see Amazon EC2 pricing.

Contents
• AMI billing information fields (p. 172)
• Finding AMI billing and usage details (p. 173)
• Verify AMI charges on your bill (p. 175)

AMI billing information fields

The following fields provide billing information associated with an AMI:

Platform details

The platform details associated with the billing code of the AMI. For example, Red Hat
Enterprise Linux.
Usage operation

The operation of the Amazon EC2 instance and the billing code that is associated with the AMI. For
example, RunInstances:0010. Usage operation corresponds to the lineitem/Operation column on
your AWS Cost and Usage Report (CUR) and in the AWS Price List API.

You can view these fields on the Instances or AMIs page in the Amazon EC2 console, or in the response
that is returned by the describe-images command.

Sample data: usage operation by platform

The following table lists some of the platform details and usage operation values that can be displayed
on the Instances or AMIs pages in the Amazon EC2 console, or in the response that is returned by the
describe-images command.

Platform details Usage operation **

Linux/UNIX RunInstances

Red Hat BYOL Linux RunInstances:00g0

Red Hat Enterprise Linux RunInstances:0010

172
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Find AMI billing information

Platform details Usage operation **

SQL Server Enterprise RunInstances:0100

SQL Server Standard RunInstances:0004

SQL Server Web RunInstances:0200

SUSE Linux RunInstances:000g

Windows RunInstances:0002

Windows BYOL RunInstances:0800

Windows with SQL Server Enterprise * RunInstances:0102

Windows with SQL Server Standard * RunInstances:0006

Windows with SQL Server Web * RunInstances:0202

* If two software licenses are associated with an AMI, the Platform details field shows both.

** If you are running Spot Instances, the lineitem/Operation on your AWS Cost and Usage Report
might be different from the Usage operation value that is listed here. For example, if lineitem/
Operation displays RunInstances:0010:SV006, it means that Amazon EC2 is running Red Hat
Enterprise Linux Spot Instance-hour in US East (Virginia) in VPC Zone #6.

Finding AMI billing and usage details

In the Amazon EC2 console, you can view the AMI billing information from the AMIs page or from the
Instances page. You can also find billing information using the AWS CLI.

The following fields can help you verify AMI charges on your bill:

• Platform details
• Usage operation
• AMI ID

Find AMI billing information (console)

Follow these steps to view AMI billing information in the Amazon EC2 console:

Look up AMI billing information from the AMIs page

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose AMIs, and then select an AMI.
3. On the Details tab, check the values for Platform details and Usage operation.

Look up AMI billing information from the Instances page

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and then select an instance.
3. On the Details tab (or the Description tab if you are using the prior version of the console), check
the values for Platform details and Usage operation.

173
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Find AMI billing information

To look up the billingProducts code for a running instance, you can use the Instance Metadata
Service to retrieve the instance identity document in plaintext JSON format. For more information, see
Instance identity documents (p. 730).

Find AMI billing information (AWS CLI)

If you know the instance ID, and want to find out what the AMI ID is for that instance, you can use the
describe-instances command to retrieve it.

aws ec2 describe-instances --instance-ids i-123456789abcde123

..."Instances": [
{
"AmiLaunchIndex": 0,
"ImageId": "ami-0123456789EXAMPLE",
"InstanceId": "i-123456789abcde123",
...
}]

If you know the AMI ID, you can use the describe-images command to show AMI platform and usage
operation details:

$ aws ec2 describe-images --image-ids ami-0123456789EXAMPLE

The following example output shows the PlatformDetails and UsageOperation fields. In this
example, the ami-0123456789EXAMPLE platform is Red Hat Enterprise Linux and the usage
operation and billing code is RunInstances:0010.

{
"Images": [
{
"VirtualizationType": "hvm",
"Description": "Provided by Red Hat, Inc.",
"Hypervisor": "xen",
"EnaSupport": true,
"SriovNetSupport": "simple",
"ImageId": "ami-0123456789EXAMPLE",
"State": "available",
"BlockDeviceMappings": [
{
"DeviceName": "/dev/sda1",
"Ebs": {
"SnapshotId": "snap-111222333444aaabb",
"DeleteOnTermination": true,
"VolumeType": "gp2",
"VolumeSize": 10,
"Encrypted": false
}
}
],
"Architecture": "x86_64",
"ImageLocation": "123456789012/RHEL-8.0.0_HVM-20190618-x86_64-1-Hourly2-GP2",
"RootDeviceType": "ebs",
"OwnerId": "123456789012",
"PlatformDetails": "Red Hat Enterprise Linux",
"UsageOperation": "RunInstances:0010",
"RootDeviceName": "/dev/sda1",
"CreationDate": "2019-05-10T13:17:12.000Z",
"Public": true,
"ImageType": "machine",
"Name": "RHEL-8.0.0_HVM-20190618-x86_64-1-Hourly2-GP2"

174
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Verify AMI charges on your bill

}
]
}

Verify AMI charges on your bill

To ensure that you're not incurring unplanned costs, you can verify that the billing information for an
instance in your AWS Cost and Usage Report (CUR) matches the billing information that's associated with
the AMI that you used to launch the instance.

To verify the billing information, find the instance ID in your CUR and check the corresponding value
in the lineitem/Operation column. That value should match the value for Usage operation that's
associated with the AMI.

For example, the AMI ami-0123456789EXAMPLE has the following billing information:

• Platform details = Red Hat Enterprise Linux

• Usage operation = RunInstances:0010

If you launched an instance using this AMI, you can find the instance ID in your CUR, and check the
corresponding value in the lineitem/Operation column. In this example, the value should be
RunInstances:0010.

Deregister your Linux AMI

You can deregister an AMI when you have finished using it. After you deregister an AMI, you can't use it
to launch new instances.

When you deregister an AMI, it doesn't affect any instances that you've already launched from the
AMI. You'll continue to incur usage costs for these instances. Therefore, if you are finished with these
instances, you should terminate them.

The procedure that you'll use to clean up your AMI depends on whether it is backed by Amazon EBS or
instance store. For more information, see Determin the root device type of your AMI (p. 102).
Note
An AMI must be owned by your account in order to deregister it.

Contents
• Clean up your Amazon EBS-backed AMI (p. 175)
• Clean up your instance store-backed AMI (p. 177)

Clean up your Amazon EBS-backed AMI

When you deregister an Amazon EBS-backed AMI, it doesn't affect the snapshot(s) that were created for
the volume(s) of the instance during the AMI creation process. You'll continue to incur storage costs for
the snapshots. Therefore, if you are finished with the snapshots, you should delete them.

The following diagram illustrates the process for cleaning up your Amazon EBS-backed AMI.

175
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Clean up your Amazon EBS-backed AMI

To clean up your Amazon EBS-backed AMI

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose AMIs. Select the AMI, and take note of its ID — this can help you find
the correct snapshot in the next step. Choose Actions, and then Deregister. When prompted for
confirmation, choose Continue.

It might take a few minutes before the console removes the AMI from the list. Choose Refresh to
refresh the status.
3. In the navigation pane, choose Snapshots, and select the snapshot (look for the AMI ID in the
Description column). Choose Actions, and then choose Delete Snapshot. When prompted for
confirmation, choose Yes, Delete.

176
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Clean up your instance store-backed AMI

4. (Optional) If you are finished with an instance that you launched from the AMI, terminate it. In the
navigation pane, choose Instances. Select the instance, choose Instance state, Terminate instance.
When prompted for confirmation, choose Terminate.

Clean up your instance store-backed AMI

When you deregister an instance store-backed AMI, it doesn't affect the files that you uploaded to
Amazon S3 when you created the AMI. You'll continue to incur usage costs for these files in Amazon S3.
Therefore, if you are finished with these files, you should delete them.

The following diagram illustrates the process for cleaning up your instance store-backed AMI.

To clean up your instance store-backed AMI

1. Deregister the AMI using the deregister-image command as follows.

177
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Amazon Linux

aws ec2 deregister-image --image-id ami_id

2. Delete the bundle in Amazon S3 using the ec2-delete-bundle (p. 151) (AMI tools) command as
follows.

ec2-delete-bundle -b myawsbucket/myami -a your_access_key_id -s your_secret_access_key

-p image

3. (Optional) If you are finished with an instance that you launched from the AMI, you can terminate it
using the terminate-instances command as follows.

aws ec2 terminate-instances --instance-ids instance_id

4. (Optional) If you are finished with the Amazon S3 bucket that you uploaded the bundle to, you can
delete the bucket. To delete an Amazon S3 bucket, open the Amazon S3 console, select the bucket,
choose Actions, and then choose Delete.

Amazon Linux
Amazon Linux is provided by Amazon Web Services (AWS). It is designed to provide a stable, secure,
and high-performance execution environment for applications running on Amazon EC2. It also includes
packages that enable easy integration with AWS, including launch configuration tools and many popular
AWS libraries and tools. AWS provides ongoing security and maintenance updates for all instances
running Amazon Linux. Many applications developed on CentOS (and similar distributions) run on
Amazon Linux.

Contents
• Amazon Linux availability (p. 178)
• Connect to an Amazon Linux instance (p. 179)
• Identify Amazon Linux images (p. 179)
• AWS command line tools (p. 180)
• Package repository (p. 181)
• Extras library (Amazon Linux 2) (p. 183)
• Access source packages for reference (p. 184)
• cloud-init (p. 184)
• Subscribe to Amazon Linux notifications (p. 186)
• Run Amazon Linux 2 as a virtual machine on premises (p. 187)
• Kernel Live Patching on Amazon Linux 2 (p. 191)

Amazon Linux availability

AWS provides Amazon Linux 2 and the Amazon Linux AMI. If you are migrating from another Linux
distribution to Amazon Linux, we recommend that you migrate to Amazon Linux 2.

The last version of the Amazon Linux AMI, 2018.03, reaches the end of standard support on December
31, 2020. For more information, see the following blog post: Amazon Linux AMI end of life. If you are
currently using the Amazon Linux AMI, we recommend that you migrate to Amazon Linux 2. To migrate
to Amazon Linux 2, launch an instance or create a virtual machine using the current Amazon Linux 2
image. Install your applications, plus any required packages. Test your application, and make any changes
required for it to run on Amazon Linux 2.

178
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect to an Amazon Linux instance

For more information, see Amazon Linux 2 and Amazon Linux AMI. For Amazon Linux Docker container
images, see amazonlinux on Docker Hub.

Connect to an Amazon Linux instance

Amazon Linux does not allow remote root SSH by default. Also, password authentication is disabled
to prevent brute-force password attacks. To enable SSH logins to an Amazon Linux instance, you must
provide your key pair to the instance at launch. You must also set the security group used to launch your
instance to allow SSH access. By default, the only account that can log in remotely using SSH is ec2-user;
this account also has sudo privileges. If you enable remote root login, be aware that it is less secure than
relying on key pairs and a secondary user.

Identify Amazon Linux images

Each image contains a unique /etc/image-id file that identifies it. This file contains the following
information about the image:

• image_name, image_version, image_arch — Values from the build recipe that Amazon used to
construct the image.
• image_stamp — A unique, random hex value generated during image creation.
• image_date — The UTC time of image creation, in YYYYMMDDhhmmss format
• recipe_name, recipe_id — The name and ID of the build recipe Amazon used to construct the
image.

Amazon Linux contains an /etc/system-release file that specifies the current release that is
installed. This file is updated using yum and is part of the system-release RPM.

Amazon Linux also contains a machine-readable version of /etc/system-release that follows the
CPE specification; see /etc/system-release-cpe.

Amazon Linux 2
The following is an example of /etc/image-id for the current version of Amazon Linux 2:

[ec2-user ~]$ cat /etc/image-id

image_name="amzn2-ami-hvm"
image_version="2"
image_arch="x86_64"
image_file="amzn2-ami-hvm-2.0.20180810-x86_64.xfs.gpt"
image_stamp="8008-2abd"
image_date="20180811020321"
recipe_name="amzn2 ami"
recipe_id="c652686a-2415-9819-65fb-4dee-9792-289d-1e2846bd"

The following is an example of /etc/system-release for the current version of Amazon Linux 2:

[ec2-user ~]$ cat /etc/system-release

Amazon Linux 2

The following is an example of /etc/os-release for Amazon Linux 2:

[ec2-user ~]$ cat /etc/os-release

NAME="Amazon Linux"
VERSION="2"

179
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AWS command line tools

ID="amzn"
ID_LIKE="centos rhel fedora"
VERSION_ID="2"
PRETTY_NAME="Amazon Linux 2"
ANSI_COLOR="0;33"
CPE_NAME="cpe:2.3:o:amazon:amazon_linux:2"
HOME_URL="https://amazonlinux.com/"

Amazon Linux AMI

The following is an example of /etc/image-id for the current Amazon Linux AMI:

[ec2-user ~]$ cat /etc/image-id

image_name="amzn-ami-hvm"
image_version="2018.03"
image_arch="x86_64"
image_file="amzn-ami-hvm-2018.03.0.20180811-x86_64.ext4.gpt"
image_stamp="cc81-f2f3"
image_date="20180811012746"
recipe_name="amzn ami"
recipe_id="5b283820-dc60-a7ea-d436-39fa-439f-02ea-5c802dbd"

The following is an example of /etc/system-release for the current Amazon Linux AMI:

[ec2-user ~]$ cat /etc/system-release

Amazon Linux AMI release 2018.03

AWS command line tools

The following command line tools for AWS integration and usage are included in the Amazon Linux AMI,
or in the default repositories for Amazon Linux 2. For the complete list of packages in the Amazon Linux
AMI, see Amazon Linux AMI 2017.09 Packages.

• aws-amitools-ec2
• aws-apitools-as
• aws-apitools-cfn
• aws-apitools-elb
• aws-apitools-mon
• aws-cfn-bootstrap
• aws-cli

Amazon Linux 2 and the minimal versions of Amazon Linux (amzn-ami-minimal-* and amzn2-ami-
minimal-*) do not always contain all of these packages; however, you can install them from the default
repositories using the following command:

[ec2-user ~]$ sudo yum install -y package_name

For instances launched using IAM roles, a simple script has been included to prepare
AWS_CREDENTIAL_FILE, JAVA_HOME, AWS_PATH, PATH, and product-specific environment variables
after a credential file has been installed to simplify the configuration of these tools.

Also, to allow the installation of multiple versions of the API and AMI tools, we have placed symbolic
links to the desired versions of these tools in /opt/aws, as described here:

180
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Package repository

/opt/aws/bin

Symbolic links to /bin directories in each of the installed tools directories.

/opt/aws/{apitools|amitools}

Products are installed in directories of the form name-version and a symbolic link name that is
attached to the most recently installed version.
/opt/aws/{apitools|amitools}/name/environment.sh

Used by /etc/profile.d/aws-apitools-common.sh to set product-specific environment

variables, such as EC2_HOME.

Package repository
Amazon Linux 2 and the Amazon Linux AMI are designed to be used with online package repositories
hosted in each Amazon EC2 AWS Region. These repositories provide ongoing updates to packages in
Amazon Linux 2 and the Amazon Linux AMI, as well as access to hundreds of additional common open-
source server applications. The repositories are available in all Regions and are accessed using yum
update tools. Hosting repositories in each Region enables us to deploy updates quickly and without any
data transfer charges.

Amazon Linux 2 and the Amazon Linux AMI are updated regularly with security and feature
enhancements. If you do not need to preserve data or customizations for your instances, you can simply
launch new instances using the current AMI. If you need to preserve data or customizations for your
instances, you can maintain those instances through the Amazon Linux package repositories. These
repositories contain all the updated packages. You can choose to apply these updates to your running
instances. Older versions of the AMI and update packages continue to be available for use, even as new
versions are released.
Important
Your instance must have access to the internet in order to access the repository.

To install packages, use the following command:

[ec2-user ~]$ sudo yum install package

For the Amazon Linux AMI, access to the Extra Packages for Enterprise Linux (EPEL) repository is
configured, but it is not enabled by default. Amazon Linux 2 is not configured to use the EPEL repository.
EPEL provides third-party packages in addition to those that are in the repositories. The third-party
packages are not supported by AWS. You can enable the EPEL repository with the following commands:

• For Amazon Linux 2:

[ec2-user ~]$ sudo yum install https://dl.fedoraproject.org/pub/epel/epel-release-

latest-7.noarch.rpm

• For the Amazon Linux AMI:

[ec2-user ~]$ sudo yum-config-manager --enable epel

If you find that Amazon Linux does not contain an application you need, you can simply install the
application directly on your Amazon Linux instance. Amazon Linux uses RPMs and yum for package
management, and that is likely the simplest way to install new applications. You should always check to
see if an application is available in our central Amazon Linux repository first, because many applications
are available there. These applications can easily be added to your Amazon Linux instance.

181
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Package repository

To upload your applications onto a running Amazon Linux instance, use scp or sftp and then configure
the application by logging on to your instance. Your applications can also be uploaded during the
instance launch by using the PACKAGE_SETUP action from the built-in cloud-init package. For more
information, see cloud-init (p. 184).

Security updates
Security updates are provided using the package repositories as well as updated AMI security alerts are
published in the Amazon Linux Security Center. For more information about AWS security policies or to
report a security problem, go to the AWS Security Center.

Amazon Linux is configured to download and install critical or important security updates at launch
time. We recommend that you make the necessary updates for your use case after launch. For example,
you may want to apply all updates (not just security updates) at launch, or evaluate each update and
apply only the ones applicable to your system. This is controlled using the following cloud-init setting:
repo_upgrade. The following snippet of cloud-init configuration shows how you can change the
settings in the user data text you pass to your instance initialization:

#cloud-config
repo_upgrade: security

The possible values for repo_upgrade are as follows:

critical

Apply outstanding critical security updates.

important

Apply outstanding critical and important security updates.

medium

Apply outstanding critical, important, and medium security updates.

low

Apply all outstanding security updates, including low-severity security updates.

security

Apply outstanding critical or important updates that Amazon marks as security updates.
bugfix

Apply updates that Amazon marks as bug fixes. Bug fixes are a larger set of updates, which include
security updates and fixes for various other minor bugs.
all

Apply all applicable available updates, regardless of their classification.

none

Do not apply any updates to the instance on startup.

The default setting for repo_upgrade is security. That is, if you don't specify a different value in your
user data, by default, Amazon Linux performs the security upgrades at launch for any packages installed
at that time. Amazon Linux also notifies you of any updates to the installed packages by listing the
number of available updates upon login using the /etc/motd file. To install these updates, you need to
run sudo yum upgrade on the instance.

182
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Extras library (Amazon Linux 2)

Repository configuration
With Amazon Linux, AMIs are treated as snapshots in time, with a repository and update structure that
always gives you the latest packages when you run yum update -y.

The repository structure is configured to deliver a continuous flow of updates that enable you to roll
from one version of Amazon Linux to the next. For example, if you launch an instance from an older
version of the Amazon Linux AMI (such as 2017.09 or earlier) and run yum update -y, you end up with
the latest packages.

You can disable rolling updates by enabling the lock-on-launch feature. The lock-on-launch feature locks
your instance to receive updates only from the specified release of the AMI. For example, you can launch
a 2017.09 AMI and have it receive only the updates that were released prior to the 2018.03 AMI, until
you are ready to migrate to the 2018.03 AMI.
Important
If you lock to a version of the repositories that is not the latest, you do not receive further
updates. To receive a continuous flow of updates, you must use the latest AMI, or consistently
update your AMI with the repositories pointed to latest.

To enable lock-on-launch in new instances, launch it with the following user data passed to cloud-init:

#cloud-config
repo_releasever: 2017.09

To lock existing instances to their current AMI version

1. Edit /etc/yum.conf.
2. Comment out releasever=latest.
3. To clear the cache, run yum clean all.

Extras library (Amazon Linux 2)

With Amazon Linux 2, you can use the Extras Library to install application and software updates on your
instances. These software updates are known as topics. You can install a specific version of a topic or
omit the version information to use the most recent version.

To list the available topics, use the following command:

[ec2-user ~]$ amazon-linux-extras list

To enable a topic and install the latest version of its package to ensure freshness, use the following
command:

[ec2-user ~]$ sudo amazon-linux-extras install topic

To enable topics and install specific versions of their packages to ensure stability, use the following
command:

[ec2-user ~]$ sudo amazon-linux-extras install topic=version topic=version

To remove a package installed from a topic, use the following command:

183
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Access source packages for reference

[ec2-user ~]$ sudo yum remove $(yum list installed | grep amzn2extra-topic | awk '{ print
$1 }')

Note
This command does not remove packages that were installed as dependencies of the extra.

To disable a topic and make the packages inaccessible to the yum package manager, use the following
command:

[ec2-user ~]$ sudo amazon-linux-extras disable topic

Important
This command is intended for advanced users. Improper usage of this command could cause
package compatibility conflicts.

Access source packages for reference

You can view the source of packages you have installed on your instance for reference purposes by
using tools provided in Amazon Linux. Source packages are available for all of the packages included in
Amazon Linux and the online package repository. Simply determine the package name for the source
package you want to install and use the yumdownloader --source command to view source within your
running instance. For example:

[ec2-user ~]$ yumdownloader --source bash

The source RPM can be unpacked, and, for reference, you can view the source tree using standard RPM
tools. After you finish debugging, the package is available for use.

cloud-init
The cloud-init package is an open-source application built by Canonical that is used to bootstrap Linux
images in a cloud computing environment, such as Amazon EC2. Amazon Linux contains a customized
version of cloud-init. It enables you to specify actions that should happen to your instance at boot time.
You can pass desired actions to cloud-init through the user data fields when launching an instance.
This means you can use common AMIs for many use cases and configure them dynamically at startup.
Amazon Linux also uses cloud-init to perform initial configuration of the ec2-user account.

For more information, see the cloud-init documentation.

Amazon Linux uses the cloud-init actions found in /etc/cloud/cloud.cfg.d and /etc/cloud/
cloud.cfg. You can create your own cloud-init action files in /etc/cloud/cloud.cfg.d. All files
in this directory are read by cloud-init. They are read in lexical order, and later files overwrite values in
earlier files.

The cloud-init package performs these (and other) common configuration tasks for instances at boot:

• Set the default locale.

• Set the hostname.
• Parse and handle user data.
• Generate host private SSH keys.
• Add a user's public SSH keys to .ssh/authorized_keys for easy login and administration.
• Prepare the repositories for package management.
• Handle package actions defined in user data.

184
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
cloud-init

• Execute user scripts found in user data.

• Mount instance store volumes, if applicable.
• By default, the ephemeral0 instance store volume is mounted at /media/ephemeral0 if it is
present and contains a valid file system; otherwise, it is not mounted.
• By default, any swap volumes associated with the instance are mounted (only for m1.small and
c1.medium instance types).
• You can override the default instance store volume mount with the following cloud-init directive:

#cloud-config
mounts:
- [ ephemeral0 ]

For more control over mounts, see Mounts in the cloud-init documentation.
• Instance store volumes that support TRIM are not formatted when an instance launches, so you
must partition and format them before you can mount them. For more information, see Instance
store volume TRIM support (p. 1282). You can use the disk_setup module to partition and
format your instance store volumes at boot. For more information, see Disk Setup in the cloud-init
documentation.

Supported user-data formats

The cloud-init package supports user-data handling of a variety of formats:

• Gzip
• If user-data is gzip compressed, cloud-init decompresses the data and handles it appropriately.
• MIME multipart
• Using a MIME multipart file, you can specify more than one type of data. For example, you could
specify both a user-data script and a cloud-config type. Each part of the multipart file can be
handled by cloud-init if it is one of the supported formats.
• Base64 decoding
• If user-data is base64-encoded, cloud-init determines if it can understand the decoded data as
one of the supported types. If it understands the decoded data, it decodes the data and handles it
appropriately. If not, it returns the base64 data intact.
• User-Data script
• Begins with #! or Content-Type: text/x-shellscript.
• The script is run by /etc/init.d/cloud-init-user-scripts during the first boot cycle. This
occurs late in the boot process (after the initial configuration actions are performed).
• Include file
• Begins with #include or Content-Type: text/x-include-url.
• This content is an include file. The file contains a list of URLs, one per line. Each of the URLs is read,
and their content passed through this same set of rules. The content read from the URL can be
gzipped, MIME-multi-part, or plaintext.
• Cloud Config Data
• Begins with #cloud-config or Content-Type: text/cloud-config.
• This content is cloud-config data. For a commented example of supported configuration formats,
see the examples.
• Upstart job
• Begins with #upstart-job or Content-Type: text/upstart-job.
• This content is stored in a file in /etc/init, and upstart consumes the content as per other upstart
jobs.

185
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Subscribe to Amazon Linux notifications

• Cloud Boothook
• Begins with #cloud-boothook or Content-Type: text/cloud-boothook.
• This content is boothook data. It is stored in a file under /var/lib/cloud and then runs
immediately.
• This is the earliest "hook" available. There is no mechanism provided for running it only one time.
The boothook must take care of this itself. It is provided with the instance ID in the environment
variable INSTANCE_ID. Use this variable to provide a once-per-instance set of boothook data.

Subscribe to Amazon Linux notifications

To be notified when new AMIs are released, you can subscribe using Amazon SNS.

To subscribe to Amazon Linux notifications

1. Open the Amazon SNS console at https://console.aws.amazon.com/sns/v3/home.

2. In the navigation bar, change the Region to US East (N. Virginia), if necessary. You must select the
Region in which the SNS notification that you are subscribing to was created.
3. In the navigation pane, choose Subscriptions, Create subscription.
4. For the Create subscription dialog box, do the following:

a. [Amazon Linux 2] For Topic ARN, copy and paste the following Amazon Resource Name (ARN):
arn:aws:sns:us-east-1:137112412989:amazon-linux-2-ami-updates.
b. [Amazon Linux] For Topic ARN, copy and paste the following Amazon Resource Name (ARN):
arn:aws:sns:us-east-1:137112412989:amazon-linux-ami-updates.
c. For Protocol, choose Email.
d. For Endpoint, enter an email address that you can use to receive the notifications.
e. Choose Create subscription.
5. You receive a confirmation email with the subject line "AWS Notification - Subscription
Confirmation". Open the email and choose Confirm subscription to complete your subscription.

Whenever AMIs are released, we send notifications to the subscribers of the corresponding topic. To stop
receiving these notifications, use the following procedure to unsubscribe.

To unsubscribe from Amazon Linux notifications

1. Open the Amazon SNS console at https://console.aws.amazon.com/sns/v3/home.

2. In the navigation bar, change the Region to US East (N. Virginia), if necessary. You must use the
Region in which the SNS notification was created.
3. In the navigation pane, choose Subscriptions, select the subscription, and choose Actions, Delete
subscriptions.
4. When prompted for confirmation, choose Delete.

Amazon Linux AMI SNS message format

The schema for the SNS message is as follows.

{
"description": "Validates output from AMI Release SNS message",
"type": "object",
"properties": {
"v1": {

186
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Run Amazon Linux 2 on premises

"type": "object",
"properties": {
"ReleaseVersion": {
"description": "Major release (ex. 2018.03)",
"type": "string"
},
"ImageVersion": {
"description": "Full release (ex. 2018.03.0.20180412)",
"type": "string"
},
"ReleaseNotes": {
"description": "Human-readable string with extra information",
"type": "string"
},
"Regions": {
"type": "object",
"description": "Each key will be a region name (ex. us-east-1)",
"additionalProperties": {
"type": "array",
"items": {
"type": "object",
"properties": {
"Name": {
"description": "AMI Name (ex. amzn-ami-
hvm-2018.03.0.20180412-x86_64-gp2)",
"type": "string"
},
"ImageId": {
"description": "AMI Name (ex.ami-467ca739)",
"type": "string"
}
},
"required": [
"Name",
"ImageId"
]
}
}
}
},
"required": [
"ReleaseVersion",
"ImageVersion",
"ReleaseNotes",
"Regions"
]
}
},
"required": [
"v1"
]
}

Run Amazon Linux 2 as a virtual machine on premises

Use the Amazon Linux 2 virtual machine (VM) images for on-premises development and testing. These
images are available for use on the following virtualization platforms:

• VMWare
• KVM
• VirtualBox (Oracle VM)
• Microsoft Hyper-V

187
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Run Amazon Linux 2 on premises

To use the Amazon Linux 2 virtual machine images with one of the supported virtualization platforms,
do the following:

• Step 1: Prepare the seed.iso boot image (p. 188)

• Step 2: Download the Amazon Linux 2 VM image (p. 189)
• Step 3: Boot and connect to your new VM (p. 190)

Step 1: Prepare the seed.iso boot image

The seed.iso boot image includes the initial configuration information that is needed to boot your new
VM, such as the network configuration, host name, and user data.
Note
The seed.iso boot image includes only the configuration information required to boot the VM.
It does not include the Amazon Linux 2 operating system files.

To generate the seed.iso boot image, you need two configuration files:

• meta-data—This file includes the hostname and static network settings for the VM.
• user-data—This file configures user accounts, and specifies their passwords, key pairs, and access
mechanisms. By default, the Amazon Linux 2 VM image creates a ec2-user user account. You use the
user-data configuration file to set the password for the default user account.

To create the seed.iso boot disc

1. Create a new folder named seedconfig and navigate into it.

2. Create the meta-data configuration file.

a. Create a new file named meta-data.

b. Open the meta-data file using your preferred editor and add the following.

local-hostname: vm_hostname
# eth0 is the default network interface enabled in the image. You can configure
static network settings with an entry like the following.
network-interfaces: |
auto eth0
iface eth0 inet static
address 192.168.1.10
network 192.168.1.0
netmask 255.255.255.0
broadcast 192.168.1.255
gateway 192.168.1.254

Replace vm_hostname with a VM host name of your choice, and configure the network settings
as required.
c. Save and close the meta-data configuration file.

For an example meta-data configuration file that specifies a VM hostname

(amazonlinux.onprem), configures the default network interface (eth0), and specifies static IP
addresses for the necessary network devices, see the sample Seed.iso file.
3. Create the user-data configuration file.

a. Create a new file named user-data.

b. Open the user-data file using your preferred editor and add the following.

188
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Run Amazon Linux 2 on premises

#cloud-config
#vim:syntax=yaml
users:
# A user by the name `ec2-user` is created in the image by default.
- default
chpasswd:
list: |
ec2-user:plain_text_password
# In the above line, do not add any spaces after 'ec2-user:'.

Replace plain_text_password with a password of your choice for the default ec2-user user
account.
c. (Optional) By default, cloud-init applies network settings each time the VM boots. Add the
following to prevent cloud-init from applying network settings at each boot, and to retain the
network settings applied during the first boot.

# NOTE: Cloud-init applies network settings on every boot by default. To retain

network settings from first
boot, add following ‘write_files’ section:
write_files:
- path: /etc/cloud/cloud.cfg.d/80_disable_network_after_firstboot.cfg
content: |
# Disable network configuration after first boot
network:
config: disabled

d. Save and close the user-data configuration file.

You can also create additional user accounts and specify their access mechanisms, passwords, and
key pairs. For more information about the supported directives, see Modules. For an example user-
data file that creates three additional users and specifies a custom password for the default ec2-
user user account, see the sample Seed.iso file.
4. Create the seed.iso boot image using the meta-data and user-data configuration files.

For Linux, use a tool such as genisoimage. Navigate into the seedconfig folder, and run the
following command.

$ genisoimage -output seed.iso -volid cidata -joliet -rock user-data meta-data

For macOS, use a tool such as hdiutil. Navigate one level up from the seedconfig folder, and run
the following command.

$ hdiutil makehybrid -o seed.iso -hfs -joliet -iso -default-volume-name cidata

seedconfig/

Step 2: Download the Amazon Linux 2 VM image

We offer a different Amazon Linux 2 VM image for each of the supported virtualization platforms.
Download the correct VM image for your chosen platform:

• VMWare
• KVM
• Oracle VirtualBox
• Microsoft Hyper-V

189
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Run Amazon Linux 2 on premises

Step 3: Boot and connect to your new VM

To boot and connect to your new VM, you must have the seed.iso boot image (created in Step 1) and
an Amazon Linux 2 VM image (downloaded in Step 2). The steps vary depending on your chosen VM
platform.

VMWare vSphere

The VM image for VMware is made available in the OVF format.

To boot the VM using VMWare vSphere

1. Create a new datastore for the seed.iso file, or add it to an existing datastore.
2. Deploy the OVF template, but do not start the VM yet.
3. In the Navigator panel, right-click the new virtual machine and choose Edit Settings.
4. On the Virtual Hardware tab, for New device, choose CD/DVD Drive, and then choose Add.
5. For New CD/DVD Drive, choose Datastore ISO File. Select the datastore to which you added the
seed.iso file, browse to and select the seed.iso file, and then choose OK.
6. For New CD/DVD Drive, select Connect, and then choose OK.

After you have associated the datastore with the VM, you should be able to boot it.
KVM

To boot the VM using KVM

1. Open the Create new VM wizard.

2. For Step 1, choose Import existing disk image.
3. For Step 2, browse to and select the VM image. For OS type and Version, choose Linux and Red
Hat Enterprise Linux 7.0 respectively.
4. For Step 3, specify the amount of RAM and the number of CPUs to use.
5. For Step 4, enter a name for the new VM and select Customize configuration before install,
and choose Finish.
6. In the Configuration window for the VM, choose Add Hardware.
7. In the Add New Virtual Hardware window, choose Storage.
8. In the Storage configuration, choose Select or create custom storage. For Device type, choose
CDROM device. Choose Manage, Browse Local, and then navigate to and select the seed.iso
file. Choose Finish.
9. Choose Begin Installation.

Oracle VirtualBox

To boot the VM using Oracle VirtualBox

1. Open Oracle VirtualBox and choose New.

2. For Name, enter a descriptive name for the virtual machine, and for Type and Version, select
Linux and Red Hat (64-bit) respectively. Choose Continue.
3. For Memory size, specify the amount of memory to allocate to the virtual machine, and then
choose Continue.
4. For Hard disk, choose Use an existing virtual hard disk file, browse to and open the VM image,
and then choose Create.
5. Before you start the VM, you must load the seed.iso file in the virtual machine's virtual optical
drive:

190
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Kernel Live Patching

a. Select the new VM, choose Settings, and then choose Storage.
b. In the Storage Devices list, under Controller: IDE, choose the Empty optical drive.
c. In the Attributes section for the optical drive, choose the browse button, select Choose
Virtual Optical Disk File, and then select the seed.iso file. Choose OK to apply the
changes and close the Settings.

After you have added the seed.iso file to the virtual optical drive, you should be able to start the
VM.
Microsoft Hyper-V

The VM image for Microsoft Hyper-V is compressed into a zip file. You must extract the contents of
the zip file.

To boot the VM using Microsoft Hyper-V

1. Open the New Virtual Machine Wizard.

2. When prompted to select a generation, select Generation 1.
3. When prompted to configure the network adapter, for Connection choose External.
4. When prompted to connect a virtual hard disk, choose Use an existing virtual hard disk, choose
Browse, and then navigate to and select the VM image. Choose Finish to create the VM.
5. Right-click the new VM and choose Settings. In the Settings window, under IDE Controller 1,
choose DVD Drive.
6. For the DVD drive, choose Image file and then browse to and select the seed.iso file.
7. Apply the changes and start the VM.

After the VM has booted, log in using one of the user accounts that is defined in the user-data
configuration file. For virtualization platforms other than VMWare, you can disconnect the seed.iso
boot image from the VM after you have logged in for the first time.

Kernel Live Patching on Amazon Linux 2

Kernel Live Patching for Amazon Linux 2 enables you to apply security vulnerability and critical bug
patches to a running Linux kernel, without reboots or disruptions to running applications. This allows you
to benefit from improved service and application availability, while keeping your infrastructure secure
and up to date.

AWS releases two types of kernel live patches for Amazon Linux 2:

• Security updates—Include updates for Linux common vulnerabilities and exposures (CVE). These
updates are typically rated as important or critical using the Amazon Linux Security Advisory ratings.
They generally map to a Common Vulnerability Scoring System (CVSS) score of 7 and higher. In some
cases, AWS might provide updates before a CVE is assigned. In these cases, the patches might appear
as bug fixes.
• Bug fixes—Include fixes for critical bugs and stability issues that are not associated with CVEs.

AWS provides kernel live patches for an Amazon Linux 2 kernel version for up to 3 months after its
release. After the 3-month period, you must update to a later kernel version to continue to receive kernel
live patches.

Amazon Linux 2 kernel live patches are made available as signed RPM packages in the existing Amazon
Linux 2 repositories. The patches can be installed on individual instances using existing yum workflows,
or they can be installed on a group of managed instances using AWS Systems Manager.

191
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Kernel Live Patching

Kernel Live Patching on Amazon Linux 2 is provided at no additional cost.

Topics
• Supported configurations and prerequisites (p. 192)
• Work with Kernel Live Patching (p. 192)
• Limitations (p. 196)
• Frequently asked questions (p. 196)

Supported configurations and prerequisites

Kernel Live Patching is supported on Amazon EC2 instances and on-premises virtual machines (p. 187)
running Amazon Linux 2.

To use Kernel Live Patching on Amazon Linux 2, you must use:

• A 64-bit (x86_64) architecture that is supported by Amazon Linux 2

• Amazon Linux 2 with kernel version 4.14.165-131.185 or later

Note
The 64-bit ARM (arm64) architecture is not supported.

Work with Kernel Live Patching

You can enable and use Kernel Live Patching on individual instances using the command line on the
instance itself, or you can enable and use Kernel Live Patching on a group of managed instances using
AWS Systems Manager.

The following sections explain how to enable and use Kernel Live Patching on individual instances using
the command line.

For more information about enabling and using Kernel Live Patching on a group of managed instances,
see Use Kernel Live Patching on Amazon Linux 2 instances in the AWS Systems Manager User Guide.

Topics
• Enable Kernel Live Patching (p. 192)
• View the available kernel live patches (p. 193)
• Apply kernel live patches (p. 194)
• View the applied kernel live patches (p. 195)
• Disable Kernel Live Patching (p. 195)

Enable Kernel Live Patching

Kernel Live Patching is disabled by default on Amazon Linux 2. To use live patching, you must install the
yum plugin for Kernel Live Patching and enable the live patching functionality.

Prerequisites

Kernel Live Patching requires binutils. If you do not have binutils installed, install it using the
following command:

$ sudo yum install binutils

192
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Kernel Live Patching

To enable Kernel Live Patching

1. Kernel live patches are available for Amazon Linux 2 with kernel version 4.14.165-131.185 or
later. To check your kernel version, run the following command.

$ sudo yum list kernel

2. If you already have a supported kernel version, skip this step. If you do not have a supported kernel
version, run the following commands to update the kernel to the latest version and to reboot the
instance.

$ sudo yum install -y kernel

$ sudo reboot

3. Install the yum plugin for Kernel Live Patching.

$ sudo yum install -y yum-plugin-kernel-livepatch

4. Enable the yum plugin for Kernel Live Patching.

$ sudo yum kernel-livepatch enable -y

This command also installs the latest version of the kernel live patch RPM from the configured
repositories.
5. To confirm that the yum plugin for kernel live patching has installed successfully, run the following
command.

$ rpm -qa | grep kernel-livepatch

When you enable Kernel Live Patching, an empty kernel live patch RPM is automatically applied. If
Kernel Live Patching was successfully enabled, this command returns a list that includes the initial
empty kernel live patch RPM.
6. Update and start the kpatch service. This service loads all of the kernel live patches upon
initialization or at boot.

$ sudo yum update kpatch-runtime

$ sudo systemctl enable kpatch.service

7. Configure the Amazon Linux 2 Kernel Live Patching repository, which contains the kernel live
patches.

$ sudo amazon-linux-extras enable livepatch

View the available kernel live patches

Amazon Linux security alerts are published to the Amazon Linux Security Center. For more information
about the Amazon Linux 2 security alerts, which include alerts for kernel live patches, see the Amazon
Linux Security Center. Kernel live patches are prefixed with ALASLIVEPATCH. The Amazon Linux Security
Center might not list kernel live patches that address bugs.

You can also discover the available kernel live patches for advisories and CVEs using the command line.

193
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Kernel Live Patching

To list all available kernel live patches for advisories

Use the following command.

$ yum updateinfo list

The following shows example output.

Loaded plugins: extras_suggestions, kernel-livepatch, langpacks, priorities, update-motd

ALAS2LIVEPATCH-2020-002 important/Sec. kernel-
livepatch-4.14.165-133.209-1.0-3.amzn2.x86_64
ALAS2LIVEPATCH-2020-005 medium/Sec. kernel-livepatch-4.14.165-133.209-1.0-4.amzn2.x86_64
updateinfo list done

To list all available kernel live patches for CVEs

Use the following command.

$ yum updateinfo list cves

The following shows example output.

Loaded plugins: extras_suggestions, kernel-livepatch, langpacks, priorities, update-

motdamzn2-core/2/x86_64 | 2.4 kB 00:00:00
CVE-2019-15918 important/Sec. kernel-livepatch-4.14.165-133.209-1.0-3.amzn2.x86_64
CVE-2019-20096 important/Sec. kernel-livepatch-4.14.165-133.209-1.0-3.amzn2.x86_64
CVE-2020-8648 medium/Sec. kernel-livepatch-4.14.165-133.209-1.0-4.amzn2.x86_64
updateinfo list done

Apply kernel live patches

You apply kernel live patches using the yum package manager in the same way that you would apply
regular updates. The yum plugin for Kernel Live Patching manages the kernel live patches that are to be
applied and eliminates the need to reboot.
Tip
We recommend that you update your kernel regularly using Kernel Live Patching to ensure that
it remains secure and up to date.

You can choose to apply a specific kernel live patch, or to apply any available kernel live patches along
with your regular security updates.

To apply a specific kernel live patch

1. Get the kernel live patch version using one of the commands described in View the available kernel
live patches (p. 193).
2. Apply the kernel live patch for your Amazon Linux 2 kernel.

$ sudo yum install kernel-livepatch-kernel_version.x86_64

For example, the following command applies a kernel live patch for Amazon Linux 2 kernel version
4.14.165-133.209.

$ sudo yum install kernel-livepatch-4.14.165-133.209-1.0-4.amzn2.x86_64

To apply any available kernel live patches along with your regular security updates

194
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Kernel Live Patching

Use the following command.

$ sudo yum update --security

Omit the --security option to include bug fixes.

Important

• The kernel version is not updated after applying kernel live patches. The version is only
updated to the new version after the instance is rebooted.
• An Amazon Linux 2 kernel receives kernel live patches for a period of three months. After the
three month period has lapsed, no new kernel live patches are released for that kernel version.
To continue to receive kernel live patches after the three-month period, you must reboot the
instance to move to the new kernel version, which will then continue receiving kernel live
patches for the next three months. To check the support window for your kernel version, run
yum kernel-livepatch supported.

View the applied kernel live patches

To view the applied kernel live patches

Use the following command.

$ kpatch list

The command returns a list of the loaded and installed security update kernel live patches. The following
is example output.

Loaded patch modules:

livepatch_cifs_lease_buffer_len [enabled]
livepatch_CVE_2019_20096 [enabled]
livepatch_CVE_2020_8648 [enabled]

Installed patch modules:

livepatch_cifs_lease_buffer_len (4.14.165-133.209.amzn2.x86_64)
livepatch_CVE_2019_20096 (4.14.165-133.209.amzn2.x86_64)
livepatch_CVE_2020_8648 (4.14.165-133.209.amzn2.x86_64)

Note
A single kernel live patch can include and install multiple live patches.

Disable Kernel Live Patching

If you no longer need to use Kernel Live Patching, you can disable it at any time.

To disable Kernel Live Patching

1. Remove the RPM packages for the applied kernel live patches.

$ sudo yum kernel-livepatch disable

2. Uninstall the yum plugin for Kernel Live Patching.

$ sudo yum remove yum-plugin-kernel-livepatch

3. Reboot the instance.

195
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
User provided kernels

$ sudo reboot

Limitations
Kernel Live Patching has the following limitations:

• While applying a kernel live patch, you can't perform hibernation, use advanced debugging tools (such
as SystemTap, kprobes, and eBPF-based tools), or access ftrace output files used by the Kernel Live
Patching infrastructure.
• Amazon Linux 2 instances with 64-bit ARM (arm64) architecture are not supported.

Frequently asked questions

For frequently asked questions about Kernel Live Patching for Amazon Linux 2, see the Amazon Linux 2
Kernel Live Patching FAQ.

User provided kernels

If you need a custom kernel on your Amazon EC2 instances, you can start with an AMI that is close to
what you want, compile the custom kernel on your instance, and update the bootloader to point to
the new kernel. This process varies depending on the virtualization type that your AMI uses. For more
information, see Linux AMI virtualization types (p. 104).

Contents
• HVM AMIs (GRUB) (p. 196)
• Paravirtual AMIs (PV-GRUB) (p. 196)

HVM AMIs (GRUB)

HVM instance volumes are treated like actual physical disks. The boot process is similar to that of a
bare metal operating system with a partitioned disk and bootloader, which enables it to work with all
currently supported Linux distributions. The most common bootloader is GRUB or GRUB2.

By default, GRUB does not send its output to the instance console because it creates an extra boot delay.
For more information, see Instance console output (p. 1364). If you are installing a custom kernel, you
should consider enabling GRUB output.

You don't need to specify a fallback kernel, but we recommend that you have a fallback when you test a
new kernel. GRUB can fall back to another kernel in the event that the new kernel fails. Having a fallback
kernel enables the instance to boot even if the new kernel isn't found.

The legacy GRUB for Amazon Linux uses /boot/grub/menu.lst. GRUB2 for Amazon Linux 2 uses /
etc/default/grub. For more information about updating the default kernel in the bootloader, see the
documentation for your Linux distribution.

Paravirtual AMIs (PV-GRUB)

Amazon Machine Images that use paravirtual (PV) virtualization use a system called PV-GRUB during the
boot process. PV-GRUB is a paravirtual bootloader that runs a patched version of GNU GRUB 0.97. When

196
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Paravirtual AMIs (PV-GRUB)

you start an instance, PV-GRUB starts the boot process and then chain loads the kernel specified by your
image's menu.lst file.

PV-GRUB understands standard grub.conf or menu.lst commands, which allows it to work with all
currently supported Linux distributions. Older distributions such as Ubuntu 10.04 LTS, Oracle Enterprise
Linux, or CentOS 5.x require a special "ec2" or "xen" kernel package, while newer distributions include the
required drivers in the default kernel package.

Most modern paravirtual AMIs use a PV-GRUB AKI by default (including all of the paravirtual Linux AMIs
available in the Amazon EC2 Launch Wizard Quick Start menu), so there are no additional steps that
you need to take to use a different kernel on your instance, provided that the kernel you want to use
is compatible with your distribution. The best way to run a custom kernel on your instance is to start
with an AMI that is close to what you want and then to compile the custom kernel on your instance and
modify the menu.lst file to boot with that kernel.

You can verify that the kernel image for an AMI is a PV-GRUB AKI. Run the following describe-images
command (substituting your kernel image ID) and check whether the Name field starts with pv-grub:

aws ec2 describe-images --filters Name=image-id,Values=aki-880531cd

Contents
• Limitations of PV-GRUB (p. 197)
• Configure GRUB for paravirtual AMIs (p. 197)
• Amazon PV-GRUB Kernel Image IDs (p. 198)
• Update PV-GRUB (p. 200)

Limitations of PV-GRUB
PV-GRUB has the following limitations:

• You can't use the 64-bit version of PV-GRUB to start a 32-bit kernel or vice versa.
• You can't specify an Amazon ramdisk image (ARI) when using a PV-GRUB AKI.
• AWS has tested and verified that PV-GRUB works with these file system formats: EXT2, EXT3, EXT4,
JFS, XFS, and ReiserFS. Other file system formats might not work.
• PV-GRUB can boot kernels compressed using the gzip, bzip2, lzo, and xz compression formats.
• Cluster AMIs don't support or need PV-GRUB, because they use full hardware virtualization (HVM).
While paravirtual instances use PV-GRUB to boot, HVM instance volumes are treated like actual disks,
and the boot process is similar to the boot process of a bare metal operating system with a partitioned
disk and bootloader.
• PV-GRUB versions 1.03 and earlier don't support GPT partitioning; they support MBR partitioning only.
• If you plan to use a logical volume manager (LVM) with Amazon Elastic Block Store (Amazon EBS)
volumes, you need a separate boot partition outside of the LVM. Then you can create logical volumes
with the LVM.

Configure GRUB for paravirtual AMIs

To boot PV-GRUB, a GRUB menu.lst file must exist in the image; the most common location for this file
is /boot/grub/menu.lst.

The following is an example of a menu.lst configuration file for booting an AMI with a PV-GRUB
AKI. In this example, there are two kernel entries to choose from: Amazon Linux 2018.03 (the original
kernel for this AMI), and Vanilla Linux 4.16.4 (a newer version of the Vanilla Linux kernel from https://
www.kernel.org/). The Vanilla entry was copied from the original entry for this AMI, and the kernel

197
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Paravirtual AMIs (PV-GRUB)

and initrd paths were updated to the new locations. The default 0 parameter points the bootloader
to the first entry it sees (in this case, the Vanilla entry), and the fallback 1 parameter points the
bootloader to the next entry if there is a problem booting the first.

default 0
fallback 1
timeout 0
hiddenmenu

title Vanilla Linux 4.16.4

root (hd0)
kernel /boot/vmlinuz-4.16.4 root=LABEL=/ console=hvc0
initrd /boot/initrd.img-4.16.4

title Amazon Linux 2018.03 (4.14.26-46.32.amzn1.x86_64)

root (hd0)
kernel /boot/vmlinuz-4.14.26-46.32.amzn1.x86_64 root=LABEL=/ console=hvc0
initrd /boot/initramfs-4.14.26-46.32.amzn1.x86_64.img

You don't need to specify a fallback kernel in your menu.lst file, but we recommend that you have a
fallback when you test a new kernel. PV-GRUB can fall back to another kernel in the event that the new
kernel fails. Having a fallback kernel allows the instance to boot even if the new kernel isn't found.

PV-GRUB checks the following locations for menu.lst, using the first one it finds:

• (hd0)/boot/grub
• (hd0,0)/boot/grub
• (hd0,0)/grub
• (hd0,1)/boot/grub
• (hd0,1)/grub
• (hd0,2)/boot/grub
• (hd0,2)/grub
• (hd0,3)/boot/grub
• (hd0,3)/grub

Note that PV-GRUB 1.03 and earlier only check one of the first two locations in this list.

Amazon PV-GRUB Kernel Image IDs

PV-GRUB AKIs are available in all Amazon EC2 regions. There are AKIs for both 32-bit and 64-bit
architecture types. Most modern AMIs use a PV-GRUB AKI by default.

We recommend that you always use the latest version of the PV-GRUB AKI, as not all versions of the PV-
GRUB AKI are compatible with all instance types. Use the following describe-images command to get a
list of the PV-GRUB AKIs for the current region:

aws ec2 describe-images --owners amazon --filters Name=name,Values=pv-grub-*.gz

PV-GRUB is the only AKI available in the ap-southeast-2 Region. You should verify that any AMI you
want to copy to this Region is using a version of PV-GRUB that is available in this Region.

The following are the current AKI IDs for each Region. Register new AMIs using an hd0 AKI.
Note
We continue to provide hd00 AKIs for backward compatibility in Regions where they were
previously available.

198
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Paravirtual AMIs (PV-GRUB)

ap-northeast-1, Asia Pacific (Tokyo)

Image ID Image Name

aki-f975a998 pv-grub-hd0_1.05-i386.gz

aki-7077ab11 pv-grub-hd0_1.05-x86_64.gz

ap-southeast-1, Asia Pacific (Singapore) Region

Image ID Image Name

aki-17a40074 pv-grub-hd0_1.05-i386.gz

aki-73a50110 pv-grub-hd0_1.05-x86_64.gz

ap-southeast-2, Asia Pacific (Sydney)

Image ID Image Name

aki-ba5665d9 pv-grub-hd0_1.05-i386.gz

aki-66506305 pv-grub-hd0_1.05-x86_64.gz

eu-central-1, Europe (Frankfurt)

Image ID Image Name

aki-1419e57b pv-grub-hd0_1.05-i386.gz

aki-931fe3fc pv-grub-hd0_1.05-x86_64.gz

eu-west-1, Europe (Ireland)

Image ID Image Name

aki-1c9fd86f pv-grub-hd0_1.05-i386.gz

aki-dc9ed9af pv-grub-hd0_1.05-x86_64.gz

sa-east-1, South America (São Paulo)

Image ID Image Name

aki-7cd34110 pv-grub-hd0_1.05-i386.gz

aki-912fbcfd pv-grub-hd0_1.05-x86_64.gz

us-east-1, US East (N. Virginia)

Image ID Image Name

aki-04206613 pv-grub-hd0_1.05-i386.gz

199
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Paravirtual AMIs (PV-GRUB)

Image ID Image Name

aki-5c21674b pv-grub-hd0_1.05-x86_64.gz

us-gov-west-1, AWS GovCloud (US-West)

Image ID Image Name

aki-5ee9573f pv-grub-hd0_1.05-i386.gz

aki-9ee55bff pv-grub-hd0_1.05-x86_64.gz

us-west-1, US West (N. California)

Image ID Image Name

aki-43cf8123 pv-grub-hd0_1.05-i386.gz

aki-59cc8239 pv-grub-hd0_1.05-x86_64.gz

us-west-2, US West (Oregon)

Image ID Image Name

aki-7a69931a pv-grub-hd0_1.05-i386.gz

aki-70cb0e10 pv-grub-hd0_1.05-x86_64.gz

Update PV-GRUB
We recommend that you always use the latest version of the PV-GRUB AKI, as not all versions of the PV-
GRUB AKI are compatible with all instance types. Also, older versions of PV-GRUB are not available in all
regions, so if you copy an AMI that uses an older version to a Region that does not support that version,
you will be unable to boot instances launched from that AMI until you update the kernel image. Use the
following procedures to check your instance's version of PV-GRUB and update it if necessary.

To check your PV-GRUB version

1. Find the kernel ID for your instance.

aws ec2 describe-instance-attribute --instance-id instance_id --attribute kernel --

region region

{
"InstanceId": "instance_id",
"KernelId": "aki-70cb0e10"
}

The kernel ID for this instance is aki-70cb0e10.

2. View the version information of that kernel ID.

aws ec2 describe-images --image-ids aki-70cb0e10 --region region

200
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Use the MATE desktop environment

"Images": [
{
"VirtualizationType": "paravirtual",
"Name": "pv-grub-hd0_1.05-x86_64.gz",
...
"Description": "PV-GRUB release 1.05, 64-bit"
}
]
}

This kernel image is PV-GRUB 1.05. If your PV-GRUB version is not the newest version (as shown in
Amazon PV-GRUB Kernel Image IDs (p. 198)), you should update it using the following procedure.

To update your PV-GRUB version

If your instance is using an older version of PV-GRUB, you should update it to the latest version.

1. Identify the latest PV-GRUB AKI for your Region and processor architecture from Amazon PV-GRUB
Kernel Image IDs (p. 198).
2. Stop your instance. Your instance must be stopped to modify the kernel image used.

aws ec2 stop-instances --instance-ids instance_id --region region

3. Modify the kernel image used for your instance.

aws ec2 modify-instance-attribute --instance-id instance_id --kernel kernel_id --

region region

4. Restart your instance.

aws ec2 start-instances --instance-ids instance_id --region region

Use the MATE desktop environment provided with

Amazon Linux 2
The MATE desktop environment is pre-installed and pre-configured in the AMI with the following
description: Amazon Linux 2 with .Net Core, PowerShell, Mono, and MATE Desktop Environment. The
environment provides an intuitive graphical user interface for administering Amazon Linux 2 instances
without using the command line. The interface uses graphical representations, such as icons, windows,
toolbars, folders, wallpapers, and desktop widgets. Built-in, GUI-based tools are available to perform
common tasks. For example, there are tools for adding and removing software, applying updates,
organizing files, launching programs, and monitoring system health.

Complete the following procedure to use the MATE desktop environment.

To configure Remote Desktop Protocol (RDP) connections and set up a password

1. Use the following describe-images command to get the ID of the AMI for Amazon Linux 2 that
includes MATE in the AMI name.

aws ec2 describe-images --filters "Name=name,Values=amzn2*MATE*" --query

'Images[*].ImageId' --output text

The following is example output:

201
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Use the MATE desktop environment

ami-0abcdef1234567890

2. Launch an EC2 instance with the AMI that you located in the previous step. Configure the security
group to allow for inbound TCP traffic to port 3389. For more information about configuring
security groups, see Security groups for your VPC. This configuration enables you to use an RDP
client to connect to the instance.
3. Connect to the instance using SSH. Run the following command to set the password for ec2-user.

[ec2-user ~]$ sudo passwd ec2-user

4. Open an RDP client on the computer from which you will connect to the instance (for example,
Remote Desktop Connection on a computer running Microsoft Windows). Enter ec2-user as the
user name and enter the password that you set in the previous step.

To disable the MATE Desktop Environment on your EC2 instance

You can turn off the GUI environment at any time by running one of the following commands.

[ec2-user ~]$ sudo systemctl disable xrdp

[ec2-user ~]$ sudo systemctl stop xrdp

To enable the MATE Desktop Environment on your EC2 instance

To turn the GUI back on, you can run one of the following commands.

[ec2-user ~]$ sudo systemctl enable xrdp

[ec2-user ~]$ sudo systemctl start xrdp

202
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance types

Amazon EC2 instances

If you're new to Amazon EC2, see the following topics to get started:

• What is Amazon EC2? (p. 1)

• Set up to use Amazon EC2 (p. 28)
• Tutorial: Get started with Amazon EC2 Linux instances (p. 32)
• Instance lifecycle (p. 520)

Before you launch a production environment, you need to answer the following questions.

Q. What instance type best meets my needs?

Amazon EC2 provides different instance types to enable you to choose the CPU, memory, storage,
and networking capacity that you need to run your applications. For more information, see Instance
types (p. 203).
Q. What purchasing option best meets my needs?

Amazon EC2 supports On-Demand Instances (the default), Spot Instances, and Reserved Instances.
For more information, see Instance purchasing options (p. 320).
Q. Which type of root volume meets my needs?

Each instance is backed by Amazon EBS or backed by instance store. Select an AMI based on which
type of root volume you need. For more information, see Storage for the root device (p. 102).
Q. Can I remotely manage a fleet of EC2 instances and machines in my hybrid environment?

AWS Systems Manager enables you to remotely and securely manage the configuration of your
Amazon EC2 instances, and your on-premises instances and virtual machines (VMs) in hybrid
environments, including VMs from other cloud providers. For more information, see the AWS
Systems Manager User Guide.

Instance types
When you launch an instance, the instance type that you specify determines the hardware of the host
computer used for your instance. Each instance type offers different compute, memory, and storage
capabilities and are grouped in instance families based on these capabilities. Select an instance type
based on the requirements of the application or software that you plan to run on your instance.

Amazon EC2 provides each instance with a consistent and predictable amount of CPU capacity,
regardless of its underlying hardware.

Amazon EC2 dedicates some resources of the host computer, such as CPU, memory, and instance
storage, to a particular instance. Amazon EC2 shares other resources of the host computer, such as the
network and the disk subsystem, among instances. If each instance on a host computer tries to use
as much of one of these shared resources as possible, each receives an equal share of that resource.
However, when a resource is underused, an instance can consume a higher share of that resource while
it's available.

Each instance type provides higher or lower minimum performance from a shared resource. For example,
instance types with high I/O performance have a larger allocation of shared resources. Allocating a larger
share of shared resources also reduces the variance of I/O performance. For most applications, moderate
I/O performance is more than enough. However, for applications that require greater or more consistent
I/O performance, consider an instance type with higher I/O performance.

203
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Available instance types

Contents
• Available instance types (p. 204)
• Hardware specifications (p. 207)
• AMI virtualization types (p. 208)
• Instances built on the Nitro System (p. 209)
• Networking and storage features (p. 209)
• Instance limits (p. 213)
• General purpose instances (p. 213)
• Compute optimized instances (p. 264)
• Memory optimized instances (p. 272)
• Storage optimized instances (p. 283)
• Linux accelerated computing instances (p. 290)
• Find an Amazon EC2 instance type (p. 310)
• Change the instance type (p. 311)
• Get recommendations for an instance type (p. 317)

Available instance types

Amazon EC2 provides a wide selection of instance types optimized for different use cases. To determine
which instance types meet your requirements, such as supported Regions, compute resources, or storage
resources, see Find an Amazon EC2 instance type (p. 310).

Current generation instances

For the best performance, we recommend that you use the following instance types when you launch
new instances. For more information, see Amazon EC2 Instance Types.

Type Sizes Use case

C4 c4.large | c4.xlarge | c4.2xlarge | c4.4xlarge | c4.8xlarge Compute

optimized (p. 264)

C5 c5.large | c5.xlarge | c5.2xlarge | c5.4xlarge | c5.9xlarge | Compute

c5.12xlarge | c5.18xlarge | c5.24xlarge | c5.metal optimized (p. 264)

C5a c5a.large | c5a.xlarge | c5a.2xlarge | c5a.4xlarge | c5a.8xlarge Compute

| c5a.12xlarge | c5a.16xlarge | c5a.24xlarge optimized (p. 264)

C5ad c5ad.large | c5ad.xlarge | c5ad.2xlarge | c5ad.4xlarge | Compute

c5ad.8xlarge | c5ad.12xlarge | c5ad.16xlarge | c5ad.24xlarge optimized (p. 264)

C5d c5d.large | c5d.xlarge | c5d.2xlarge | c5d.4xlarge | c5d.9xlarge Compute

| c5d.12xlarge | c5d.18xlarge | c5d.24xlarge | c5d.metal optimized (p. 264)

C5n c5n.large | c5n.xlarge | c5n.2xlarge | c5n.4xlarge | c5n.9xlarge Compute

| c5n.18xlarge | c5n.metal optimized (p. 264)

C6g c6g.medium | c6g.large | c6g.xlarge | c6g.2xlarge | c6g.4xlarge | Compute

c6g.8xlarge | c6g.12xlarge | c6g.16xlarge | c6g.metal optimized (p. 264)

C6gd c6gd.medium | c6gd.large | c6gd.xlarge | c6gd.2xlarge | Compute

c6gd.4xlarge | c6gd.8xlarge | c6gd.12xlarge | c6gd.16xlarge | optimized (p. 264)
c6gd.metal

204
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Available instance types

Type Sizes Use case

C6gn c6gn.medium | c6gn.large | c6gn.xlarge | c6gn.2xlarge | Compute

c6gn.4xlarge | c6gn.8xlarge | c6gn.12xlarge | c6gn.16xlarge optimized (p. 264)

D2 d2.xlarge | d2.2xlarge | d2.4xlarge | d2.8xlarge Storage

optimized (p. 283)

D3 d3.xlarge | d3.2xlarge | d3.4xlarge | d3.8xlarge Storage

optimized (p. 283)

D3en d3en.large | d3en.xlarge | d3en.2xlarge | d3en.4xlarge | Storage

d3en.6xlarge | d3en.8xlarge | d3en.12xlarge optimized (p. 283)

F1 f1.2xlarge | f1.4xlarge | f1.16xlarge Accelerated

computing (p. 290)

G3 g3s.xlarge | g3.4xlarge | g3.8xlarge | g3.16xlarge Accelerated

computing (p. 290)

G4ad g4ad.4xlarge | g4ad.8xlarge | g4ad.16xlarge Accelerated

computing (p. 290)

G4dn g4dn.xlarge | g4dn.2xlarge | g4dn.4xlarge | g4dn.8xlarge | Accelerated

g4dn.12xlarge | g4dn.16xlarge | g4dn.metal computing (p. 290)

H1 h1.2xlarge | h1.4xlarge | h1.8xlarge | h1.16xlarge Storage

optimized (p. 283)

I3 i3.large | i3.xlarge | i3.2xlarge | i3.4xlarge | i3.8xlarge | Storage

i3.16xlarge | i3.metal optimized (p. 283)

I3en i3en.large | i3en.xlarge | i3en.2xlarge | i3en.3xlarge | Storage

i3en.6xlarge | i3en.12xlarge | i3en.24xlarge | i3en.metal optimized (p. 283)

Inf1 inf1.xlarge | inf1.2xlarge | inf1.6xlarge | inf1.24xlarge Accelerated

computing (p. 290)

M4 m4.large | m4.xlarge | m4.2xlarge | m4.4xlarge | m4.10xlarge | General

m4.16xlarge purpose (p. 213)

M5 m5.large | m5.xlarge | m5.2xlarge | m5.4xlarge | m5.8xlarge | General

m5.12xlarge | m5.16xlarge | m5.24xlarge | m5.metal purpose (p. 213)

M5a m5a.large | m5a.xlarge | m5a.2xlarge | m5a.4xlarge | m5a.8xlarge General

| m5a.12xlarge | m5a.16xlarge | m5a.24xlarge purpose (p. 213)

M5ad m5ad.large | m5ad.xlarge | m5ad.2xlarge | m5ad.4xlarge | General

m5ad.8xlarge | m5ad.12xlarge | m5ad.16xlarge | m5ad.24xlarge purpose (p. 213)

M5d m5d.large | m5d.xlarge | m5d.2xlarge | m5d.4xlarge | m5d.8xlarge General

| m5d.12xlarge | m5d.16xlarge | m5d.24xlarge | m5d.metal purpose (p. 213)

M5dn m5dn.large | m5dn.xlarge | m5dn.2xlarge | m5dn.4xlarge | General

m5dn.8xlarge | m5dn.12xlarge | m5dn.16xlarge | m5dn.24xlarge purpose (p. 213)

M5n m5n.large | m5n.xlarge | m5n.2xlarge | m5n.4xlarge | m5n.8xlarge General

| m5n.12xlarge | m5n.16xlarge | m5n.24xlarge purpose (p. 213)

M5zn m5zn.large | m5zn.xlarge | m5zn.2xlarge | m5zn.3xlarge | General

m5zn.6xlarge | m5zn.12xlarge | m5zn.metal purpose (p. 213)

205
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Available instance types

Type Sizes Use case

M6g m6g.medium | m6g.large | m6g.xlarge | m6g.2xlarge | m6g.4xlarge | General

m6g.8xlarge | m6g.12xlarge | m6g.16xlarge | m6g.metal purpose (p. 213)

M6gd m6gd.medium | m6gd.large | m6gd.xlarge | m6gd.2xlarge | General

m6gd.4xlarge | m6gd.8xlarge | m6gd.12xlarge | m6gd.16xlarge | purpose (p. 213)
m6gd.metal

Mac1 mac1.metal General

purpose (p. 213)

P2 p2.xlarge | p2.8xlarge | p2.16xlarge Accelerated

computing (p. 290)

P3 p3.2xlarge | p3.8xlarge | p3.16xlarge Accelerated

computing (p. 290)

P3dn p3dn.24xlarge Accelerated

computing (p. 290)

P4d p4d.24xlarge Accelerated

computing (p. 290)

R4 r4.large | r4.xlarge | r4.2xlarge | r4.4xlarge | r4.8xlarge | Memory

r4.16xlarge optimized (p. 272)

R5 r5.large | r5.xlarge | r5.2xlarge | r5.4xlarge | r5.8xlarge | Memory

r5.12xlarge | r5.16xlarge | r5.24xlarge | r5.metal optimized (p. 272)

R5a r5a.large | r5a.xlarge | r5a.2xlarge | r5a.4xlarge | r5a.8xlarge Memory

| r5a.12xlarge | r5a.16xlarge | r5a.24xlarge optimized (p. 272)

R5ad r5ad.large | r5ad.xlarge | r5ad.2xlarge | r5ad.4xlarge | Memory

r5ad.8xlarge | r5ad.12xlarge | r5ad.16xlarge | r5ad.24xlarge optimized (p. 272)

R5b r5b.large | r5b.xlarge | r5b.2xlarge | r5b.4xlarge | r5b.8xlarge Memory

| r5b.12xlarge | r5b.16xlarge | r5b.24xlarge | r5b.metal optimized (p. 272)

R5d r5d.large | r5d.xlarge | r5d.2xlarge | r5d.4xlarge | r5d.8xlarge Memory

| r5d.12xlarge | r5d.16xlarge | r5d.24xlarge | r5d.metal optimized (p. 272)

R5dn r5dn.large | r5dn.xlarge | r5dn.2xlarge | r5dn.4xlarge | Memory

r5dn.8xlarge | r5dn.12xlarge | r5dn.16xlarge | r5dn.24xlarge optimized (p. 272)

R5n r5n.large | r5n.xlarge | r5n.2xlarge | r5n.4xlarge | r5n.8xlarge Memory

| r5n.12xlarge | r5n.16xlarge | r5n.24xlarge optimized (p. 272)

R6g r6g.medium | r6g.large | r6g.xlarge | r6g.2xlarge | r6g.4xlarge | Memory

r6g.8xlarge | r6g.12xlarge | r6g.16xlarge | r6g.metal optimized (p. 272)

R6gd r6gd.medium | r6gd.large | r6gd.xlarge | r6gd.2xlarge | Memory

r6gd.4xlarge | r6gd.8xlarge | r6gd.12xlarge | r6gd.16xlarge | optimized (p. 272)
r6gd.metal

T2 t2.nano | t2.micro | t2.small | t2.medium | t2.large | t2.xlarge | General

t2.2xlarge purpose (p. 213)

T3 t3.nano | t3.micro | t3.small | t3.medium | t3.large | t3.xlarge | General

t3.2xlarge purpose (p. 213)

206
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hardware specifications

Type Sizes Use case

T3a t3a.nano | t3a.micro | t3a.small | t3a.medium | t3a.large | General

t3a.xlarge | t3a.2xlarge purpose (p. 213)

T4g t4g.nano | t4g.micro | t4g.small | t4g.medium | t4g.large | General

t4g.xlarge | t4g.2xlarge purpose (p. 213)

u-xtb1 u-6tb1.metal | u-9tb1.metal | u-12tb1.metal | u-18tb1.metal | Memory

u-24tb1.metal optimized (p. 272)

X1 x1.16xlarge | x1.32xlarge Memory

optimized (p. 272)

X1e x1e.xlarge | x1e.2xlarge | x1e.4xlarge | x1e.8xlarge | Memory

x1e.16xlarge | x1e.32xlarge optimized (p. 272)

z1d z1d.large | z1d.xlarge | z1d.2xlarge | z1d.3xlarge | z1d.6xlarge Memory

| z1d.12xlarge | z1d.metal optimized (p. 272)

Previous generation instances

Amazon Web Services offers previous generation instance types for users who have optimized their
applications around them and have yet to upgrade. We encourage you to use current generation instance
types to get the best performance, but we continue to support the following previous generation
instance types. For more information about which current generation instance type would be a suitable
upgrade, see Previous Generation Instances.

Type Sizes

A1 a1.medium | a1.large | a1.xlarge | a1.2xlarge | a1.4xlarge | a1.metal

C1 c1.medium | c1.xlarge

C3 c3.large | c3.xlarge | c3.2xlarge | c3.4xlarge | c3.8xlarge

G2 g2.2xlarge | g2.8xlarge

I2 i2.xlarge | i2.2xlarge | i2.4xlarge | i2.8xlarge

M1 m1.small | m1.medium | m1.large | m1.xlarge

M2 m2.xlarge | m2.2xlarge | m2.4xlarge

M3 m3.medium | m3.large | m3.xlarge | m3.2xlarge

R3 r3.large | r3.xlarge | r3.2xlarge | r3.4xlarge | r3.8xlarge

T1 t1.micro

Hardware specifications
For more information about the hardware specifications for each Amazon EC2 instance type, see Amazon
EC2 Instance Types.

207
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI virtualization types

To determine which instance type best meets your needs, we recommend that you launch an instance
and use your own benchmark application. Because you pay by the instance second, it's convenient and
inexpensive to test multiple instance types before making a decision.

If your needs change, even after you make a decision, you can resize your instance later. For more
information, see Change the instance type (p. 311).
Note
Amazon EC2 instances typically run on 64-bit virtual Intel processors as specified in the instance
type product pages. For more information about the hardware specifications for each Amazon
EC2 instance type, see Amazon EC2 Instance Types. However, confusion may result from
industry naming conventions for 64-bit CPUs. Chip manufacturer Advanced Micro Devices
(AMD) introduced the first commercially successful 64-bit architecture based on the Intel x86
instruction set. Consequently, the architecture is widely referred to as AMD64 regardless of the
chip manufacturer. Windows and several Linux distributions follow this practice. This explains
why the internal system information on an Ubuntu or Windows EC2 instance displays the CPU
architecture as AMD64 even though the instances are running on Intel hardware.

Processor features
Intel processor features

Amazon EC2 instances that run on Intel processors may include the following features. Not all of the
following processor features are supported by all instance types. For detailed information about which
features are available for each instance type, see Amazon EC2 Instance Types.

• Intel AES New Instructions (AES-NI) — Intel AES-NI encryption instruction set improves upon the
original Advanced Encryption Standard (AES) algorithm to provide faster data protection and greater
security. All current generation EC2 instances support this processor feature.
• Intel Advanced Vector Extensions (Intel AVX, Intel AVX2, and Intel AVX-512) — Intel AVX and Intel
AVX2 are 256-bit, and Intel AVX-512 is a 512-bit instruction set extension designed for applications
that are Floating Point (FP) intensive. Intel AVX instructions improve performance for applications like
image and audio/video processing, scientific simulations, financial analytics, and 3D modeling and
analysis. These features are only available on instances launched with HVM AMIs.
• Intel Turbo Boost Technology — Intel Turbo Boost Technology processors automatically run cores
faster than the base operating frequency.
• Intel Deep Learning Boost (Intel DL Boost) — Accelerates AI deep learning use cases. The 2nd Gen
Intel Xeon Scalable processors extend Intel AVX-512 with a new Vector Neural Network Instruction
(VNNI/INT8) that significantly increases deep learning inference performance over previous generation
Intel Xeon Scalable processors (with FP32) for image recognition/segmentation, object detection,
speech recognition, language translation, recommendation systems, reinforcement learning, and more.
VNNI may not be compatible with all Linux distributions.

The following instances support VNNI: M5n, R5n, M5dn, M5zn, R5b, R5dn, D3, and D3en. C5 and C5d
instances support VNNI for only 12xlarge, 24xlarge, and metal instances.

AMI virtualization types

The virtualization type of your instance is determined by the AMI that you use to launch it. Current
generation instance types support hardware virtual machine (HVM) only. Some previous generation
instance types support paravirtual (PV) and some AWS regions support PV instances. For more
information, see Linux AMI virtualization types (p. 104).

For best performance, we recommend that you use an HVM AMI. In addition, HVM AMIs are required to
take advantage of enhanced networking. HVM virtualization uses hardware-assist technology provided
by the AWS platform. With HVM virtualization, the guest VM runs as if it were on a native hardware
platform, except that it still uses PV network and storage drivers for improved performance.

208
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instances built on the Nitro System

Instances built on the Nitro System

The Nitro System is a collection of AWS-built hardware and software components that enable high
performance, high availability, and high security. For more information, see AWS Nitro System.

The Nitro System provides bare metal capabilities that eliminate virtualization overhead and support
workloads that require full access to host hardware. Bare metal instances are well suited for the
following:

• Workloads that require access to low-level hardware features (for example, Intel VT) that are not
available or fully supported in virtualized environments
• Applications that require a non-virtualized environment for licensing or support

Nitro components
The following components are part of the Nitro System:

• Nitro card
• Local NVMe storage volumes
• Networking hardware support
• Management
• Monitoring
• Security
• Nitro security chip, integrated into the motherboard
• Nitro hypervisor - A lightweight hypervisor that manages memory and CPU allocation and delivers
performance that is indistinguishable from bare metal for most workloads.

Instance types
The following instances are built on the Nitro System:

• Virtualized: A1, C5, C5a, C5ad, C5d, C5n, C6g, C6gd, C6gn, D3, D3en, G4, I3en, Inf1, M5, M5a, M5ad,
M5d, M5dn, M5n, M5zn, M6g, M6gd, p3dn.24xlarge, P4, R5, R5a, R5ad, R5b, R5d, R5dn, R5n, R6g,
R6gd, T3, T3a, T4g, and z1d
• Bare metal: a1.metal, c5.metal, c5d.metal, c5n.metal, c6g.metal, c6gd.metal, i3.metal,
i3en.metal, m5.metal, m5d.metal, m5zn.metal, m6g.metal, m6gd.metal, mac1.metal,
r5.metal, r5b.metal, r5d.metal, r6g.metal, r6gd.metal, u-6tb1.metal, u-9tb1.metal,
u-12tb1.metal, u-18tb1.metal, u-24tb1.metal, and z1d.metal

Learn more
For more information, see the following videos:

• AWS re:Invent 2017: The Amazon EC2 Nitro System Architecture

• AWS re:Invent 2017: Amazon EC2 Bare Metal Instances
• AWS re:Invent 2019: Powering next-gen Amazon EC2: Deep dive into the Nitro system
• AWS re:Inforce 2019: Security Benefits of the Nitro Architecture

Networking and storage features

When you select an instance type, this determines the networking and storage features that are
available. To describe an instance type, use the describe-instance-types command.

209
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Networking and storage features

Networking features

• IPv6 is supported on all current generation instance types and the C3, R3, and I2 previous generation
instance types.
• To maximize the networking and bandwidth performance of your instance type, you can do the
following:
• Launch supported instance types into a cluster placement group to optimize your instances for
high performance computing (HPC) applications. Instances in a common cluster placement group
can benefit from high-bandwidth, low-latency networking. For more information, see Placement
groups (p. 927).
• Enable enhanced networking for supported current generation instance types to get significantly
higher packet per second (PPS) performance, lower network jitter, and lower latencies. For more
information, see Enhanced networking on Linux (p. 864).
• Current generation instance types that are enabled for enhanced networking have the following
networking performance attributes:
• Traffic within the same Region over private IPv4 or IPv6 can support 5 Gbps for single-flow traffic
and up to 25 Gbps for multi-flow traffic (depending on the instance type).
• Traffic to and from Amazon S3 buckets within the same Region over the public IP address space or
through a VPC endpoint can use all available instance aggregate bandwidth.
• The maximum transmission unit (MTU) supported varies across instance types. All Amazon EC2
instance types support standard Ethernet V2 1500 MTU frames. All current generation instances
support 9001 MTU, or jumbo frames, and some previous generation instances support them as well.
For more information, see Network maximum transmission unit (MTU) for your EC2 instance (p. 939).

Storage features

• Some instance types support EBS volumes and instance store volumes, while other instance types
support only EBS volumes. Some instance types that support instance store volumes use solid state
drives (SSD) to deliver very high random I/O performance. Some instance types support NVMe
instance store volumes. Some instance types support NVMe EBS volumes. For more information, see
Amazon EBS and NVMe on Linux instances (p. 1214) and NVMe SSD volumes (p. 1280).
• To obtain additional, dedicated capacity for Amazon EBS I/O, you can launch some instance types as
EBS–optimized instances. Some instance types are EBS–optimized by default. For more information,
see Amazon EBS–optimized instances (p. 1217).

Summary of networking and storage features

The following table summarizes the networking and storage features supported by current generation
instance types.

  EBS only NVMe EBS Instance store Placement Enhanced

group networking

C4 Yes No No Yes Intel 82599 VF

C5 Yes Yes No Yes ENA

C5a Yes Yes No Yes ENA

C5ad No Yes NVMe * Yes ENA

C5d No Yes NVMe * Yes ENA

C5n Yes Yes No Yes ENA

210
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Networking and storage features

  EBS only NVMe EBS Instance store Placement Enhanced

group networking

C6g Yes Yes No Yes ENA

C6gd No Yes NVME * Yes ENA

C6gn Yes Yes No Yes ENA

D2 No No HDD Yes Intel 82599 VF

D3 No Yes NVMe * Yes ENA

D3en No Yes NVMe * Yes ENA

F1 No No NVMe * Yes ENA

G3 Yes No No Yes ENA

G4ad No Yes NVMe * Yes ENA

G4dn No Yes NVMe * Yes ENA

H1 No No HDD * Yes ENA

I3 No No NVMe * Yes ENA

I3en No Yes NVMe * Yes ENA

Inf1 Yes Yes No Yes ENA

M4 Yes No No Yes m4.16xlarge:

ENA

All other sizes:

Intel 82599 VF

M5 Yes Yes No Yes ENA

M5a Yes Yes No Yes ENA

M5ad No Yes NVMe * Yes ENA

M5d No Yes NVMe * Yes ENA

M5dn No Yes NVMe * Yes ENA

M5n Yes Yes No Yes ENA

M5zn Yes Yes No Yes ENA

M6g Yes Yes No Yes ENA

M6gd No Yes NVME * Yes ENA

Mac1 Yes Yes No No ENA

P2 Yes No No Yes ENA

P3 Yes No No Yes ENA

P3dn No Yes NVMe * Yes ENA

211
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Networking and storage features

  EBS only NVMe EBS Instance store Placement Enhanced

group networking

P4d No Yes NVMe * Yes ENA

R4 Yes No No Yes ENA

R5 Yes Yes No Yes ENA

R5a Yes Yes No Yes ENA

R5ad No Yes NVMe * Yes ENA

R5b Yes Yes No Yes ENA

R5d No Yes NVMe * Yes ENA

R5dn No Yes NVMe * Yes ENA

R5n Yes Yes No Yes ENA

R6g Yes Yes No Yes ENA

R6gd No Yes NVME * Yes ENA

T2 Yes No No No No

T3 Yes Yes No No ENA

T3a Yes Yes No No ENA

T4g Yes Yes No No ENA

u-xtb1.metal Yes Yes No No ENA

X1 No No SSD * Yes ENA

X1e No No SSD * Yes ENA

z1d No Yes NVMe * Yes ENA

* The root device volume must be an Amazon EBS volume.

The following table summarizes the networking and storage features supported by previous generation
instance types.

  Instance store Placement group Enhanced networking

C3 SSD Yes Intel 82599 VF

G2 SSD Yes No

I2 SSD Yes Intel 82599 VF

M3 SSD No No

R3 SSD Yes Intel 82599 VF

212
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance limits

Instance limits
There is a limit on the total number of instances that you can launch in a region, and there are additional
limits on some instance types.

For more information about the default limits, see How many instances can I run in Amazon EC2?

For more information about viewing your current limits or requesting an increase in your current limits,
see Amazon EC2 service quotas (p. 1323).

General purpose instances

General purpose instances provide a balance of compute, memory, and networking resources, and can be
used for a wide range of workloads.

M5 and M5a instances

These instances provide an ideal cloud infrastructure, offering a balance of compute, memory, and
networking resources for a broad range of applications that are deployed in the cloud. They are well-
suited for the following:

• Small and midsize databases

• Data processing tasks that require additional memory
• Caching fleets
• Backend servers for SAP, Microsoft SharePoint, cluster computing, and other enterprise applications

For more information, see Amazon EC2 M5 Instances.

Bare metal instances, such as m5.metal, provide your applications with direct access to physical
resources of the host server, such as processors and memory.

M5zn
These instances are ideal for applications that benefit from extremely high single-thread performance,
high throughput, and low latency networking. They are well-suited for the following:

• Gaming
• High performance computing
• Simulation modeling

For more information, see Amazon EC2 M5 Instances.

Bare metal instances, such as m5zn.metal, provide your applications with direct access to physical
resources of the host server, such as processors and memory.

M6g and M6gd instances

These instances are powered by AWS Graviton2 processors and deliver balanced compute, memory, and
networking for a broad range a general purpose workloads. They are well suited for the following:

• Application servers
• Microservices
• Gaming servers
• Midsize data stores
• Caching fleets

213
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Bare metal instances, such as m6g.metal, provide your applications with direct access to physical
resources of the host server, such as processors and memory.

For more information, see Amazon EC2 M6g Instances.

Mac1 instances

These instances are powered by Apple Mac mini computers. They provide up to 10 Gbps of network
bandwidth and 8 Gbps EBS bandwidth through high-speed Thunderbolt 3 connections. They are well
suited to develop, build, test, and sign applications for Apple devices, such as iPhone, iPad, iPod, Mac,
Apple Watch, and Apple TV.

For more information, see Amazon EC2 Mac instances (p. 258).

T2, T3, T3a, and T4g instances

These instances provide a baseline level of CPU performance with the ability to burst to a higher level
when required by your workload. An Unlimited instance can sustain high CPU performance for any
period of time whenever required. For more information, see Burstable performance instances (p. 224).
They are well-suited for the following:

• Websites and web applications

• Code repositories
• Development, build, test, and staging environments
• Microservices

For more information, see Amazon EC2 T2 Instances, Amazon EC2 T3 Instances, and Amazon EC2 T4g
Instances.

Contents
• Hardware specifications (p. 214)
• Instance performance (p. 218)
• Network performance (p. 218)
• SSD I/O performance (p. 220)
• Instance features (p. 222)
• Release notes (p. 222)
• Burstable performance instances (p. 224)
• Amazon EC2 Mac instances (p. 258)

Hardware specifications
The following is a summary of the hardware specifications for general purpose instances.

Instance type Default vCPUs Memory (GiB)

m4.large 2 8

m4.xlarge 4 16

m4.2xlarge 8 32

m4.4xlarge 16 64

m4.10xlarge 40 160

214
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Instance type Default vCPUs Memory (GiB)

m4.16xlarge 64 256

m5.large 2 8

m5.xlarge 4 16

m5.2xlarge 8 32

m5.4xlarge 16 64

m5.8xlarge 32 128

m5.12xlarge 48 192

m5.16xlarge 64 256

m5.24xlarge 96 384

m5.metal 96 384

m5a.large 2 8

m5a.xlarge 4 16

m5a.2xlarge 8 32

m5a.4xlarge 16 64

m5a.8xlarge 32 128

m5a.12xlarge 48 192

m5a.16xlarge 64 256

m5a.24xlarge 96 384

m5ad.large 2 8

m5ad.xlarge 4 16

m5ad.2xlarge 8 32

m5ad.4xlarge 16 64

m5ad.8xlarge 32 128

m5ad.12xlarge 48 192

m5ad.16xlarge 64 256

m5ad.24xlarge 96 384

m5d.large 2 8

m5d.xlarge 4 16

m5d.2xlarge 8 32

m5d.4xlarge 16 64

m5d.8xlarge 32 128

215
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Instance type Default vCPUs Memory (GiB)

m5d.12xlarge 48 192

m5d.16xlarge 64 256

m5d.24xlarge 96 384

m5d.metal 96 384

m5dn.large 2 8

m5dn.xlarge 4 16

m5dn.2xlarge 8 32

m5dn.4xlarge 16 64

m5dn.8xlarge 32 128

m5dn.12xlarge 48 192

m5dn.16xlarge 64 256

m5dn.24xlarge 96 384

m5n.large 2 8

m5n.xlarge 4 16

m5n.2xlarge 8 32

m5n.4xlarge 16 64

m5n.8xlarge 32 128

m5n.12xlarge 48 192

m5n.16xlarge 64 256

m5n.24xlarge 96 384

m5zn.large 2 8

m5zn.xlarge 4 16

m5zn.2xlarge 8 32

m5zn.3xlarge 12 48

m5zn.6xlarge 24 96

m5zn.12xlarge 48 192

m5zn.metal 48 192

m6g.medium 1 4

m6g.large 2 8

m6g.xlarge 4 16

m6g.2xlarge 8 32

216
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Instance type Default vCPUs Memory (GiB)

m6g.4xlarge 16 64

m6g.8xlarge 32 128

m6g.12xlarge 48 192

m6g.16xlarge 64 256

m6g.metal 64 256

m6gd.medium 1 4

m6gd.large 2 8

m6gd.xlarge 4 16

m6gd.2xlarge 8 32

m6gd.4xlarge 16 64

m6gd.8xlarge 32 128

m6gd.12xlarge 48 192

m6gd.16xlarge 64 256

m6gd.metal 64 256

mac1.metal 12 32

t2.nano 1 0.5

t2.micro 1 1

t2.small 1 2

t2.medium 2 4

t2.large 2 8

t2.xlarge 4 16

t2.2xlarge 8 32

t3.nano 2 0.5

t3.micro 2 1

t3.small 2 2

t3.medium 2 4

t3.large 2 8

t3.xlarge 4 16

t3.2xlarge 8 32

t3a.nano 2 0.5

t3a.micro 2 1

217
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Instance type Default vCPUs Memory (GiB)

t3a.small 2 2

t3a.medium 2 4

t3a.large 2 8

t3a.xlarge 4 16

t3a.2xlarge 8 32

t4g.nano 2 0.5

t4g.micro 2 1

t4g.small 2 2

t4g.medium 2 4

t4g.large 2 8

t4g.xlarge 4 16

t4g.2xlarge 8 32

For more information about the hardware specifications for each Amazon EC2 instance type, see Amazon
EC2 Instance Types.

For more information about specifying CPU options, see Optimize CPU options (p. 668).

Instance performance
EBS-optimized instances enable you to get consistently high performance for your EBS volumes by
eliminating contention between Amazon EBS I/O and other network traffic from your instance. Some
general purpose instances are EBS-optimized by default at no additional cost. For more information, see
Amazon EBS–optimized instances (p. 1217).

Some general purpose instance types provide the ability to control processor C-states and P-states on
Linux. C-states control the sleep levels that a core can enter when it is inactive, while P-states control the
desired performance (in CPU frequency) from a core. For more information, see Processor state control
for your EC2 instance (p. 658).

Network performance
You can enable enhanced networking on supported instance types to provide lower latencies, lower
network jitter, and higher packet-per-second (PPS) performance. Most applications do not consistently
need a high level of network performance, but can benefit from access to increased bandwidth when
they send or receive data. For more information, see Enhanced networking on Linux (p. 864).

The following is a summary of network performance for general purpose instances that support
enhanced networking.

Instance type Network performance Enhanced networking

t2.nano | t2.micro | Up to 1 Gbps Not supported

t2.small | t2.medium |
t2.large | t2.xlarge |
t2.2xlarge

218
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Instance type Network performance Enhanced networking

t3.nano | t3.micro | Up to 5 Gbps † ENA (p. 865)

t3.small | t3.medium |
t3.large | t3.xlarge |
t3.2xlarge | t3a.nano |
t3a.micro | t3a.small |
t3a.medium | t3a.large |
t3a.xlarge | t3a.2xlarge
| t4g.nano | t4g.micro |
t4g.small | t4g.medium |
t4g.large | t4g.xlarge |
t4g.2xlarge

m4.large Moderate Intel 82599 VF (p. 875)

m4.xlarge | m4.2xlarge | High Intel 82599 VF (p. 875)

m4.4xlarge

m5.4xlarge and smaller | Up to 10 Gbps † ENA (p. 865)

m5a.8xlarge and smaller |
m5ad.8xlarge and smaller
| m5d.4xlarge and smaller
| m6g.4xlarge and smaller |
m6gd.4xlarge and smaller

m4.10xlarge 10 Gbps Intel 82599 VF (p. 875)

m5.8xlarge | m5a.12xlarge 10 Gbps ENA (p. 865)

| m5ad.12xlarge |
m5d.8xlarge | mac1.metal

m5.12xlarge | m5a.16xlarge 12 Gbps ENA (p. 865)

| m5ad.16xlarge |
m5d.12xlarge | m6g.8xlarge
| m6gd.8xlarge

m5.16xlarge | m5a.24xlarge 20 Gbps ENA (p. 865)

| m5ad.24xlarge
| m5d.16xlarge
| m6g.12xlarge |
m6gd.12xlarge

m5dn.4xlarge and smaller Up to 25 Gbps † ENA (p. 865)

| m5n.4xlarge and smaller |
m5zn.3xlarge and smaller

m4.16xlarge | m5.24xlarge 25 Gbps ENA (p. 865)

| m5.metal | m5d.24xlarge |
m5d.metal | m5dn.8xlarge |
m5n.8xlarge | m6g.16xlarge
| m6g.metal | m6gd.16xlarge
| m6gd.metal

m5dn.12xlarge 50 Gbps ENA (p. 865)

| m5n.12xlarge |
m5zn.6xlarge

219
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Instance type Network performance Enhanced networking

m5dn.16xlarge | 75 Gbps ENA (p. 865)

m5n.16xlarge

m5dn.24xlarge 100 Gbps ENA (p. 865)

| m5n.24xlarge |
m5zn.12xlarge | m5zn.metal

† These instances use a network I/O credit mechanism to allocate network bandwidth to instances based
on average bandwidth utilization. They accrue credits when their bandwidth is below their baseline
bandwidth, and can use these credits when they perform network data transfers. For more information,
open a support case and ask about baseline bandwidth for the specific instance types that you are
interested in.

SSD I/O performance

If you use a Linux AMI with kernel version 4.4 or later and use all the SSD-based instance store volumes
available to your instance, you get the IOPS (4,096 byte block size) performance listed in the following
table (at queue depth saturation). Otherwise, you get lower IOPS performance.

Instance Size 100% Random Read IOPS Write IOPS

m5ad.large * 30,000 15,000

m5ad.xlarge * 59,000 29,000

m5ad.2xlarge * 117,000 57,000

m5ad.4xlarge * 234,000 114,000

m5ad.8xlarge 466,666 233,333

m5ad.12xlarge 700,000 340,000

m5ad.16xlarge 933,333 466,666

m5ad.24xlarge 1,400,000 680,000

m5d.large * 30,000 15,000

m5d.xlarge * 59,000 29,000

m5d.2xlarge * 117,000 57,000

m5d.4xlarge * 234,000 114,000

m5d.8xlarge 466,666 233,333

m5d.12xlarge 700,000 340,000

m5d.16xlarge 933,333 466,666

m5d.24xlarge 1,400,000 680,000

m5d.metal 1,400,000 680,000

m5dn.large * 30,000 15,000

220
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Instance Size 100% Random Read IOPS Write IOPS

m5dn.xlarge * 59,000 29,000

m5dn.2xlarge * 117,000 57,000

m5dn.4xlarge * 234,000 114,000

m5dn.8xlarge 466,666 233,333

m5dn.12xlarge 700,000 340,000

m5dn.16xlarge 933,333 466,666

m5dn.24xlarge 1,400,000 680,000

m6gd.medium 13,438 5,625

m6gd.large 26,875 11,250

m6gd.xlarge 53,750 22,500

m6gd.2xlarge 107,500 45,000

m6gd.4xlarge 215,000 90,000

m6gd.8xlarge 430,000 180,000

m6gd.12xlarge 645,000 270,000

m6gd.16xlarge 860,000 360,000

m6gd.metal 860,000 360,000

* For these instances, you can get up to the specified performance.

As you fill the SSD-based instance store volumes for your instance, the number of write IOPS that
you can achieve decreases. This is due to the extra work the SSD controller must do to find available
space, rewrite existing data, and erase unused space so that it can be rewritten. This process of
garbage collection results in internal write amplification to the SSD, expressed as the ratio of SSD write
operations to user write operations. This decrease in performance is even larger if the write operations
are not in multiples of 4,096 bytes or not aligned to a 4,096-byte boundary. If you write a smaller
amount of bytes or bytes that are not aligned, the SSD controller must read the surrounding data and
store the result in a new location. This pattern results in significantly increased write amplification,
increased latency, and dramatically reduced I/O performance.

SSD controllers can use several strategies to reduce the impact of write amplification. One such strategy
is to reserve space in the SSD instance storage so that the controller can more efficiently manage the
space available for write operations. This is called over-provisioning. The SSD-based instance store
volumes provided to an instance don't have any space reserved for over-provisioning. To reduce write
amplification, we recommend that you leave 10% of the volume unpartitioned so that the SSD controller
can use it for over-provisioning. This decreases the storage that you can use, but increases performance
even if the disk is close to full capacity.

For instance store volumes that support TRIM, you can use the TRIM command to notify the SSD
controller whenever you no longer need data that you've written. This provides the controller with more
free space, which can reduce write amplification and increase performance. For more information, see
Instance store volume TRIM support (p. 1282).

221
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Instance features
The following is a summary of features for general purpose instances:

  EBS only NVMe EBS Instance store Placement group

M4 Yes No No Yes

M5 Yes Yes No Yes

M5a Yes Yes No Yes

M5ad No Yes NVMe * Yes

M5d No Yes NVMe * Yes

M5dn No Yes NVMe * Yes

M5n Yes Yes No Yes

M5zn Yes Yes No Yes

M6g Yes Yes No Yes

M6gd No Yes NVMe * Yes

Mac1 Yes Yes No No

T2 Yes No No No

T3 Yes Yes No No

T3a Yes Yes No No

T4g Yes Yes No No

* The root device volume must be an Amazon EBS volume.

For more information, see the following:

• Amazon EBS and NVMe on Linux instances (p. 1214)

• Amazon EC2 instance store (p. 1268)
• Placement groups (p. 927)

Release notes
• M5, M5d, and T3 instances feature a 3.1 GHz Intel Xeon Platinum 8000 series processor from either the
first generation (Skylake-SP) or second generation (Cascade Lake).
• M5a, M5ad, and T3a instances feature a 2.5 GHz AMD EPYC 7000 series processor.
• M5zn instances are powered by Intel Cascade Lake CPUs that deliver all-core turbo frequency of up to
4.5 GHz and up to 100 Gbps network bandwidth.
• M6g and M6gd instances feature an AWS Graviton2 processor based on 64-bit Arm architecture.
• Mac1 instances feature a 3.2 GHz Intel eighth-generation (Coffee Lake) Core i7 processor.
• T4g instances feature an AWS Graviton2 processor based on 64-bit Arm architecture.
• M4, M5, M5a, M5ad, M5d, t2.large and larger, and t3.large and larger, and t3a.large and larger
instance types require 64-bit HVM AMIs. They have high-memory, and require a 64-bit operating

222
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

system to take advantage of that capacity. HVM AMIs provide superior performance in comparison to
paravirtual (PV) AMIs on high-memory instance types. In addition, you must use an HVM AMI to take
advantage of enhanced networking.
• Instances built on the Nitro System (p. 209) have the following requirements:
• NVMe drivers (p. 1214) must be installed
• Elastic Network Adapter (ENA) drivers (p. 865) must be installed

The following Linux AMIs meet these requirements:

• Amazon Linux 2
• Amazon Linux AMI 2018.03
• Ubuntu 14.04 (with linux-aws kernel) or later
• Red Hat Enterprise Linux 7.4 or later
• SUSE Linux Enterprise Server 12 SP2 or later
• CentOS 7.4.1708 or later
• FreeBSD 11.1 or later
• Debian GNU/Linux 9 or later
• Instances with an AWS Graviton Processor have the following requirements:
• Use an AMI for the 64-bit Arm architecture.
• Support booting through UEFI with ACPI tables and support ACPI hot-plug of PCI devices.

The following Linux AMIs meet these requirements:

• Amazon Linux 2 (64-bit Arm)
• Ubuntu 16.04 or later (64-bit Arm)
• Red Hat Enterprise Linux 8.0 or later (64-bit Arm)
• SUSE Linux Enterprise Server 15 or later (64-bit Arm)
• Debian 10 or later (64-bit Arm)
• Amazon EC2 Mac instances support macOS Mojave (version 10.14) and macOS Catalina (version
10.15).
• Instances built on the Nitro System support a maximum of 28 attachments, including network
interfaces, EBS volumes, and NVMe instance store volumes. For more information, see Nitro System
volume limits (p. 1291).
• Launching a bare metal instance boots the underlying server, which includes verifying all hardware and
firmware components. This means that it can take 20 minutes from the time the instance enters the
running state until it becomes available over the network.
• To attach or detach EBS volumes or secondary network interfaces from a bare metal instance requires
PCIe native hotplug support. Amazon Linux 2 and the latest versions of the Amazon Linux AMI
support PCIe native hotplug, but earlier versions do not. You must enable the following Linux kernel
configuration options:

CONFIG_HOTPLUG_PCI_PCIE=y
CONFIG_PCIEASPM=y

• Bare metal instances use a PCI-based serial device rather than an I/O port-based serial device. The
upstream Linux kernel and the latest Amazon Linux AMIs support this device. Bare metal instances also
provide an ACPI SPCR table to enable the system to automatically use the PCI-based serial device. The
latest Windows AMIs automatically use the PCI-based serial device.
• Instances built on the Nitro System should have system-logind or acpid installed to support clean
shutdown through API requests.
• There is a limit on the total number of instances that you can launch in a Region, and there are
additional limits on some instance types. For more information, see How many instances can I run in
Amazon EC2? in the Amazon EC2 FAQ.

223
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Burstable performance instances

Burstable performance instances are designed to provide a baseline level of CPU performance with
the ability to burst to a higher level when required by your workload. Burstable performance instances
are well suited for a wide range of general-purpose applications. Examples include microservices, low-
latency interactive applications, small and medium databases, virtual desktops, development, build, and
stage environments, code repositories, and product prototypes.

Burstable performance instances are the only instance types that use credits for CPU usage. For more
information about instance pricing and additional hardware details, see Amazon EC2 Pricing and Amazon
EC2 Instance Types.

If your account is less than 12 months old, you can use a t2.micro instance for free (or a t3.micro
instance in Regions where t2.micro is unavailable) within certain usage limits. For more information,
see AWS Free Tier.

Contents
• Burstable performance instance requirements (p. 224)
• Best practices (p. 224)
• CPU credits and baseline utilization for burstable performance instances (p. 225)
• Unlimited mode for burstable performance instances (p. 228)
• Standard mode for burstable performance instances (p. 236)
• Work with burstable performance instances (p. 250)
• Monitor your CPU credits (p. 255)

Burstable performance instance requirements

The following are the requirements for these instances:

• The supported instance families are: T2, T3, T3a, and T4g.
• The supported purchasing options are: On-Demand Instances, Reserved Instances, Dedicated
Instances, and Spot Instances. These instances are not supported on a Dedicated Host. For more
information, see Instance purchasing options (p. 320).
• Ensure that the instance size you choose passes the minimum memory requirements of your operating
system and applications. Operating systems with graphical user interfaces that consume significant
memory and CPU resources (for example, Windows) might require a t2.micro or larger instance size
for many use cases. As the memory and CPU requirements of your workload grow over time, you can
scale to larger instance sizes of the same instance type, or another instance type.
• For additional requirements, see General Purpose Instances Release Notes (p. 222).

Best practices
Follow these best practices to get the maximum benefit from burstable performance instances.

• Use a recommended AMI – Use an AMI that provides the required drivers. For more information, see
Release notes (p. 222).
• Turn on instance recovery – Create a CloudWatch alarm that monitors an EC2 instance and
automatically recovers it if it becomes impaired for any reason. For more information, see Add recover
actions to Amazon CloudWatch alarms (p. 789).

224
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

CPU credits and baseline utilization for burstable performance instances

Traditional Amazon EC2 instance types provide fixed CPU utilization, while burstable performance
instances provide a baseline level of CPU utilization with the ability to burst CPU utilization above the
baseline level. The baseline utilization and ability to burst are governed by CPU credits.

The CPU credits used depends on CPU utilization. The following scenarios all use one CPU credit:

• One vCPU at 100% utilization for one minute

• One vCPU at 50% utilization for two minutes
• Two vCPUs at 25% utilization for two minutes

Contents
• Earn CPU credits (p. 225)
• CPU credit earn rate (p. 227)
• CPU credit accrual limit (p. 227)
• Accrued CPU credits life span (p. 227)
• Baseline utilization (p. 228)

Earn CPU credits

Each burstable performance instance continuously earns (at a millisecond-level resolution) a set rate
of CPU credits per hour, depending on the instance size. The accounting process for whether credits
are accrued or spent also happens at a millisecond-level resolution, so you don't have to worry about
overspending CPU credits; a short burst of CPU uses a small fraction of a CPU credit.

If a burstable performance instance uses fewer CPU resources than is required for baseline utilization
(such as when it is idle), the unspent CPU credits are accrued in the CPU credit balance. If a burstable
performance instance needs to burst above the baseline utilization level, it spends the accrued credits.
The more credits that a burstable performance instance has accrued, the more time it can burst beyond
its baseline when more CPU utilization is needed.

The following table lists the burstable performance instance types, the rate at which CPU credits are
earned per hour, the maximum number of earned CPU credits that an instance can accrue, the number of
vCPUs per instance, and the baseline utilization as a percentage of a full core (using a single vCPU).

Instance type CPU credits Maximum earned vCPUs Baseline utilization

earned per hour credits that can per vCPU
be accrued*

T2

t2.nano 3 72 1 5%

t2.micro 6 144 1 10%

t2.small 12 288 1 20%

t2.medium 24 576 2 20%**

t2.large 36 864 2 30%**

t2.xlarge 54 1296 4 22.5%**

t2.2xlarge 81.6 1958.4 8 17%**

225
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Instance type CPU credits Maximum earned vCPUs Baseline utilization

earned per hour credits that can per vCPU
be accrued*

T3

t3.nano 6 144 2 5%**

t3.micro 12 288 2 10%**

t3.small 24 576 2 20%**

t3.medium 24 576 2 20%**

t3.large 36 864 2 30%**

t3.xlarge 96 2304 4 40%**

t3.2xlarge 192 4608 8 40%**

T3a

t3a.nano 6 144 2 5%**

t3a.micro 12 288 2 10%**

t3a.small 24 576 2 20%**

t3a.medium 24 576 2 20%**

t3a.large 36 864 2 30%**

t3a.xlarge 96 2304 4 40%**

t3a.2xlarge 192 4608 8 40%**

T4g        

t4g.nano 6 144 2 5%**

t4g.micro 12 288 2 10%**

t4g.small 24 576 2 20%**

t4g.medium 24 576 2 20%**

t4g.large 36 864 2 30%**

t4g.xlarge 96 2304 4 40%**

t4g.2xlarge 192 4608 8 40%**

* The number of credits that can be accrued is equivalent to the number of credits that can be earned
in a 24-hour period.

** The percentage baseline utilization in the table is per vCPU. In CloudWatch, CPU utilization is shown
per vCPU. For example, the CPU utilization for a t3.large instance operating at the baseline level
is shown as 30% in CloudWatch CPU metrics. For information about how to calculate the baseline
utilization, see Baseline utilization (p. 228).

226
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

CPU credit earn rate

The number of CPU credits earned per hour is determined by the instance size. For example, a t3.nano
earns six credits per hour, while a t3.small earns 24 credits per hour. The preceding table lists the
credit earn rate for all instances.

CPU credit accrual limit

While earned credits never expire on a running instance, there is a limit to the number of earned credits
that an instance can accrue. The limit is determined by the CPU credit balance limit. After the limit is
reached, any new credits that are earned are discarded, as indicated by the following image. The full
bucket indicates the CPU credit balance limit, and the spillover indicates the newly earned credits that
exceed the limit.

The CPU credit balance limit differs for each instance size. For example, a t3.micro instance can accrue
a maximum of 288 earned CPU credits in the CPU credit balance. The preceding table lists the maximum
number of earned credits that each instance can accrue.

T2 Standard instances also earn launch credits. Launch credits do not count towards the CPU credit
balance limit. If a T2 instance has not spent its launch credits, and remains idle over a 24-hour period
while accruing earned credits, its CPU credit balance appears as over the limit. For more information, see
Launch credits (p. 237).

T3 and T4g instances do not earn launch credits. These instances launch as unlimited by default, and
therefore can burst immediately upon start without any launch credits.

Accrued CPU credits life span

CPU credits on a running instance do not expire.

For T2, the CPU credit balance does not persist between instance stops and starts. If you stop a T2
instance, the instance loses all its accrued credits.

For T3 and T4g, the CPU credit balance persists for seven days after an instance stops and the credits are
lost thereafter. If you start the instance within seven days, no credits are lost.

For more information, see CPUCreditBalance in the CloudWatch metrics table (p. 255).

227
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Baseline utilization

The baseline utilization is the level at which the CPU can be utilized for a net credit balance of zero, when
the number CPU credits being earned matches the number of CPU credits being used. Baseline utilization
is also known as the baseline.

Baseline utilization is expressed as a percentage of vCPU utilization, which is calculated as follows:

(number of credits earned/number of vCPUs)/60 minutes = % baseline utilization

For example, a t3.nano instance, with 2 vCPUs, earns 6 credits per hour, resulting in a baseline
utilization of 5% , which is calculated as follows:

(6 credits earned/2 vCPUs)/60 minutes = 5% baseline utilization

A t3.xlarge instance, with 4 vCPUs, earns 96 credits per hour, resulting in a baseline utilization of 40%
((96/4)/60).

Unlimited mode for burstable performance instances

A burstable performance instance configured as unlimited can sustain high CPU utilization for any
period of time whenever required. The hourly instance price automatically covers all CPU usage spikes
if the average CPU utilization of the instance is at or below the baseline over a rolling 24-hour period or
the instance lifetime, whichever is shorter.

For the vast majority of general-purpose workloads, instances configured as unlimited provide
ample performance without any additional charges. If the instance runs at higher CPU utilization for a
prolonged period, it can do so for a flat additional rate per vCPU-hour. For information about instance
pricing, see Amazon EC2 Pricing and the section for Unlimited Mode Pricing on the Amazon EC2 On-
Demand Pricing page.

If you use a t2.micro or t3.micro instance under the AWS Free Tier offer and use it in unlimited
mode, charges might apply if your average utilization over a rolling 24-hour period exceeds the baseline
utilization (p. 228) of the instance.

T3 and T4g instances launch as unlimited by default. If the average CPU usage over a 24-hour
period exceeds the baseline, you incur charges for surplus credits. If you launch Spot Instances as
unlimited and plan to use them immediately and for a short duration, with no idle time for accruing
CPU credits, you incur charges for surplus credits. We recommend that you launch your Spot Instances
in standard (p. 236) mode to avoid paying higher costs. For more information, see Surplus credits can
incur charges (p. 231) and Burstable performance instances (p. 465).

Contents
• Unlimited mode concepts (p. 229)
• How Unlimited burstable performance instances work (p. 229)
• When to use unlimited mode versus fixed CPU (p. 229)
• Surplus credits can incur charges (p. 231)
• No launch credits for T2 Unlimited instances (p. 232)
• Enable unlimited mode (p. 232)
• What happens to credits when switching between Unlimited and Standard (p. 232)
• Monitor credit usage (p. 232)
• Unlimited mode examples (p. 233)
• Example 1: Explain credit use with T3 Unlimited (p. 233)
• Example 2: Explain credit use with T2 Unlimited (p. 234)

228
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Unlimited mode concepts

The unlimited mode is a credit configuration option for burstable performance instances. It can be
enabled or disabled at any time for a running or stopped instance. You can set unlimited as the default
credit option at the account level per AWS Region, per burstable performance instance family, so that all
new burstable performance instances in the account launch using the default credit option.

How Unlimited burstable performance instances work

If a burstable performance instance configured as unlimited depletes its CPU credit balance, it can
spend surplus credits to burst beyond the baseline (p. 228). When its CPU utilization falls below
the baseline, it uses the CPU credits that it earns to pay down the surplus credits that it spent earlier.
The ability to earn CPU credits to pay down surplus credits enables Amazon EC2 to average the CPU
utilization of an instance over a 24-hour period. If the average CPU usage over a 24-hour period exceeds
the baseline, the instance is billed for the additional usage at a flat additional rate per vCPU-hour.

The following graph shows the CPU usage of a t3.large. The baseline CPU utilization for a t3.large
is 30%. If the instance runs at 30% CPU utilization or less on average over a 24-hour period, there is
no additional charge because the cost is already covered by the instance hourly price. However, if the
instance runs at 40% CPU utilization on average over a 24-hour period, as shown in the graph, the
instance is billed for the additional 10% CPU usage at a flat additional rate per vCPU-hour.

For more information about the baseline utilization per vCPU for each instance type and how many
credits each instance type earns, see the credit table (p. 225).

When to use unlimited mode versus fixed CPU

When determining whether you should use a burstable performance instance in unlimited mode,
such as T3, or a fixed performance instance, such as M5, you need to determine the breakeven CPU
usage. The breakeven CPU usage for a burstable performance instance is the point at which a burstable
performance instance costs the same as a fixed performance instance. The breakeven CPU usage helps
you determine the following:

• If the average CPU usage over a 24-hour period is at or below the breakeven CPU usage, use a
burstable performance instance in unlimited mode so that you can benefit from the lower price of a
burstable performance instance while getting the same performance as a fixed performance instance.
• If the average CPU usage over a 24-hour period is above the breakeven CPU usage, the burstable
performance instance will cost more than the equivalently-sized fixed performance instance. If a T3
instance continuously bursts at 100% CPU, you end up paying approximately 1.5 times the price of an
equivalently-sized M5 instance.

The following graph shows the breakeven CPU usage point where a t3.large costs the same as an
m5.large. The breakeven CPU usage point for a t3.large is 42.5%. If the average CPU usage is at
42.5%, the cost of running the t3.large is the same as an m5.large, and is more expensive if the
average CPU usage is above 42.5%. If the workload needs less than 42.5% average CPU usage, you can
benefit from the lower price of the t3.large while getting the same performance as an m5.large.

229
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

The following table shows how to calculate the breakeven CPU usage threshold so that you can
determine when it's less expensive to use a burstable performance instance in unlimited mode or a
fixed performance instance. The columns in the table are labeled A through K.

Instance vCPUs T3 M5 Price T3 Charge Charge AdditionalAdditional

Breakeven
type price*/ price*/ differencebaseline per per burst CPU % CPU %
hour hour utilization vCPU vCPU minutes available
per hour minute available
vCPU for per
(%) surplus vCPU
credits

A B C D E= F G H= I= J = (I / K=
D-C G / 60 E/H 60) / B F+J

t3.large 2 $0.0835 $0.096 $0.0125 30% $0.05 $0.000833 15 12.5% 42.5%

230
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

* Price is based on us-east-1 and Linux OS.

The table provides the following information:

• Column A shows the instance type, t3.large.

• Column B shows the number of vCPUs for the t3.large.
• Column C shows the price of a t3.large per hour.
• Column D shows the price of an m5.large per hour.
• Column E shows the price difference between the t3.large and the m5.large.
• Column F shows the baseline utilization per vCPU of the t3.large, which is 30%. At the baseline, the
hourly cost of the instance covers the cost of the CPU usage.
• Column G shows the flat additional rate per vCPU-hour that an instance is charged if it bursts at 100%
CPU after it has depleted its earned credits.
• Column H shows the flat additional rate per vCPU-minute that an instance is charged if it bursts at
100% CPU after it has depleted its earned credits.
• Column I shows the number of additional minutes that the t3.large can burst per hour at 100% CPU
while paying the same price per hour as an m5.large.
• Column J shows the additional CPU usage (in %) over baseline that the instance can burst while paying
the same price per hour as an m5.large.
• Column K shows the breakeven CPU usage (in %) that the t3.large can burst without paying more
than the m5.large. Anything above this, and the t3.large costs more than the m5.large.

The following table shows the breakeven CPU usage (in %) for T3 instance types compared to the
similarly-sized M5 instance types.

T3 instance type Breakeven CPU usage (in %) for T3 compared to

M5

t3.large 42.5%

t3.xlarge 52.5%

t3.2xlarge 52.5%

Surplus credits can incur charges

If the average CPU utilization of an instance is at or below the baseline, the instance incurs no additional
charges. Because an instance earns a maximum number of credits (p. 225) in a 24-hour period (for
example, a t3.micro instance can earn a maximum of 288 credits in a 24-hour period), it can spend
surplus credits up to that maximum without being charged.

However, if CPU utilization stays above the baseline, the instance cannot earn enough credits to pay
down the surplus credits that it has spent. The surplus credits that are not paid down are charged at a
flat additional rate per vCPU-hour.

Surplus credits that were spent earlier are charged when any of the following occurs:

• The spent surplus credits exceed the maximum number of credits (p. 225) the instance can earn in a
24-hour period. Spent surplus credits above the maximum are charged at the end of the hour.
• The instance is stopped or terminated.

231
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

• The instance is switched from unlimited to standard.

Spent surplus credits are tracked by the CloudWatch metric CPUSurplusCreditBalance. Surplus
credits that are charged are tracked by the CloudWatch metric CPUSurplusCreditsCharged. For more
information, see Additional CloudWatch metrics for burstable performance instances (p. 255).

No launch credits for T2 Unlimited instances

T2 Standard instances receive launch credits (p. 237), but T2 Unlimited instances do not. A T2
Unlimited instance can burst beyond the baseline at any time with no additional charge, as long as
its average CPU utilization is at or below the baseline over a rolling 24-hour window or its lifetime,
whichever is shorter. As such, T2 Unlimited instances do not require launch credits to achieve high
performance immediately after launch.

If a T2 instance is switched from standard to unlimited, any accrued launch credits are removed from
the CPUCreditBalance before the remaining CPUCreditBalance is carried over.

T3 and T4g instances never receive launch credits.

Enable unlimited mode

You can switch from unlimited to standard, and from standard to unlimited, at any time on
a running or stopped instance. For more information, see Launch a burstable performance instance
as Unlimited or Standard (p. 250) and Modify the credit specification of a burstable performance
instance (p. 253).

You can set unlimited as the default credit option at the account level per AWS Region, per burstable
performance instance family, so that all new burstable performance instances in the account launch
using the default credit option. For more information, see Set the default credit specification for the
account (p. 254).

You can check whether your burstable performance instance is configured as unlimited or standard
using the Amazon EC2 console or the AWS CLI. For more information, see View the credit specification of
a burstable performance instance (p. 252) and View the default credit specification (p. 254).

What happens to credits when switching between Unlimited and Standard

CPUCreditBalance is a CloudWatch metric that tracks the number of credits accrued by an instance.
CPUSurplusCreditBalance is a CloudWatch metric that tracks the number of surplus credits spent by
an instance.

When you change an instance configured as unlimited to standard, the following occurs:

• The CPUCreditBalance value remains unchanged and is carried over.

• The CPUSurplusCreditBalance value is immediately charged.

When a standard instance is switched to unlimited, the following occurs:

• The CPUCreditBalance value containing accrued earned credits is carried over.

• For T2 Standard instances, any launch credits are removed from the CPUCreditBalance value, and
the remaining CPUCreditBalance value containing accrued earned credits is carried over.

Monitor credit usage

To see if your instance is spending more credits than the baseline provides, you can use CloudWatch
metrics to track usage, and you can set up hourly alarms to be notified of credit usage. For more
information, see Monitor your CPU credits (p. 255).

232
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Unlimited mode examples

The following examples explain credit use for instances that are configured as unlimited.

Examples
• Example 1: Explain credit use with T3 Unlimited (p. 233)
• Example 2: Explain credit use with T2 Unlimited (p. 234)

Example 1: Explain credit use with T3 Unlimited

In this example, you see the CPU utilization of a t3.nano instance launched as unlimited, and how it
spends earned and surplus credits to sustain CPU utilization.

A t3.nano instance earns 144 CPU credits over a rolling 24-hour period, which it can redeem for 144
minutes of vCPU use. When it depletes its CPU credit balance (represented by the CloudWatch metric
CPUCreditBalance), it can spend surplus CPU credits—that it has not yet earned—to burst for as long
as it needs. Because a t3.nano instance earns a maximum of 144 credits in a 24-hour period, it can
spend surplus credits up to that maximum without being charged immediately. If it spends more than
144 CPU credits, it is charged for the difference at the end of the hour.

The intent of the example, illustrated by the following graph, is to show how an instance can burst using
surplus credits even after it depletes its CPUCreditBalance. The following workflow references the
numbered points on the graph:

P1 – At 0 hours on the graph, the instance is launched as unlimited and immediately begins to earn
credits. The instance remains idle from the time it is launched—CPU utilization is 0%—and no credits are
spent. All unspent credits are accrued in the credit balance. For the first 24 hours, CPUCreditUsage is at
0, and the CPUCreditBalance value reaches its maximum of 144.

P2 – For the next 12 hours, CPU utilization is at 2.5%, which is below the 5% baseline. The instance
earns more credits than it spends, but the CPUCreditBalance value cannot exceed its maximum of 144
credits.

P3 – For the next 24 hours, CPU utilization is at 7% (above the baseline), which requires a spend of 57.6
credits. The instance spends more credits than it earns, and the CPUCreditBalance value reduces to
86.4 credits.

P4 – For the next 12 hours, CPU utilization decreases to 2.5% (below the baseline), which requires a
spend of 36 credits. In the same time, the instance earns 72 credits. The instance earns more credits than
it spends, and the CPUCreditBalance value increases to 122 credits.

P5 – For the next 5 hours, the instance bursts at 100% CPU utilization, and spends a total of 570 credits
to sustain the burst. About an hour into this period, the instance depletes its entire CPUCreditBalance
of 122 credits, and starts to spend surplus credits to sustain the high CPU utilization, totaling 448
surplus credits in this period (570-122=448). When the CPUSurplusCreditBalance value reaches
144 CPU credits (the maximum a t3.nano instance can earn in a 24-hour period), any surplus credits
spent thereafter cannot be offset by earned credits. The surplus credits spent thereafter amounts to 304
credits (448-144=304), which results in a small additional charge at the end of the hour for 304 credits.

P6 – For the next 13 hours, CPU utilization is at 5% (the baseline). The instance earns as
many credits as it spends, with no excess to pay down the CPUSurplusCreditBalance. The
CPUSurplusCreditBalance value remains at 144 credits.

P7 – For the last 24 hours in this example, the instance is idle and CPU utilization is 0%. During this time,
the instance earns 144 credits, which it uses to pay down the CPUSurplusCreditBalance.

233
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Example 2: Explain credit use with T2 Unlimited

In this example, you see the CPU utilization of a t2.nano instance launched as unlimited, and how it
spends earned and surplus credits to sustain CPU utilization.

A t2.nano instance earns 72 CPU credits over a rolling 24-hour period, which it can redeem for 72
minutes of vCPU use. When it depletes its CPU credit balance (represented by the CloudWatch metric
CPUCreditBalance), it can spend surplus CPU credits—that it has not yet earned—to burst for as long
as it needs. Because a t2.nano instance earns a maximum of 72 credits in a 24-hour period, it can spend
surplus credits up to that maximum without being charged immediately. If it spends more than 72 CPU
credits, it is charged for the difference at the end of the hour.

The intent of the example, illustrated by the following graph, is to show how an instance can burst using
surplus credits even after it depletes its CPUCreditBalance. You can assume that, at the start of the
time line in the graph, the instance has an accrued credit balance equal to the maximum number of
credits it can earn in 24 hours. The following workflow references the numbered points on the graph:

1 – In the first 10 minutes, CPUCreditUsage is at 0, and the CPUCreditBalance value remains at its
maximum of 72.

2 – At 23:40, as CPU utilization increases, the instance spends CPU credits and the CPUCreditBalance
value decreases.

3 – At around 00:47, the instance depletes its entire CPUCreditBalance, and starts to spend surplus
credits to sustain high CPU utilization.

4 – Surplus credits are spent until 01:55, when the CPUSurplusCreditBalance value reaches 72 CPU
credits. This is equal to the maximum a t2.nano instance can earn in a 24-hour period. Any surplus

234
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

credits spent thereafter cannot be offset by earned credits within the 24-hour period, which results in a
small additional charge at the end of the hour.

5 – The instance continues to spend surplus credits until around 02:20. At this time, CPU utilization
falls below the baseline, and the instance starts to earn credits at 3 credits per hour (or 0.25
credits every 5 minutes), which it uses to pay down the CPUSurplusCreditBalance. After the
CPUSurplusCreditBalance value reduces to 0, the instance starts to accrue earned credits in its
CPUCreditBalance at 0.25 credits every 5 minutes.

Calculating the bill

Surplus credits cost $0.05 per vCPU-hour. The instance spent approximately 25 surplus credits between
01:55 and 02:20, which is equivalent to 0.42 vCPU-hours.

Additional charges for this instance are 0.42 vCPU-hours x $0.05/vCPU-hour = $0.021, rounded to $0.02.

Here is the month-end bill for this T2 Unlimited instance:

235
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

You can set billing alerts to be notified every hour of any accruing charges, and take action if required.

Standard mode for burstable performance instances

A burstable performance instance configured as standard is suited to workloads with an average
CPU utilization that is consistently below the baseline CPU utilization of the instance. To burst above
the baseline, the instance spends credits that it has accrued in its CPU credit balance. If the instance
is running low on accrued credits, CPU utilization is gradually lowered to the baseline level, so that
the instance does not experience a sharp performance drop-off when its accrued CPU credit balance
is depleted. For more information, see CPU credits and baseline utilization for burstable performance
instances (p. 225).

Contents
• Standard mode concepts (p. 236)
• How standard burstable performance instances work (p. 236)
• Launch credits (p. 237)
• Launch credit limits (p. 237)
• Differences between launch credits and earned credits (p. 237)
• Standard mode examples (p. 238)
• Example 1: Explain credit use with T3 Standard (p. 238)
• Example 2: Explain credit use with T2 Standard (p. 240)
• Period 1: 1 – 24 hours (p. 241)
• Period 2: 25 – 36 hours (p. 242)
• Period 3: 37 – 61 hours (p. 243)
• Period 4: 62 – 72 hours (p. 244)
• Period 5: 73 – 75 hours (p. 246)
• Period 6: 76 – 90 hours (p. 247)
• Period 7: 91 – 96 hours (p. 249)

Standard mode concepts

The standard mode is a configuration option for burstable performance instances. It can be enabled
or disabled at any time for a running or stopped instance. You can set standard as the default credit
option at the account level per AWS Region, per burstable performance instance family, so that all new
burstable performance instances in the account launch using the default credit option.

How standard burstable performance instances work

When a burstable performance instance configured as standard is in a running state, it continuously

earns (at a millisecond-level resolution) a set rate of earned credits per hour. For T2 Standard, when
the instance is stopped, it loses all its accrued credits, and its credit balance is reset to zero. When it is
restarted, it receives a new set of launch credits, and begins to accrue earned credits. For T3 and T4g
Standard instances, the CPU credit balance persists for seven days after the instance stops and the
credits are lost thereafter. If you start the instance within seven days, no credits are lost.
236
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

T2 Standard instances receive two types of CPU credits: earned credits and launch credits. When a T2
Standard instance is in a running state, it continuously earns (at a millisecond-level resolution) a set rate
of earned credits per hour. At start, it has not yet earned credits for a good startup experience; therefore,
to provide a good startup experience, it receives launch credits at start, which it spends first while it
accrues earned credits.

T3 and T4g Standard instances do not receive launch credits.

Launch credits

T2 Standard instances get 30 launch credits per vCPU at launch or start. For example, a t2.micro
instance has one vCPU and gets 30 launch credits, while a t2.xlarge instance has four vCPUs and gets
120 launch credits. Launch credits are designed to provide a good startup experience to allow instances
to burst immediately after launch before they have accrued earned credits.

Launch credits are spent first, before earned credits. Unspent launch credits are accrued in the CPU
credit balance, but do not count towards the CPU credit balance limit. For example, a t2.micro instance
has a CPU credit balance limit of 144 earned credits. If it is launched and remains idle for 24 hours,
its CPU credit balance reaches 174 (30 launch credits + 144 earned credits), which is over the limit.
However, after the instance spends the 30 launch credits, the credit balance cannot exceed 144. For more
information about the CPU credit balance limit for each instance size, see the credit table (p. 225).

The following table lists the initial CPU credit allocation received at launch or start, and the number of
vCPUs.

Instance type Launch credits vCPUs

t1.micro 15 1

t2.nano 30 1

t2.micro 30 1

t2.small 30 1

t2.medium 60 2

t2.large 60 2

t2.xlarge 120 4

t2.2xlarge 240 8

Launch credit limits

There is a limit to the number of times T2 Standard instances can receive launch credits. The default limit
is 100 launches or starts of all T2 Standard instances combined per account, per Region, per rolling 24-
hour period. For example, the limit is reached when one instance is stopped and started 100 times within
a 24-hour period, or when 100 instances are launched within a 24-hour period, or other combinations
that equate to 100 starts. New accounts may have a lower limit, which increases over time based on your
usage.
Tip
To ensure that your workloads always get the performance they need, switch to Unlimited mode
for burstable performance instances (p. 228) or consider using a larger instance size.

Differences between launch credits and earned credits

The following table lists the differences between launch credits and earned credits.

237
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Launch credits Earned credits

Credit earn T2 Standard instances get 30 launch Each T2 instance continuously earns (at
rate credits per vCPU at launch or start. a millisecond-level resolution) a set rate
of CPU credits per hour, depending on
If a T2 instance is switched from the instance size. For more information
unlimited to standard, it does not get about the number of CPU credits
launch credits at the time of switching. earned per instance size, see the credit
table (p. 225).

Credit earn The limit for receiving launch credits is A T2 instance cannot accrue more credits
limit 100 launches or starts of all T2 Standard than the CPU credit balance limit. If the
instances combined per account, per CPU credit balance has reached its limit,
Region, per rolling 24-hour period. New any credits that are earned after the limit
accounts may have a lower limit, which is reached are discarded. Launch credits
increases over time based on your usage. do not count towards the limit. For more
information about the CPU credit balance
limit for each T2 instance size, see the
credit table (p. 225).

Credit use Launch credits are spent first, before Earned credits are spent only after all
earned credits. launch credits are spent.

Credit When a T2 Standard instance is running, When a T2 instance is running, earned

expiration launch credits do not expire. When a T2 credits that have accrued do not expire.
Standard instance stops or is switched to When the T2 instance stops, all accrued
T2 Unlimited, all launch credits are lost. earned credits are lost.

The number of accrued launch credits and accrued earned credits is tracked by the CloudWatch metric
CPUCreditBalance. For more information, see CPUCreditBalance in the CloudWatch metrics
table (p. 255).

Standard mode examples

The following examples explain credit use when instances are configured as standard.

Examples
• Example 1: Explain credit use with T3 Standard (p. 238)
• Example 2: Explain credit use with T2 Standard (p. 240)

Example 1: Explain credit use with T3 Standard

In this example, you see how a t3.nano instance launched as standard earns, accrues, and spends
earned credits. You see how the credit balance reflects the accrued earned credits.

A running t3.nano instance earns 144 credits every 24 hours. Its credit balance limit is 144 earned
credits. After the limit is reached, new credits that are earned are discarded. For more information about
the number of credits that can be earned and accrued, see the credit table (p. 225).

You might launch a T3 Standard instance and use it immediately. Or, you might launch a T3 Standard
instance and leave it idle for a few days before running applications on it. Whether an instance is used or
remains idle determines if credits are spent or accrued. If an instance remains idle for 24 hours from the
time it is launched, the credit balance reaches it limit, which is the maximum number of earned credits
that can be accrued.

238
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

This example describes an instance that remains idle for 24 hours from the time it is launched, and walks
you through seven periods of time over a 96-hour period, showing the rate at which credits are earned,
accrued, spent, and discarded, and the value of the credit balance at the end of each period.

The following workflow references the numbered points on the graph:

P1 – At 0 hours on the graph, the instance is launched as standard and immediately begins to earn
credits. The instance remains idle from the time it is launched—CPU utilization is 0%—and no credits are
spent. All unspent credits are accrued in the credit balance. For the first 24 hours, CPUCreditUsage is at
0, and the CPUCreditBalance value reaches its maximum of 144.

P2 – For the next 12 hours, CPU utilization is at 2.5%, which is below the 5% baseline. The instance
earns more credits than it spends, but the CPUCreditBalance value cannot exceed its maximum of 144
credits. Any credits that are earned in excess of the limit are discarded.

P3 – For the next 24 hours, CPU utilization is at 7% (above the baseline), which requires a spend of 57.6
credits. The instance spends more credits than it earns, and the CPUCreditBalance value reduces to
86.4 credits.

P4 – For the next 12 hours, CPU utilization decreases to 2.5% (below the baseline), which requires a
spend of 36 credits. In the same time, the instance earns 72 credits. The instance earns more credits than
it spends, and the CPUCreditBalance value increases to 122 credits.

P5 – For the next two hours, the instance bursts at 100% CPU utilization, and depletes its entire
CPUCreditBalance value of 122 credits. At the end of this period, with the CPUCreditBalance at
zero, CPU utilization is forced to drop to the baseline utilization level of 5%. At the baseline, the instance
earns as many credits as it spends.

P6 – For the next 14 hours, CPU utilization is at 5% (the baseline). The instance earns as many credits as
it spends. The CPUCreditBalance value remains at 0.

P7 – For the last 24 hours in this example, the instance is idle and CPU utilization is 0%. During this time,
the instance earns 144 credits, which it accrues in its CPUCreditBalance.

239
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Example 2: Explain credit use with T2 Standard

In this example, you see how a t2.nano instance launched as standard earns, accrues, and spends
launch and earned credits. You see how the credit balance reflects not only accrued earned credits, but
also accrued launch credits.

A t2.nano instance gets 30 launch credits when it is launched, and earns 72 credits every 24 hours. Its
credit balance limit is 72 earned credits; launch credits do not count towards the limit. After the limit is
reached, new credits that are earned are discarded. For more information about the number of credits
that can be earned and accrued, see the credit table (p. 225). For more information about limits, see
Launch credit limits (p. 237).

You might launch a T2 Standard instance and use it immediately. Or, you might launch a T2 Standard
instance and leave it idle for a few days before running applications on it. Whether an instance is used
or remains idle determines if credits are spent or accrued. If an instance remains idle for 24 hours from
the time it is launched, the credit balance appears to exceed its limit because the balance reflects both
accrued earned credits and accrued launch credits. However, after CPU is used, the launch credits are
spent first. Thereafter, the limit always reflects the maximum number of earned credits that can be
accrued.

This example describes an instance that remains idle for 24 hours from the time it is launched, and walks
you through seven periods of time over a 96-hour period, showing the rate at which credits are earned,
accrued, spent, and discarded, and the value of the credit balance at the end of each period.

240
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Period 1: 1 – 24 hours

At 0 hours on the graph, the T2 instance is launched as standard and immediately gets 30 launch
credits. It earns credits while in the running state. The instance remains idle from the time it is launched
—CPU utilization is 0%—and no credits are spent. All unspent credits are accrued in the credit balance.
At approximately 14 hours after launch, the credit balance is 72 (30 launch credits + 42 earned credits),
which is equivalent to what the instance can earn in 24 hours. At 24 hours after launch, the credit
balance exceeds 72 credits because the unspent launch credits are accrued in the credit balance—the
credit balance is 102 credits: 30 launch credits + 72 earned credits.

Credit Spend Rate 0 credits per 24 hours (0% CPU utilization)

Credit Earn Rate 72 credits per 24 hours

Credit Discard Rate 0 credits per 24 hours

Credit Balance 102 credits (30 launch credits + 72 earned credits)

241
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Conclusion

If there is no CPU utilization after launch, the instance accrues more credits than what it can earn in 24
hours (30 launch credits + 72 earned credits = 102 credits).

In a real-world scenario, an EC2 instance consumes a small number of credits while launching and
running, which prevents the balance from reaching the maximum theoretical value in this example.

Period 2: 25 – 36 hours
For the next 12 hours, the instance continues to remain idle and earn credits, but the credit balance does
not increase. It plateaus at 102 credits (30 launch credits + 72 earned credits). The credit balance has
reached its limit of 72 accrued earned credits, so newly earned credits are discarded.

Credit Spend Rate 0 credits per 24 hours (0% CPU utilization)

Credit Earn Rate 72 credits per 24 hours (3 credits per hour)

Credit Discard Rate 72 credits per 24 hours (100% of credit earn rate)

242
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Credit Balance 102 credits (30 launch credits + 72 earned credits)

—balance is unchanged

Conclusion

An instance constantly earns credits, but it cannot accrue more earned credits if the credit balance
has reached its limit. After the limit is reached, newly earned credits are discarded. Launch credits do
not count towards the credit balance limit. If the balance includes accrued launch credits, the balance
appears to be over the limit.

Period 3: 37 – 61 hours

For the next 25 hours, the instance uses 2% CPU, which requires 30 credits. In the same period, it earns
75 credits, but the credit balance decreases. The balance decreases because the accrued launch credits
are spent first, while newly earned credits are discarded because the credit balance is already at its limit
of 72 earned credits.

243
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Credit Spend Rate 28.8 credits per 24 hours (1.2 credits per hour,
2% CPU utilization, 40% of credit earn rate)—30
credits over 25 hours

Credit Earn Rate 72 credits per 24 hours

Credit Discard Rate 72 credits per 24 hours (100% of credit earn rate)

Credit Balance 72 credits (30 launch credits were spent; 72

earned credits remain unspent)

Conclusion

An instance spends launch credits first, before spending earned credits. Launch credits do not count
towards the credit limit. After the launch credits are spent, the balance can never go higher than what
can be earned in 24 hours. Furthermore, while an instance is running, it cannot get more launch credits.

Period 4: 62 – 72 hours

For the next 11 hours, the instance uses 2% CPU, which requires 13.2 credits. This is the same CPU
utilization as in the previous period, but the balance does not decrease. It stays at 72 credits.

The balance does not decrease because the credit earn rate is higher than the credit spend rate. In the
time that the instance spends 13.2 credits, it also earns 33 credits. However, the balance limit is 72
credits, so any earned credits that exceed the limit are discarded. The balance plateaus at 72 credits,
which is different from the plateau of 102 credits during Period 2, because there are no accrued launch
credits.

244
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Credit Spend Rate 28.8 credits per 24 hours (1.2 credits per hour, 2%
CPU utilization, 40% of credit earn rate)—13.2
credits over 11 hours

Credit Earn Rate 72 credits per 24 hours

Credit Discard Rate 43.2 credits per 24 hours (60% of credit earn rate)

Credit Balance 72 credits (0 launch credits, 72 earned credits)—

balance is at its limit

Conclusion

After launch credits are spent, the credit balance limit is determined by the number of credits that an
instance can earn in 24 hours. If the instance earns more credits than it spends, newly earned credits over
the limit are discarded.

245
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Period 5: 73 – 75 hours

For the next three hours, the instance bursts at 20% CPU utilization, which requires 36 credits. The
instance earns nine credits in the same three hours, which results in a net balance decrease of 27 credits.
At the end of three hours, the credit balance is 45 accrued earned credits.

Credit Spend Rate 288 credits per 24 hours (12 credits per hour, 20%
CPU utilization, 400% of credit earn rate)—36
credits over 3 hours

Credit Earn Rate 72 credits per 24 hours (9 credits over 3 hours)

Credit Discard Rate 0 credits per 24 hours

Credit Balance 45 credits (previous balance (72) - spent credits

(36) + earned credits (9))—balance decreases at
a rate of 216 credits per 24 hours (spend rate

246
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

288/24 + earn rate 72/24 = balance decrease rate

216/24)

Conclusion

If an instance spends more credits than it earns, its credit balance decreases.

Period 6: 76 – 90 hours

For the next 15 hours, the instance uses 2% CPU, which requires 18 credits. This is the same CPU
utilization as in Periods 3 and 4. However, the balance increases in this period, whereas it decreased in
Period 3 and plateaued in Period 4.

In Period 3, the accrued launch credits were spent, and any earned credits that exceeded the credit limit
were discarded, resulting in a decrease in the credit balance. In Period 4, the instance spent fewer credits
than it earned. Any earned credits that exceeded the limit were discarded, so the balance plateaued at its
maximum of 72 credits.

In this period, there are no accrued launch credits, and the number of accrued earned credits in the
balance is below the limit. No earned credits are discarded. Furthermore, the instance earns more credits
than it spends, resulting in an increase in the credit balance.

247
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Credit Spend Rate 28.8 credits per 24 hours (1.2 credits per hour,
2% CPU utilization, 40% of credit earn rate)—18
credits over 15 hours

Credit Earn Rate 72 credits per 24 hours (45 credits over 15 hours)

Credit Discard Rate 0 credits per 24 hours

Credit Balance 72 credits (balance increases at a rate of 43.2

credits per 24 hours—change rate = spend rate
28.8/24 + earn rate 72/24)

Conclusion

If an instance spends fewer credits than it earns, its credit balance increases.

248
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Period 7: 91 – 96 hours

For the next six hours, the instance remains idle—CPU utilization is 0%—and no credits are spent. This is
the same CPU utilization as in Period 2, but the balance does not plateau at 102 credits—it plateaus at
72 credits, which is the credit balance limit for the instance.

In Period 2, the credit balance included 30 accrued launch credits. The launch credits were spent in
Period 3. A running instance cannot get more launch credits. After its credit balance limit is reached, any
earned credits that exceed the limit are discarded.

Credit Spend Rate 0 credits per 24 hours (0% CPU utilization)

Credit Earn Rate 72 credits per 24 hours

Credit Discard Rate 72 credits per 24 hours (100% of credit earn rate)

Credit Balance 72 credits (0 launch credits, 72 earned credits)

249
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Conclusion

An instance constantly earns credits, but cannot accrue more earned credits if the credit balance limit has
been reached. After the limit is reached, newly earned credits are discarded. The credit balance limit is
determined by the number of credits that an instance can earn in 24 hours. For more information about
credit balance limits, see the credit table (p. 225).

Work with burstable performance instances

The steps for launching, monitoring, and modifying these instances are similar. The key difference is the
default credit specification when they launch. If you do not change the default credit specification, the
default is that:

• T3 and T4g instances launch as unlimited

• T2 instances launch as standard

Contents
• Launch a burstable performance instance as Unlimited or Standard (p. 250)
• Use an Auto Scaling group to launch a burstable performance instance as Unlimited (p. 251)
• View the credit specification of a burstable performance instance (p. 252)
• Modify the credit specification of a burstable performance instance (p. 253)
• Set the default credit specification for the account (p. 254)
• View the default credit specification (p. 254)

Launch a burstable performance instance as Unlimited or Standard

You can launch your instances as unlimited or standard using the Amazon EC2 console, an AWS SDK,
a command line tool, or with an Auto Scaling group. For more information, see Use an Auto Scaling
group to launch a burstable performance instance as Unlimited (p. 251).

Requirements

• You must launch your instances using an Amazon EBS volume as the root device. For more
information, see Amazon EC2 root device volume (p. 21).
• For more information about AMI and driver requirements for these instances, see Release
notes (p. 222).

To launch a burstable performance instance as Unlimited or Standard (console)

1. Follow the Launch an instance using the Launch Instance Wizard (p. 526) procedure.
2. On the Choose an Instance Type page, select an instance type, and choose Next: Configure
Instance Details.
3. Choose a credit specification.

a. To launch a T3 or T4g instance as standard, clear Unlimited.

b. To launch a T2 instance as unlimited, select Unlimited.
4. Continue as prompted by the wizard. When you've finished reviewing your options on the Review
Instance Launch page, choose Launch. For more information, see Launch an instance using the
Launch Instance Wizard (p. 526).

To launch a burstable performance instance as Unlimited or Standard (AWS CLI)

250
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Use the run-instances command to launch your instances. Specify the credit specification using the --
credit-specification CpuCredits= parameter. Valid credit specifications are unlimited and
standard.

• For T3 and T4g, if you do not include the --credit-specification parameter, the instance
launches as unlimited by default.
• For T2, if you do not include the --credit-specification parameter, the instance launches as
standard by default.

aws ec2 run-instances --image-id ami-abc12345 --count 1 --instance-type t3.micro --key-

name MyKeyPair --credit-specification "CpuCredits=unlimited"

Use an Auto Scaling group to launch a burstable performance instance as Unlimited

When burstable performance instances are launched or started, they require CPU credits for a good
bootstrapping experience. If you use an Auto Scaling group to launch your instances, we recommend
that you configure your instances as unlimited. If you do, the instances use surplus credits when
they are automatically launched or restarted by the Auto Scaling group. Using surplus credits prevents
performance restrictions.

Create a launch template

You must use a launch template for launching instances as unlimited in an Auto Scaling group. A
launch configuration does not support launching instances as unlimited.

To create a launch template that launches instances as Unlimited (console)

1. Follow the Creating a Launch Template for an Auto Scaling Group procedure.
2. In Launch template contents, for Instance type, choose an instance size.
3. To launch instances as unlimited in an Auto Scaling group, under Advanced details, for Credit
specification, choose Unlimited.
4. When you've finished defining the launch template parameters, choose Create launch template.
For more information, see Creating a Launch Template for an Auto Scaling Group in the Amazon EC2
Auto Scaling User Guide.

To create a launch template that launches instances as Unlimited (AWS CLI)

Use the create-launch-template command and specify unlimited as the credit specification.

• For T3 and T4g, if you do not include the CreditSpecification={CpuCredits=unlimited}

value, the instance launches as unlimited by default.
• For T2, if you do not include the CreditSpecification={CpuCredits=unlimited} value, the
instance launches as standard by default.

aws ec2 create-launch-template --launch-template-name MyLaunchTemplate

--version-description FirstVersion --launch-template-data
ImageId=ami-8c1be5f6,InstanceType=t3.medium,CreditSpecification={CpuCredits=unlimited}

Associate an Auto Scaling group with a launch template

To associate the launch template with an Auto Scaling group, create the Auto Scaling group using the
launch template, or add the launch template to an existing Auto Scaling group.

251
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

To create an Auto Scaling group using a launch template (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. On the navigation bar at the top of the screen, select the same Region that you used when you
created the launch template.
3. In the navigation pane, choose Auto Scaling Groups, Create Auto Scaling group.
4. Choose Launch Template, select your launch template, and then choose Next Step.
5. Complete the fields for the Auto Scaling group. When you've finished reviewing your configuration
settings on the Review page, choose Create Auto Scaling group. For more information, see Creating
an Auto Scaling Group Using a Launch Template in the Amazon EC2 Auto Scaling User Guide.

To create an Auto Scaling group using a launch template (AWS CLI)

Use the create-auto-scaling-group AWS CLI command and specify the --launch-template parameter.

To add a launch template to an existing Auto Scaling group (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. On the navigation bar at the top of the screen, select the same Region that you used when you
created the launch template.
3. In the navigation pane, choose Auto Scaling Groups.
4. From the Auto Scaling group list, select an Auto Scaling group, and choose Actions, Edit.
5. On the Details tab, for Launch Template, choose a launch template, and then choose Save.

To add a launch template to an existing Auto Scaling group (AWS CLI)

Use the update-auto-scaling-group AWS CLI command and specify the --launch-template parameter.

View the credit specification of a burstable performance instance

You can view the credit specification (unlimited or standard) of a running or stopped instance.

New console

To view the credit specification of a burstable instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the left navigation pane, choose Instances.
3. Select the instance.
4. Choose Details and view the Credit specification field. The value is either unlimited or
standard.

Old console

To view the credit specification of a burstable instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the left navigation pane, choose Instances.
3. Select the instance.
4. Choose Description and view the T2/T3 Unlimited field.

• If the value is Enabled, then your instance is configured as unlimited.

• If the value is Disabled, then your instance is configured as standard.

252
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

To describe the credit specification of a burstable performance instance (AWS CLI)

Use the describe-instance-credit-specifications command. If you do not specify one or more instance
IDs, all instances with the credit specification of unlimited are returned, as well as instances that were
previously configured with the unlimited credit specification. For example, if you resize a T3 instance
to an M4 instance, while it is configured as unlimited, Amazon EC2 returns the M4 instance.

Example

aws ec2 describe-instance-credit-specifications --instance-id i-1234567890abcdef0

The following is example output:

{
"InstanceCreditSpecifications": [
{
"InstanceId": "i-1234567890abcdef0",
"CpuCredits": "unlimited"
}
]
}

Modify the credit specification of a burstable performance instance

You can switch the credit specification of a running or stopped instance at any time between unlimited
and standard.

New console

To modify the credit specification of a burstable performance instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the left navigation pane, choose Instances.
3. Select the instance. To modify the credit specification for several instances at one time, select all
applicable instances.
4. Choose Actions, Instance settings, Change credit specification. This option is enabled only if
you selected a burstable performance instance.
5. To change the credit specification to unlimited, select the check box next to the instance ID.
To change the credit specification to standard, clear the check box next to the instance ID.

Old console

To modify the credit specification of a burstable performance instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the left navigation pane, choose Instances.
3. Select the instance. To modify the credit specification for several instances at one time, select all
applicable instances.
4. Choose Actions, Instance Settings, Change T2/T3 Unlimited. This option is enabled only if you
selected a burstable performance instance.
5. The current credit specification appears in parentheses after the instance ID. To change the
credit specification to unlimited, choose Enable. To change the credit specification to
standard, choose Disable.

To modify the credit specification of a burstable performance instance (AWS CLI)

253
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Use the modify-instance-credit-specification command. Specify the instance and its credit specification
using the --instance-credit-specification parameter. Valid credit specifications are unlimited
and standard.

Example

aws ec2 modify-instance-credit-specification --region us-east-1 --instance-credit-

specification "InstanceId=i-1234567890abcdef0,CpuCredits=unlimited"

The following is example output:

{
"SuccessfulInstanceCreditSpecifications": [
{
"InstanceId": "i- 1234567890abcdef0"
}
],
"UnsuccessfulInstanceCreditSpecifications": []
}

Set the default credit specification for the account

You can set the default credit specification at the account level per AWS Region. You specify the default
credit specification per instance family (for example, T2 or T3).

If you use the Launch Instance Wizard in the AWS Management Console to launch instances, the value
you select for the credit specification overrides the account-level default credit specification. If you
use the AWS CLI to launch instances, all new burstable performance instances in the account launch
using the default credit option. The credit specification for existing running or stopped instances is not
affected.

The modify-default-credit-specification API is an asynchronous operation, which works at

an AWS Region level and modifies the credit option for each Availability Zone. All zones in a Region are
updated within five minutes. But if instances are launched during this operation, they might not get
the new credit option until the zone is updated. To verify whether the update has occurred, you can call
get-default-credit-specification and check the default credit specification for updates. For
more information, see View the default credit specification (p. 254).

Consideration

The default credit specification for an instance family can be modified only once in a rolling 5-minute
period, and up to four times in a rolling 24-hour period.

To set the default credit specification at the account level (AWS CLI)

Use the modify-default-credit-specification command. Specify the AWS Region, instance family, and the
default credit specification using the --cpu-credits parameter. Valid default credit specifications are
unlimited and standard.

aws ec2 modify-default-credit-specification --region us-east-1 --instance-family t2 --cpu-

credits unlimited

View the default credit specification

You can view the default credit specification of a burstable performance instance family at the account
level per AWS Region.

To view the default credit specification at the account level (AWS CLI)

Use the get-default-credit-specification command. Specify the AWS Region and instance family.

254
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

aws ec2 get-default-credit-specification --region us-east-1 --instance-family t2

Monitor your CPU credits

You can see the credit balance for each instance in the Amazon EC2 per-instance metrics of the
CloudWatch console.

Contents
• Additional CloudWatch metrics for burstable performance instances (p. 255)
• Calculate CPU credit usage (p. 256)

Additional CloudWatch metrics for burstable performance instances

Burstable performance instances have these additional CloudWatch metrics, which are updated every
five minutes:

• CPUCreditUsage – The number of CPU credits spent during the measurement period.
• CPUCreditBalance – The number of CPU credits that an instance has accrued. This balance is
depleted when the CPU bursts and CPU credits are spent more quickly than they are earned.
• CPUSurplusCreditBalance – The number of surplus CPU credits spent to sustain CPU utilization
when the CPUCreditBalance value is zero.
• CPUSurplusCreditsCharged – The number of surplus CPU credits exceeding the maximum number
of CPU credits (p. 225) that can be earned in a 24-hour period, and thus attracting an additional
charge.

The last two metrics apply only to instances configured as unlimited.

The following table describes the CloudWatch metrics for burstable performance instances. For more
information, see List the available CloudWatch metrics for your instances (p. 763).

Metric Description

CPUCreditUsage The number of CPU credits spent by the instance for CPU
utilization. One CPU credit equals one vCPU running at 100%
utilization for one minute or an equivalent combination of vCPUs,
utilization, and time (for example, one vCPU running at 50%
utilization for two minutes or two vCPUs running at 25% utilization
for two minutes).

CPU credit metrics are available at a five-minute frequency only. If

you specify a period greater than five minutes, use the Sum statistic
instead of the Average statistic.

Units: Credits (vCPU-minutes)

CPUCreditBalance The number of earned CPU credits that an instance has

accrued since it was launched or started. For T2 Standard, the
CPUCreditBalance also includes the number of launch credits
that have been accrued.

Credits are accrued in the credit balance after they are earned,
and removed from the credit balance when they are spent. The
credit balance has a maximum limit, determined by the instance
size. After the limit is reached, any new credits that are earned are

255
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Metric Description
discarded. For T2 Standard, launch credits do not count towards the
limit.

The credits in the CPUCreditBalance are available for the

instance to spend to burst beyond its baseline CPU utilization.

When an instance is running, credits in the CPUCreditBalance

do not expire. When a T3 or T4g instance stops, the
CPUCreditBalance value persists for seven days. Thereafter,
all accrued credits are lost. When a T2 instance stops, the
CPUCreditBalance value does not persist, and all accrued credits
are lost.

CPU credit metrics are available at a five-minute frequency only.

Units: Credits (vCPU-minutes)

CPUSurplusCreditBalance The number of surplus credits that have been spent by an

unlimited instance when its CPUCreditBalance value is zero.

The CPUSurplusCreditBalance value is paid down by earned

CPU credits. If the number of surplus credits exceeds the maximum
number of credits that the instance can earn in a 24-hour period,
the spent surplus credits above the maximum incur an additional
charge.

Units: Credits (vCPU-minutes)

CPUSurplusCreditsCharged The number of spent surplus credits that are not paid down by
earned CPU credits, and which thus incur an additional charge.

Spent surplus credits are charged when any of the following occurs:

• The spent surplus credits exceed the maximum number of credits

that the instance can earn in a 24-hour period. Spent surplus
credits above the maximum are charged at the end of the hour.
• The instance is stopped or terminated.
• The instance is switched from unlimited to standard.

Units: Credits (vCPU-minutes)

Calculate CPU credit usage

The CPU credit usage of instances is calculated using the instance CloudWatch metrics described in the
preceding table.

Amazon EC2 sends the metrics to CloudWatch every five minutes. A reference to the prior value of a
metric at any point in time implies the previous value of the metric, sent five minutes ago.

Calculate CPU credit usage for Standard instances

• The CPU credit balance increases if CPU utilization is below the baseline, when the credits spent are
less than the credits earned in the prior five-minute interval.
• The CPU credit balance decreases if CPU utilization is above the baseline, when the credits spent are
more than the credits earned in the prior five-minute interval.

256
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Mathematically, this is captured by the following equation:

Example

CPUCreditBalance = prior CPUCreditBalance + [Credits earned per hour * (5/60) -

CPUCreditUsage]

The size of the instance determines the number of credits that the instance can earn per hour and the
number of earned credits that it can accrue in the credit balance. For information about the number of
credits earned per hour, and the credit balance limit for each instance size, see the credit table (p. 225).

Example

This example uses a t3.nano instance. To calculate the CPUCreditBalance value of the instance, use
the preceding equation as follows:

• CPUCreditBalance – The current credit balance to calculate.

• prior CPUCreditBalance – The credit balance five minutes ago. In this example, the instance had
accrued two credits.
• Credits earned per hour – A t3.nano instance earns six credits per hour.
• 5/60 – Represents the five-minute interval between CloudWatch metric publication. Multiply the
credits earned per hour by 5/60 (five minutes) to get the number of credits that the instance earned in
the past five minutes. A t3.nano instance earns 0.5 credits every five minutes.
• CPUCreditUsage – How many credits the instance spent in the past five minutes. In this example, the
instance spent one credit in the past five minutes.

Using these values, you can calculate the CPUCreditBalance value:

Example

CPUCreditBalance = 2 + [0.5 - 1] = 1.5

Calculate CPU credit usage for Unlimited instances

When a burstable performance instance needs to burst above the baseline, it always spends accrued
credits before spending surplus credits. When it depletes its accrued CPU credit balance, it can spend
surplus credits to burst CPU for as long as it needs. When CPU utilization falls below the baseline, surplus
credits are always paid down before the instance accrues earned credits.

We use the term Adjusted balance in the following equations to reflect the activity that occurs in
this five-minute interval. We use this value to arrive at the values for the CPUCreditBalance and
CPUSurplusCreditBalance CloudWatch metrics.

Example

Adjusted balance = [prior CPUCreditBalance - prior CPUSurplusCreditBalance] + [Credits

earned per hour * (5/60) - CPUCreditUsage]

A value of 0 for Adjusted balance indicates that the instance spent all its earned credits
for bursting, and no surplus credits were spent. As a result, both CPUCreditBalance and
CPUSurplusCreditBalance are set to 0.

A positive Adjusted balance value indicates that the instance accrued earned credits, and previous
surplus credits, if any, were paid down. As a result, the Adjusted balance value is assigned to
CPUCreditBalance, and the CPUSurplusCreditBalance is set to 0. The instance size determines the
maximum number of credits (p. 225) that it can accrue.

257
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Example

CPUCreditBalance = min [max earned credit balance, Adjusted balance]

CPUSurplusCreditBalance = 0

A negative Adjusted balance value indicates that the instance spent all its earned credits that it
accrued and, in addition, also spent surplus credits for bursting. As a result, the Adjusted balance
value is assigned to CPUSurplusCreditBalance and CPUCreditBalance is set to 0. Again, the
instance size determines the maximum number of credits (p. 225) that it can accrue.

Example

CPUSurplusCreditBalance = min [max earned credit balance, -Adjusted balance]

CPUCreditBalance = 0

If the surplus credits spent exceed the maximum credits that the instance can accrue, the surplus credit
balance is set to the maximum, as shown in the preceding equation. The remaining surplus credits are
charged as represented by the CPUSurplusCreditsCharged metric.

Example

CPUSurplusCreditsCharged = max [-Adjusted balance - max earned credit balance, 0]

Finally, when the instance terminates, any surplus credits tracked by the CPUSurplusCreditBalance
are charged. If the instance is switched from unlimited to standard, any remaining
CPUSurplusCreditBalance is also charged.

Amazon EC2 Mac instances

Mac1 instances natively support the macOS operating system. They are built on Mac mini hardware and
powered by 3.2 GHz Intel eighth-generation (Coffee Lake) Core i7 processors. These instances are ideal
for developing, building, testing, and signing applications for Apple devices, such as iPhone, iPad, iPod,
Mac, Apple Watch, and Apple TV. You can connect to your Mac instance using SSH or Virtual Network
Computing (VNC).

For more information, see Amazon EC2 Mac Instances.

Contents
• Considerations (p. 258)
• Launch a Mac instance using the console (p. 259)
• Launch a Mac instance using the AWS CLI (p. 260)
• Connect to your instance using SSH (p. 260)
• Connect to your instance using VNC (p. 261)
• EC2 macOS AMIs (p. 262)
• Update the operating system and software (p. 262)
• EC2 macOS Init (p. 263)
• EC2 System Monitoring for macOS (p. 263)
• Clean up your instance (p. 263)

Considerations
The following considerations apply to Mac instances:

258
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

• Mac instances are available only as bare metal instances on Dedicated Hosts, with a minimum
allocation period of 24 hours before you can release the Dedicated Host. You can launch one
Mac instance per Dedicated Host. You can share the Dedicated Host with the AWS accounts or
organizational units within your AWS organization, or the entire AWS organization.
• Mac instances are available only as On-Demand Instances. They are not available as Spot Instances or
Reserved Instances. You can save money on Mac instances by purchasing a Savings Plan.
• Mac instances run either macOS Mojave (version 10.14) or macOS Catalina (version 10.15).
• If you attach an EBS volume to a running Mac instance, you must reboot the instance to make the
volume available.
• If you attach a network interface to a running Mac instance, you must reboot the instance to make the
network interface available.
• AWS does not manage or support the internal SSD on the Apple hardware. We strongly recommend
that you use Amazon EBS volumes instead. EBS volumes provide the same elasticity, availability, and
durability benefits on Mac instances as they do on any other EC2 instance.
• We recommend using General Purpose SSD (gp2 and gp3) and Provisioned IOPS SSD (io1 and io2)
with Mac instances for optimal EBS performance.
• You cannot use Mac instances with Amazon EC2 Auto Scaling.
• Automatic software updates are disabled. We recommend that you apply updates and test them on
your instance before you put the instance into production. For more information, see Update the
operating system and software (p. 262).

Launch a Mac instance using the console

You can launch a Mac instance using the AWS Management Console as described in the following
procedure. Mac instances require a Dedicated Host.

To launch a Mac instance onto a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. Choose Allocate Dedicated Host and then do the following:

a. For Instance family, choose mac1. If mac1 doesn’t appear in the list, it’s not supported in the
currently selected Region.
b. For Support multiple instance types, clear Enable. For Instance type, select mac1.metal.
c. For Availability Zone, choose the Availability Zone for the Dedicated Host.
d. For Quantity, keep 1.
e. Choose Allocate.
4. Select the Dedicated Host that you created and then do the following:

a. Choose Actions, Launch instances onto host.

b. Select a macOS AMI.
c. Select the mac1.metal instance type.
d. On the Configure Instance Details page, verify that Tenancy and Host are preconfigured based
on the Dedicated Host you created. Update Affinity as needed.
e. Complete the wizard, specifying EBS volumes, security groups, and key pairs as needed.
5. A confirmation page lets you know that your instance is launching. Choose View Instances to close
the confirmation page and return to the console. The initial state of an instance is pending. The
instance is ready when its state changes to running and it passes status checks.

259
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

Launch a Mac instance using the AWS CLI

Use the following allocate-hosts command to allocate a Dedicated Host for your Mac instance.

aws ec2 allocate-hosts --instance-type mac1.metal --availability-zone us-east-1b --quantity

1

Use the following run-instances command to launch a Mac instance.

aws ec2 run-instances --instance-type mac1.metal --placement Tenancy=host --availability-

zone us-east-1b --image-id ami_id --key-name my-key-pair

The initial state of an instance is pending. The instance is ready when its state changes to running
and it passes status checks. Use the following describe-instance-status command to display status
information for your instance:

aws ec2 describe-instance-status --instance-ids i-017f8354e2dc69c4f

The following is example output for an instance that is running and has passed status checks.

{
"InstanceStatuses": [
{
"AvailabilityZone": "us-east-1b",
"InstanceId": "i-017f8354e2dc69c4f",
"InstanceState": {
"Code": 16,
"Name": "running"
},
"InstanceStatus": {
"Details": [
{
"Name": "reachability",
"Status": "passed"
}
],
"Status": "ok"
},
"SystemStatus": {
"Details": [
{
"Name": "reachability",
"Status": "passed"
}
],
"Status": "ok"
}
}
]
}

Connect to your instance using SSH

Amazon EC2 Mac instances do not allow remote root SSH by default. Password authentication is disabled
to prevent brute-force password attacks. The ec2-user account is configured to log in remotely using
SSH. The ec2-user account also has sudo privileges. After you connect to your instance, you can add
other users.

260
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

To support connecting to your instance using SSH, launch the instance using a key pair and a security
group that allows SSH access, and ensure that the instance has internet connectivity. You provide the
.pem file for the key pair when you connect to the instance.

Use the following procedure to connect to your Mac instance using an SSH client. If you receive an error
while attempting to connect to your instance, see Troubleshoot connecting to your instance (p. 1329).

To connect to your instance using SSH

1. Verify that your local computer has an SSH client installed by entering ssh at the command line. If
your computer doesn't recognize the command, search for an SSH client for your operating system
and install it.
2. Get the public DNS name of your instance. Using the Amazon EC2 console, you can find the public
DNS name on both the Details and the Networking tabs. Using the AWS CLI, you can find the public
DNS name using the describe-instances command.
3. Locate the .pem file for the key pair that you specified when you launched the instance.
4. Connect to your instance using the following ssh command, specifying the public DNS name of the
instance and the .pem file.

ssh -i /path/my-key-pair.pem ec2-user@my-instance-public-dns-name

Connect to your instance using VNC

Use the following procedure to connect to your instance using Virtual Network Computing (VNC).

To connect to your instance using VNC

1. Verify that your local computer has a VNC client installed. On macOS, you can leverage the built-in
Screen Sharing application. Otherwise, search for a VNC client for your operating system and install
it.
2. From your local computer, connect to your instance using SSH (p. 260).
3. Set up a password for the ec2-user account using the passwd command as follows.

[ec2-user ~]$ sudo passwd ec2-user

4. Start the VNC server and enable remote desktop access as follows.

[ec2-user ~]$ sudo /System/Library/CoreServices/RemoteManagement/ARDAgent.app/Contents/

Resources/kickstart \
-activate -configure -access -on \
-restart -agent -privs -all

5. From your computer, connect to your instance using the following ssh command. In addition to the
options shown in the previous section, use the -L option to enable port forwarding and forward all
traffic on local port 5900 to the VNC server on the instance.

ssh -L 5900:localhost:5900 -i /path/my-key-pair.pem ec2-user@my-instance-public-dns-

name

6. From your local computer, use the VNC client to connect to localhost on port 5900. For example, use
the Screen Sharing application on macOS as follows:

a. Open Finder and launch the Screen Sharing application.

b. For Connect to, enter localhost.

261
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

c. Log in as prompted, using ec2-user as the user name and the password that you created for
the ec2-user account.

EC2 macOS AMIs

Amazon EC2 macOS is designed to provide a stable, secure, and high-performance environment for
developer workloads running on Amazon EC2 Mac instances. EC2 macOS AMIs includes packages that
enable easy integration with AWS, such as launch configuration tools and popular AWS libraries and
tools. EC2 macOS AMIs include the following by default:

• ENA drivers
• EC2 macOS Init
• EC2 System Monitoring for macOS
• SSM Agent for macOS
• AWS Command Line Interface (AWS CLI) version 2
• Command Line Tools for Xcode
• Homebrew

AWS provides updated EC2 macOS AMIs on a regular basis that include updates to AWS-owned packages
and the latest fully-tested macOS version. Additionally, AWS provides updated AMIs with the latest
minor version updates or major version updates as soon as they can be fully tested and vetted. If you
do not need to preserve data or customizations to your Mac instances, you can get the latest updates by
launching a new instance using the current AMI and then terminating the previous instance. Otherwise,
you can choose which updates to apply to your Mac instances.

Update the operating system and software

You can install operating system updates from Apple using the softwareupdate command.

To install operating system updates from Apple

1. List the packages with available updates using the following command.

[ec2-user ~]$ softwareupdate --list

2. Install all updates or only specific updates. To install specific updates, use the following command.

[ec2-user ~]$ sudo softwareupdate --install label

To install all updates instead, use the following command.

[ec2-user ~]$ sudo softwareupdate --install

System administrators can use AWS Systems Manager to roll out pre-approved operating system
updates. For more information, see the AWS Systems Manager User Guide.

You can use Homebrew to install updates to packages in the EC2 macOS AMIs, so that you have the
latest version of these packages on your instances. You can also use Homebrew to install and run
common macOS applications on Amazon EC2 macOS. For more information, see the Homebrew
Documentation.

262
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General purpose

To install updates using Homebrew

1. Update Homebrew using the following command.

[ec2-user ~]$ brew update

2. List the packages with available updates using the following command.

[ec2-user ~]$ brew outdated

3. Install all updates or only specific updates. To install specific updates, use the following command.

[ec2-user ~]$ brew upgrade formula

To install all updates instead, use the following command.

[ec2-user ~]$ brew upgrade

EC2 macOS Init

EC2 macOS Init is used to initialize EC2 Mac instances at launch. It uses priority groups to run logical
groups of tasks at the same time.

The launchd plist file is /Library/LaunchDaemons/com.amazon.ec2.macos-init.plist. The files

for EC2 macOS Init are located in /usr/local/aws/ec2-macos-init.

For more information, see https://github.com/aws/ec2-macos-init.

EC2 System Monitoring for macOS

EC2 System Monitoring for macOS provides CPU utilization metrics to Amazon CloudWatch. It sends
these metrics to CloudWatch over a custom serial device in 1-minute periods. You can enable or disable
this agent as follows. It is enabled by default.

sudo setup-ec2monitoring [enable | disable]

Clean up your instance

When you are finished with your Mac instance, you can clean up by releasing the Dedicated Host. Before
you can release the Dedicated Host, you must stop or terminate the Mac instance. You cannot release the
host until the allocation period exceeds the 24-hour minimum.

To release a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Instance state, then choose either Stop instance or Terminate
instance.
4. In the navigation pane, choose Dedicated Hosts.
5. Select the Dedicated Host and choose Actions, Release host.
6. When prompted for confirmation, choose Release.

263
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Compute optimized

Compute optimized instances

Compute optimized instances are ideal for compute-bound applications that benefit from high-
performance processors.

C5 and C5n instances

These instances are well suited for the following:

• Batch processing workloads

• Media transcoding
• High-performance web servers
• High-performance computing (HPC)
• Scientific modeling
• Dedicated gaming servers and ad serving engines
• Machine learning inference and other compute-intensive applications

Bare metal instances, such as c5.metal, provide your applications with direct access to physical
resources of the host server, such as processors and memory.

For more information, see Amazon EC2 C5 Instances.

C6g, C6gd, and C6gn instances

These instances are powered by AWS Graviton2 processors and are ideal for running advanced, compute-
intensive workloads, such as the following:

• High-performance computing (HPC)

• Batch processing
• Ad serving
• Video encoding
• Gaming servers
• Scientific modeling
• Distributed analytics
• CPU-based machine learning inference

Bare metal instances, such as c6g.metal, provide your applications with direct access to physical
resources of the host server, such as processors and memory.

For more information, see Amazon EC2 C6g Instances.

Contents
• Hardware specifications (p. 265)
• Instance performance (p. 267)
• Network performance (p. 267)
• SSD I/O performance (p. 268)
• Instance features (p. 270)
• Release notes (p. 270)

264
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Compute optimized

Hardware specifications
The following is a summary of the hardware specifications for compute optimized instances.

Instance type Default vCPUs Memory (GiB)

c4.large 2 3.75

c4.xlarge 4 7.5

c4.2xlarge 8 15

c4.4xlarge 16 30

c4.8xlarge 36 60

c5.large 2 4

c5.xlarge 4 8

c5.2xlarge 8 16

c5.4xlarge 16 32

c5.9xlarge 36 72

c5.12xlarge 48 96

c5.18xlarge 72 144

c5.24xlarge 96 192

c5.metal 96 192

c5a.large 2 4

c5a.xlarge 4 8

c5a.2xlarge 8 16

c5a.4xlarge 16 32

c5a.8xlarge 32 64

c5a.12xlarge 48 96

c5a.16xlarge 64 128

c5a.24xlarge 96 192

c5ad.large 2 4

c5ad.xlarge 4 8

c5ad.2xlarge 8 16

c5ad.4xlarge 16 32

c5ad.8xlarge 32 64

c5ad.12xlarge 48 96

265
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Compute optimized

Instance type Default vCPUs Memory (GiB)

c5ad.16xlarge 64 128

c5ad.24xlarge 96 192

c5d.large 2 4

c5d.xlarge 4 8

c5d.2xlarge 8 16

c5d.4xlarge 16 32

c5d.9xlarge 36 72

c5d.12xlarge 48 96

c5d.18xlarge 72 144

c5d.24xlarge 96 192

c5d.metal 96 192

c5n.large 2 5.25

c5n.xlarge 4 10.5

c5n.2xlarge 8 21

c5n.4xlarge 16 42

c5n.9xlarge 36 96

c5n.18xlarge 72 192

c5n.metal 72 192

c6g.medium 1 2

c6g.large 2 4

c6g.xlarge 4 8

c6g.2xlarge 8 16

c6g.4xlarge 16 32

c6g.8xlarge 32 64

c6g.12xlarge 48 96

c6g.16xlarge 64 128

c6g.metal 64 128

c6gd.medium 1 2

c6gd.large 2 4

c6gd.xlarge 4 8

c6gd.2xlarge 8 16

266
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Compute optimized

Instance type Default vCPUs Memory (GiB)

c6gd.4xlarge 16 32

c6gd.8xlarge 32 64

c6gd.12xlarge 48 96

c6gd.16xlarge 64 128

c6gd.metal 64 128

c6gn.medium 1 2

c6gn.large 2 4

c6gn.xlarge 4 8

c6gn.2xlarge 8 16

c6gn.4xlarge 16 32

c6gn.8xlarge 32 64

c6gn.12xlarge 48 96

c6gn.16xlarge 64 128

For more information about the hardware specifications for each Amazon EC2 instance type, see Amazon
EC2 Instance Types.

For more information about specifying CPU options, see Optimize CPU options (p. 668).

Instance performance
EBS-optimized instances enable you to get consistently high performance for your EBS volumes by
eliminating contention between Amazon EBS I/O and other network traffic from your instance. Some
compute optimized instances are EBS-optimized by default at no additional cost. For more information,
see Amazon EBS–optimized instances (p. 1217).

Some compute optimized instance types provide the ability to control processor C-states and P-states on
Linux. C-states control the sleep levels that a core can enter when it is inactive, while P-states control the
desired performance (in CPU frequency) from a core. For more information, see Processor state control
for your EC2 instance (p. 658).

Network performance
You can enable enhanced networking on supported instance types to provide lower latencies, lower
network jitter, and higher packet-per-second (PPS) performance. Most applications do not consistently
need a high level of network performance, but can benefit from access to increased bandwidth when
they send or receive data. For more information, see Enhanced networking on Linux (p. 864).

The following is a summary of network performance for compute optimized instances that support
enhanced networking.

Instance type Network performance Enhanced networking

c4.large Moderate Intel 82599 VF (p. 875)

267
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Compute optimized

Instance type Network performance Enhanced networking

c4.xlarge | c4.2xlarge | High Intel 82599 VF (p. 875)

c4.4xlarge

c5.4xlarge and smaller | Up to 10 Gbps † ENA (p. 865)

c5d.4xlarge and smaller |
c6g.4xlarge and smaller |
c6gd.4xlarge and smaller

c4.8xlarge 10 Gbps Intel 82599 VF (p. 875)

c5.9xlarge | c5d.9xlarge 10 Gbps ENA (p. 865)

c5.12xlarge | c5d.12xlarge 12 Gbps ENA (p. 865)

| c6g.8xlarge |
c6gd.8xlarge

c6g.12xlarge | 20 Gbps ENA (p. 865)

c6gd.12xlarge

c5n.4xlarge and smaller | Up to 25 Gbps † ENA (p. 865)

c6gn.2xlarge and smaller

c5.18xlarge | c5.24xlarge 25 Gbps ENA (p. 865)

| c5.metal | c5d.18xlarge |
c5d.24xlarge | c5d.metal |
c6g.16xlarge | c6g.metal |
c6gd.16xlarge | c6gd.metal
| c6gn.4xlarge

c5n.9xlarge | c6gn.8xlarge 50 Gbps ENA (p. 865)

c6gn.12xlarge 75 Gbps ENA (p. 865)

c5n.18xlarge | c5n.metal | 100 Gbps ENA (p. 865)

c6gn.16xlarge

† These instances use a network I/O credit mechanism to allocate network bandwidth to instances based
on average bandwidth utilization. They accrue credits when their bandwidth is below their baseline
bandwidth, and can use these credits when they perform network data transfers. For more information,
open a support case and ask about baseline bandwidth for the specific instance types that you are
interested in.

SSD I/O performance

If you use a Linux AMI with kernel version 4.4 or later and use all the SSD-based instance store volumes
available to your instance, you get the IOPS (4,096 byte block size) performance listed in the following
table (at queue depth saturation). Otherwise, you get lower IOPS performance.

Instance Size 100% Random Read IOPS Write IOPS

c5ad.large 16,283 7,105

c5ad.xlarge 32,566 14,211

c5ad.2xlarge 65,132 28,421

268
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Compute optimized

Instance Size 100% Random Read IOPS Write IOPS

c5ad.4xlarge 130,263 56,842

c5ad.8xlarge 260,526 113,684

c5ad.12xlarge 412,500 180,000

c5ad.16xlarge 521,053 227,368

c5ad.24xlarge 825,000 360,000

c5d.large * 20,000 9,000

c5d.xlarge * 40,000 18,000

c5d.2xlarge * 80,000 37,000

c5d.4xlarge * 175,000 75,000

c5d.9xlarge 350,000 170,000

c5d.12xlarge 700,000 340,000

c5d.18xlarge 700,000 340,000

c5d.24xlarge 1,400,000 680,000

c5d.metal 1,400,000 680,000

c6gd.medium 13,438 5,625

c6gd.large 26,875 11,250

c6gd.xlarge 53,750 22,500

c6gd.2xlarge 107,500 45,000

c6gd.4xlarge 215,000 90,000

c6gd.8xlarge 430,000 180,000

c6gd.12xlarge 645,000 270,000

c6gd.16xlarge 860,000 360,000

c6gd.metal 860,000 360,000

* For these instances, you can get up to the specified performance.

As you fill the SSD-based instance store volumes for your instance, the number of write IOPS that
you can achieve decreases. This is due to the extra work the SSD controller must do to find available
space, rewrite existing data, and erase unused space so that it can be rewritten. This process of
garbage collection results in internal write amplification to the SSD, expressed as the ratio of SSD write
operations to user write operations. This decrease in performance is even larger if the write operations
are not in multiples of 4,096 bytes or not aligned to a 4,096-byte boundary. If you write a smaller
amount of bytes or bytes that are not aligned, the SSD controller must read the surrounding data and
store the result in a new location. This pattern results in significantly increased write amplification,
increased latency, and dramatically reduced I/O performance.

269
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Compute optimized

SSD controllers can use several strategies to reduce the impact of write amplification. One such strategy
is to reserve space in the SSD instance storage so that the controller can more efficiently manage the
space available for write operations. This is called over-provisioning. The SSD-based instance store
volumes provided to an instance don't have any space reserved for over-provisioning. To reduce write
amplification, we recommend that you leave 10% of the volume unpartitioned so that the SSD controller
can use it for over-provisioning. This decreases the storage that you can use, but increases performance
even if the disk is close to full capacity.

For instance store volumes that support TRIM, you can use the TRIM command to notify the SSD
controller whenever you no longer need data that you've written. This provides the controller with more
free space, which can reduce write amplification and increase performance. For more information, see
Instance store volume TRIM support (p. 1282).

Instance features
The following is a summary of features for compute optimized instances:

  EBS only NVMe EBS Instance store Placement group

C4 Yes No No Yes

C5 Yes Yes No Yes

C5a Yes Yes No Yes

C5ad No Yes NVMe * Yes

C5d No Yes NVMe * Yes

C5n Yes Yes No Yes

C6g Yes Yes No Yes

C6gd No Yes NVMe * Yes

C6gn Yes Yes No Yes

* The root device volume must be an Amazon EBS volume.

For more information, see the following:

• Amazon EBS and NVMe on Linux instances (p. 1214)

• Amazon EC2 instance store (p. 1268)
• Placement groups (p. 927)

Release notes
• C5 and C5d instances feature a 3.1 GHz Intel Xeon Platinum 8000 series processor from either the first
generation (Skylake-SP) or second generation (Cascade Lake).
• C5a and C5ad instances feature a second-generation AMD EPYC processor (Rome) running at
frequencies as high as 3.3. GHz.
• C6g, C6gd, and C6gn instances feature an AWS Graviton2 processor based on 64-bit Arm architecture.
• C4 instances and instances built on the Nitro System (p. 209) require 64-bit EBS-backed HVM AMIs.
They have high-memory and require a 64-bit operating system to take advantage of that capacity.
HVM AMIs provide superior performance in comparison to paravirtual (PV) AMIs on high-memory
instance types. In addition, you must use an HVM AMI to take advantage of enhanced networking.

270
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Compute optimized

• Instances built on the Nitro System have the following requirements:

• NVMe drivers (p. 1214) must be installed
• Elastic Network Adapter (ENA) drivers (p. 865) must be installed

The following Linux AMIs meet these requirements:

• Amazon Linux 2
• Amazon Linux AMI 2018.03
• Ubuntu 14.04 (with linux-aws kernel) or later
• Red Hat Enterprise Linux 7.4 or later
• SUSE Linux Enterprise Server 12 SP2 or later
• CentOS 7.4.1708 or later
• FreeBSD 11.1 or later
• Debian GNU/Linux 9 or later
• Instances with an AWS Graviton processors have the following requirements:
• Use an AMI for the 64-bit Arm architecture.
• Support booting through UEFI with ACPI tables and support ACPI hot-plug of PCI devices.

The following AMIs meet these requirements:

• Amazon Linux 2 (64-bit Arm)
• Ubuntu 16.04 or later (64-bit Arm)
• Red Hat Enterprise Linux 8.0 or later (64-bit Arm)
• SUSE Linux Enterprise Server 15 or later (64-bit Arm)
• Debian 10 or later (64-bit Arm)
• Instances built on the Nitro System instances support a maximum of 28 attachments, including
network interfaces, EBS volumes, and NVMe instance store volumes. For more information, see Nitro
System volume limits (p. 1291).
• To get the best performance from your C6gn instances, ensure that they have ENA driver version 2.2.9
or later.
• Traffic Mirroring is not supported on C6gn instances.
• To launch AMIs for all Linux distributions on C6gn instances, use AMIs with the latest version and run
an update for the latest driver. For earlier versions, download the latest driver from GitHub.
• Launching a bare metal instance boots the underlying server, which includes verifying all hardware and
firmware components. This means that it can take 20 minutes from the time the instance enters the
running state until it becomes available over the network.
• To attach or detach EBS volumes or secondary network interfaces from a bare metal instance requires
PCIe native hotplug support. Amazon Linux 2 and the latest versions of the Amazon Linux AMI
support PCIe native hotplug, but earlier versions do not. You must enable the following Linux kernel
configuration options:

CONFIG_HOTPLUG_PCI_PCIE=y
CONFIG_PCIEASPM=y

• Bare metal instances use a PCI-based serial device rather than an I/O port-based serial device. The
upstream Linux kernel and the latest Amazon Linux AMIs support this device. Bare metal instances also
provide an ACPI SPCR table to enable the system to automatically use the PCI-based serial device. The
latest Windows AMIs automatically use the PCI-based serial device.
• Instances built on the Nitro System should have acpid installed to support clean shutdown through API
requests.
• There is a limit on the total number of instances that you can launch in a Region, and there are
additional limits on some instance types. For more information, see How many instances can I run in
Amazon EC2? in the Amazon EC2 FAQ.

271
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Memory optimized

Memory optimized instances

Memory optimized instances are designed to deliver fast performance for workloads that process large
data sets in memory.

R5, R5a, R5b, and R5n instances

These instances are well suited for the following:

• High-performance, relational (MySQL) and NoSQL (MongoDB, Cassandra) databases.

• Distributed web scale cache stores that provide in-memory caching of key-value type data
(Memcached and Redis).
• In-memory databases using optimized data storage formats and analytics for business intelligence (for
example, SAP HANA).
• Applications performing real-time processing of big unstructured data (financial services, Hadoop/
Spark clusters).
• High-performance computing (HPC) and Electronic Design Automation (EDA) applications.

Bare metal instances, such as r5.metal, provide your applications with direct access to physical
resources of the host server, such as processors and memory.

For more information, see Amazon EC2 R5 Instances.

R6g and R6gd instances

These instances are powered by AWS Graviton2 processors and are ideal for running memory-intensive
workloads, such as the following:

• Open-source databases (for example, MySQL, MariaDB, and PostgreSQL)

• In-memory caches (for example, Memcached, Redis, and KeyDB)

Bare metal instances, such as r6g.metal, provide your applications with direct access to physical
resources of the host server, such as processors and memory.

For more information, see Amazon EC2 R6g Instances.

High memory instances

High memory instances (u-6tb1.metal, u-9tb1.metal, u-12tb1.metal, u-18tb1.metal, and

u-24tb1.metal) offer 6 TiB, 9 TiB, 12 TiB, 18 TiB, and 24 TiB of memory per instance. These instances
are designed to run large in-memory databases, including production deployments of the SAP HANA in-
memory database, in the cloud. They offer bare metal performance with direct access to host hardware.

For more information, see Amazon EC2 High Memory Instances and Storage Configuration for SAP
HANA.

X1 instances

These instances are well suited for the following:

• In-memory databases such as SAP HANA, including SAP-certified support for Business Suite S/4HANA,
Business Suite on HANA (SoH), Business Warehouse on HANA (BW), and Data Mart Solutions on HANA.
For more information, see SAP HANA on the AWS Cloud.
• Big-data processing engines such as Apache Spark or Presto.
• High-performance computing (HPC) applications.

272
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Memory optimized

For more information, see Amazon EC2 X1 Instances.

X1e instances
These instances are well suited for the following:

• High-performance databases.
• In-memory databases such as SAP HANA. For more information, see SAP HANA on the AWS Cloud.
• Memory-intensive enterprise applications.

For more information, see Amazon EC2 X1e Instances.

z1d instances
These instances deliver both high compute and high memory and are well-suited for the following:

• Electronic Design Automation (EDA)

• Relational database workloads

z1d.metal instances provide your applications with direct access to physical resources of the host
server, such as processors and memory.

For more information, see Amazon EC2 z1d Instances.

Contents
• Hardware specifications (p. 273)
• Memory performance (p. 277)
• Instance performance (p. 277)
• Network performance (p. 277)
• SSD I/O performance (p. 278)
• Instance features (p. 280)
• Support for vCPUs (p. 281)
• Release notes (p. 281)

Hardware specifications
The following is a summary of the hardware specifications for memory optimized instances.

Instance type Default vCPUs Memory (GiB)

r4.large 2 15.25

r4.xlarge 4 30.5

r4.2xlarge 8 61

r4.4xlarge 16 122

r4.8xlarge 32 244

r4.16xlarge 64 488

r5.large 2 16

r5.xlarge 4 32

273
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Memory optimized

Instance type Default vCPUs Memory (GiB)

r5.2xlarge 8 64

r5.4xlarge 16 128

r5.8xlarge 32 256

r5.12xlarge 48 384

r5.16xlarge 64 512

r5.24xlarge 96 768

r5.metal 96 768

r5a.large 2 16

r5a.xlarge 4 32

r5a.2xlarge 8 64

r5a.4xlarge 16 128

r5a.8xlarge 32 256

r5a.12xlarge 48 384

r5a.16xlarge 64 512

r5a.24xlarge 96 768

r5ad.large 2 16

r5ad.xlarge 4 32

r5ad.2xlarge 8 64

r5ad.4xlarge 16 128

r5ad.8xlarge 32 256

r5ad.12xlarge 48 384

r5ad.16xlarge 64 512

r5ad.24xlarge 96 768

r5b.large 2 16

r5b.xlarge 4 32

r5b.2xlarge 8 64

r5b.4xlarge 16 128

r5b.8xlarge 32 256

r5b.12xlarge 48 384

r5b.16xlarge 64 512

r5b.24xlarge 96 768

274
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Memory optimized

Instance type Default vCPUs Memory (GiB)

r5b.metal 96 768

r5d.large 2 16

r5d.xlarge 4 32

r5d.2xlarge 8 64

r5d.4xlarge 16 128

r5d.8xlarge 32 256

r5d.12xlarge 48 384

r5d.16xlarge 64 512

r5d.24xlarge 96 768

r5d.metal 96 768

r5dn.large 2 16

r5dn.xlarge 4 32

r5dn.2xlarge 8 64

r5dn.4xlarge 16 128

r5dn.8xlarge 32 256

r5dn.12xlarge 48 384

r5dn.16xlarge 64 512

r5dn.24xlarge 96 768

r5n.large 2 16

r5n.xlarge 4 32

r5n.2xlarge 8 64

r5n.4xlarge 16 128

r5n.8xlarge 32 256

r5n.12xlarge 48 384

r5n.16xlarge 64 512

r5n.24xlarge 96 768

r6g.medium 1 8

r6g.large 2 16

r6g.xlarge 4 32

r6g.2xlarge 8 64

r6g.4xlarge 16 128

275
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Memory optimized

Instance type Default vCPUs Memory (GiB)

r6g.8xlarge 32 256

r6g.12xlarge 48 384

r6g.16xlarge 64 512

r6gd.medium 1 8

r6gd.large 2 16

r6gd.xlarge 4 32

r6gd.2xlarge 8 64

r6gd.4xlarge 16 128

r6gd.8xlarge 32 256

r6gd.12xlarge 48 384

r6gd.16xlarge 64 512

u-6tb1.metal 448 * 6,144

u-9tb1.metal 448 * 9,216

u-12tb1.metal 448 * 12,288

u-18tb1.metal 448 * 18,432

u-24tb1.metal 448 * 24,576

x1.16xlarge 64 976

x1.32xlarge 128 1,952

x1e.xlarge 4 122

x1e.2xlarge 8 244

x1e.4xlarge 16 488

x1e.8xlarge 32 976

x1e.16xlarge 64 1,952

x1e.32xlarge 128 3,904

z1d.large 2 16

z1d.xlarge 4 32

z1d.2xlarge 8 64

z1d.3xlarge 12 96

z1d.6xlarge 24 192

z1d.12xlarge 48 384

z1d.metal 48 384

276
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Memory optimized

* Each logical processor is a hyperthread on 224 cores.

For more information about the hardware specifications for each Amazon EC2 instance type, see Amazon
EC2 Instance Types.

For more information about specifying CPU options, see Optimize CPU options (p. 668).

Memory performance
X1 instances include Intel Scalable Memory Buffers, providing 300 GiB/s of sustainable memory-read
bandwidth and 140 GiB/s of sustainable memory-write bandwidth.

For more information about how much RAM can be enabled for memory optimized instances, see
Hardware specifications (p. 273).

Memory optimized instances have high memory and require 64-bit HVM AMIs to take advantage of that
capacity. HVM AMIs provide superior performance in comparison to paravirtual (PV) AMIs on memory
optimized instances. For more information, see Linux AMI virtualization types (p. 104).

Instance performance
Memory optimized instances enable increased cryptographic performance through the latest Intel AES-
NI feature, support Intel Transactional Synchronization Extensions (TSX) to boost the performance of in-
memory transactional data processing, and support Advanced Vector Extensions 2 (Intel AVX2) processor
instructions to expand most integer commands to 256 bits.

Some memory optimized instances provide the ability to control processor C-states and P-states on
Linux. C-states control the sleep levels that a core can enter when it is inactive, while P-states control
the desired performance (measured by CPU frequency) from a core. For more information, see Processor
state control for your EC2 instance (p. 658).

Network performance
You can enable enhanced networking on supported instance types to provide lower latencies, lower
network jitter, and higher packet-per-second (PPS) performance. Most applications do not consistently
need a high level of network performance, but can benefit from access to increased bandwidth when
they send or receive data. For more information, see Enhanced networking on Linux (p. 864).

The following is a summary of network performance for memory optimized instances that support
enhanced networking.

Instance type Network performance Enhanced networking

r4.4xlarge and smaller | r5.4xlarge Up to 10 Gbps † ENA (p. 865)

and smaller | r5a.8xlarge and smaller |
r5ad.8xlarge and smaller | r5b.4xlarge
and smaller | r5d.4xlarge and smaller |
r6g.4xlarge and smaller | r6gd.4xlarge
and smaller | x1e.8large and smaller |
z1d.3xlarge and smaller

r4.8xlarge | r5.8xlarge | r5.12xlarge 10 Gbps ENA (p. 865)

| r5a.12xlarge | r5ad.12xlarge |
r5b.8xlarge | r5b.12xlarge | r5d.8xlarge |
r5d.12xlarge | x1.16xlarge | x1e.16xlarge
| z1d.6xlarge

r5a.16xlarge | r5ad.16xlarge | 12 Gbps ENA (p. 865)

r6g.8xlarge | r6gd.8xlarge

277
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Memory optimized

Instance type Network performance Enhanced networking

r5.16xlarge | r5a.24xlarge | 20 Gbps ENA (p. 865)

r5ad.24xlarge | r5b.16xlarge |
r5d.16xlarge | r6g.12xlarge |
r6gd.12xlarge

r5dn.4xlarge and smaller | r5n.4xlarge and Up to 25 Gbps † ENA (p. 865)

smaller

r4.16xlarge | r5.24xlarge | r5.metal | 25 Gbps ENA (p. 865)

r5b.24xlarge | r5b.metal | r5d.24xlarge
| r5d.metal | r5dn.8xlarge | r5n.8xlarge |
r6g.16xlarge | r6g.metal | r6gd.16xlarge
| r6gd.metal | x1.32xlarge | x1e.32xlarge |
z1d.12xlarge | z1d.metal

r5dn.12xlarge | r5n.12xlarge 50 Gbps ENA (p. 865)

r5dn.16xlarge | r5n.16xlarge 75 Gbps ENA (p. 865)

r5dn.24xlarge | r5n.24xlarge | 100 Gbps ENA (p. 865)

u-6tb1.metal * | u-9tb1.metal * |
u-12tb1.metal * | u-18tb1.metal |
u-24tb1.metal

* Instances of this type launched after March 12, 2020 provide network performance of 100 Gbps.
Instances of this type launched before March 12, 2020 might only provide network performance of 25
Gbps. To ensure that instances launched before March 12, 2020 have a network performance of 100
Gbps, contact your account team to upgrade your instance at no additional cost.

† These instances use a network I/O credit mechanism to allocate network bandwidth to instances based
on average bandwidth utilization. They accrue credits when their bandwidth is below their baseline
bandwidth, and can use these credits when they perform network data transfers. For more information,
open a support case and ask about baseline bandwidth for the specific instance types that you are
interested in.

SSD I/O performance

If you use a Linux AMI with kernel version 4.4 or later and use all the SSD-based instance store volumes
available to your instance, you get the IOPS (4,096 byte block size) performance listed in the following
table (at queue depth saturation). Otherwise, you get lower IOPS performance.

Instance Size 100% Random Read IOPS Write IOPS

r5ad.large * 30,000 15,000

r5ad.xlarge * 59,000 29,000

r5ad.2xlarge * 117,000 57,000

r5ad.4xlarge * 234,000 114,000

r5ad.8xlarge 466,666 233,333

r5ad.12xlarge 700,000 340,000

r5ad.16xlarge 933,333 466,666

278
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Memory optimized

Instance Size 100% Random Read IOPS Write IOPS

r5ad.24xlarge 1,400,000 680,000

r5d.large * 30,000 15,000

r5d.xlarge * 59,000 29,000

r5d.2xlarge * 117,000 57,000

r5d.4xlarge * 234,000 114,000

r5d.8xlarge 466,666 233,333

r5d.12xlarge 700,000 340,000

r5d.16xlarge 933,333 466,666

r5d.24xlarge 1,400,000 680,000

r5d.metal 1,400,000 680,000

r5dn.large * 30,000 15,000

r5dn.xlarge * 59,000 29,000

r5dn.2xlarge * 117,000 57,000

r5dn.4xlarge * 234,000 114,000

r5dn.8xlarge 466,666 233,333

r5dn.12xlarge 700,000 340,000

r5dn.16xlarge 933,333 466,666

r5dn.24xlarge 1,400,000 680,000

r6gd.medium 13,438 5,625

r6gd.large 26,875 11,250

r6gd.xlarge 53,750 22,500

r6gd.2xlarge 107,500 45,000

r6gd.4xlarge 215,000 90,000

r6gd.8xlarge 430,000 180,000

r6gd.12xlarge 645,000 270,000

r6gd.16xlarge 860,000 360,000

r6gd.metal 860,000 360,000

z1d.large * 30,000 15,000

z1d.xlarge * 59,000 29,000

z1d.2xlarge * 117,000 57,000

z1d.3xlarge * 175,000 75,000

279
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Memory optimized

Instance Size 100% Random Read IOPS Write IOPS

z1d.6xlarge 350,000 170,000

z1d.12xlarge 700,000 340,000

z1d.metal 700,000 340,000

* For these instances, you can get up to the specified performance.

As you fill the SSD-based instance store volumes for your instance, the number of write IOPS that
you can achieve decreases. This is due to the extra work the SSD controller must do to find available
space, rewrite existing data, and erase unused space so that it can be rewritten. This process of
garbage collection results in internal write amplification to the SSD, expressed as the ratio of SSD write
operations to user write operations. This decrease in performance is even larger if the write operations
are not in multiples of 4,096 bytes or not aligned to a 4,096-byte boundary. If you write a smaller
amount of bytes or bytes that are not aligned, the SSD controller must read the surrounding data and
store the result in a new location. This pattern results in significantly increased write amplification,
increased latency, and dramatically reduced I/O performance.

SSD controllers can use several strategies to reduce the impact of write amplification. One such strategy
is to reserve space in the SSD instance storage so that the controller can more efficiently manage the
space available for write operations. This is called over-provisioning. The SSD-based instance store
volumes provided to an instance don't have any space reserved for over-provisioning. To reduce write
amplification, we recommend that you leave 10% of the volume unpartitioned so that the SSD controller
can use it for over-provisioning. This decreases the storage that you can use, but increases performance
even if the disk is close to full capacity.

For instance store volumes that support TRIM, you can use the TRIM command to notify the SSD
controller whenever you no longer need data that you've written. This provides the controller with more
free space, which can reduce write amplification and increase performance. For more information, see
Instance store volume TRIM support (p. 1282).

Instance features
The following is a summary of features for memory optimized instances.

  EBS only NVMe EBS Instance store Placement group

R4 Yes No No Yes

R5 Yes Yes No Yes

R5a Yes Yes No Yes

R5ad No Yes NVME * Yes

R5b Yes Yes No Yes

R5d No Yes NVME * Yes

R5dn No Yes NVME * Yes

R5n Yes Yes No Yes

R6g Yes Yes No Yes

R6gd No Yes NVMe * Yes

280
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Memory optimized

  EBS only NVMe EBS Instance store Placement group

Yes
u-6tb1.metal Yes No No

Yes
u-9tb1.metal Yes No No

Yes
u-12tb1.metal Yes No No

Yes
u-18tb1.metal Yes No No

Yes
u-24tb1.metal Yes No No

X1 No No SSD Yes

X1e No No SSD * Yes

z1d No Yes NVME * Yes

* The root device volume must be an Amazon EBS volume.

For more information, see the following:

• Amazon EBS and NVMe on Linux instances (p. 1214)

• Amazon EC2 instance store (p. 1268)
• Placement groups (p. 927)

Support for vCPUs

Memory optimized instances provide a high number of vCPUs, which can cause launch issues with
operating systems that have a lower vCPU limit. We strongly recommend that you use the latest AMIs
when you launch memory optimized instances.

The following AMIs support launching memory optimized instances:

• Amazon Linux 2 (HVM)

• Amazon Linux AMI 2016.03 (HVM) or later
• Ubuntu Server 14.04 LTS (HVM)
• Red Hat Enterprise Linux 7.1 (HVM)
• SUSE Linux Enterprise Server 12 SP1 (HVM)
• Windows Server 2019
• Windows Server 2016
• Windows Server 2012 R2
• Windows Server 2012
• Windows Server 2008 R2 64-bit
• Windows Server 2008 SP2 64-bit

Release notes
• R4 instances feature up to 64 vCPUs and are powered by two AWS-customized Intel XEON processors
based on E5-2686v4 that feature high-memory bandwidth and larger L3 caches to boost the
performance of in-memory applications.
• R5, R5b, and R5d instances feature a 3.1 GHz Intel Xeon Platinum 8000 series processor from either
the first generation (Skylake-SP) or second generation (Cascade Lake).

281
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Memory optimized

• R5a and R5ad instances feature a 2.5 GHz AMD EPYC 7000 series processor.
• R6g and R6gd instances feature an AWS Graviton2 processor based on 64-bit Arm architecture.
• High memory instances (u-6tb1.metal, u-9tb1.metal, and u-12tb1.metal) are the first
instances to be powered by an eight-socket platform with the latest generation Intel Xeon Platinum
8176M (Skylake) processors that are optimized for mission-critical enterprise workloads. High Memory
instances with 18 TB and 24 TB of memory (u-18tb1.metal and u-24tb1.metal) are the first
instances powered by an 8-socket platform with 2nd Generation Intel Xeon Scalable 8280L (Cascade
Lake) processors.
• X1e and X1 instances feature up to 128 vCPUs and are powered by four Intel Xeon E7-8880 v3
processors that feature high-memory bandwidth and larger L3 caches to boost the performance of in-
memory applications.
• Instances built on the Nitro System have the following requirements:
• NVMe drivers (p. 1214) must be installed
• Elastic Network Adapter (ENA) drivers (p. 865) must be installed

The following Linux AMIs meet these requirements:

• Amazon Linux 2
• Amazon Linux AMI 2018.03
• Ubuntu 14.04 (with linux-aws kernel) or later
• Red Hat Enterprise Linux 7.4 or later
• SUSE Linux Enterprise Server 12 SP2 or later
• CentOS 7.4.1708 or later
• FreeBSD 11.1 or later
• Debian GNU/Linux 9 or later
• Instances with an AWS Graviton processors have the following requirements:
• Use an AMI for the 64-bit Arm architecture.
• Support booting through UEFI with ACPI tables and support ACPI hot-plug of PCI devices.

The following AMIs meet these requirements:

• Amazon Linux 2 (64-bit Arm)
• Ubuntu 16.04 or later (64-bit Arm)
• Red Hat Enterprise Linux 8.0 or later (64-bit Arm)
• SUSE Linux Enterprise Server 15 or later (64-bit Arm)
• Debian 10 or later (64-bit Arm)
• Instances built on the Nitro System instances support a maximum of 28 attachments, including
network interfaces, EBS volumes, and NVMe instance store volumes. For more information, see Nitro
System volume limits (p. 1291).
• R5b instances do not support io2 volumes or volumes with Multi-Attach enabled.
• Launching a bare metal instance boots the underlying server, which includes verifying all hardware and
firmware components. This means that it can take 20 minutes from the time the instance enters the
running state until it becomes available over the network.
• To attach or detach EBS volumes or secondary network interfaces from a bare metal instance requires
PCIe native hotplug support. Amazon Linux 2 and the latest versions of the Amazon Linux AMI
support PCIe native hotplug, but earlier versions do not. You must enable the following Linux kernel
configuration options:

CONFIG_HOTPLUG_PCI_PCIE=y
CONFIG_PCIEASPM=y

• Bare metal instances use a PCI-based serial device rather than an I/O port-based serial device. The
upstream Linux kernel and the latest Amazon Linux AMIs support this device. Bare metal instances also

282
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Storage optimized

provide an ACPI SPCR table to enable the system to automatically use the PCI-based serial device. The
latest Windows AMIs automatically use the PCI-based serial device.
• You can't launch X1 instances using a Windows Server 2008 SP2 64-bit AMI, except for x1.16xlarge
instances.
• You can't launch X1e instances using a Windows Server 2008 SP2 64-bit AMI.
• With earlier versions of the Windows Server 2008 R2 64-bit AMI, you can't launch r4.large and
r4.4xlarge instances. If you experience this issue, update to the latest version of this AMI.
• There is a limit on the total number of instances that you can launch in a Region, and there are
additional limits on some instance types. For more information, see How many instances can I run in
Amazon EC2? in the Amazon EC2 FAQ.

Storage optimized instances

Storage optimized instances are designed for workloads that require high, sequential read and write
access to very large data sets on local storage. They are optimized to deliver tens of thousands of low-
latency, random I/O operations per second (IOPS) to applications.

D2 instances

These instances are well suited for the following:

• Massive parallel processing (MPP) data warehouse

• MapReduce and Hadoop distributed computing
• Log or data processing applications

D3 and D3en instances

These instances offer scale out of instance storage and are well suited for the following:

• Distributed file systems for Hadoop workloads

• File storage workloads such as GPFC and BeeFS
• Large data lakes for HPC workloads

H1 instances

These instances are well suited for the following:

• Data-intensive workloads such as MapReduce and distributed file systems

• Applications requiring sequential access to large amounts of data on direct-attached instance storage
• Applications that require high-throughput access to large quantities of data

I3 and I3en instances

These instances are well suited for the following:

• High frequency online transaction processing (OLTP) systems

• Relational databases
• NoSQL databases
• Cache for in-memory databases (for example, Redis)

283
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Storage optimized

• Data warehousing applications

• Distributed file systems

Bare metal instances provide your applications with direct access to physical resources of the host server,
such as processors and memory.

For more information, see Amazon EC2 I3 Instances.

Contents
• Hardware specifications (p. 284)
• Instance performance (p. 285)
• Network performance (p. 286)
• SSD I/O performance (p. 287)
• Instance features (p. 288)
• Support for vCPUs (p. 288)
• Release notes (p. 289)

Hardware specifications
The primary data storage for D2, D3, and D3en instances is HDD instance store volumes. The primary
data storage for I3 and I3en instances is non-volatile memory express (NVMe) SSD instance store
volumes.

Instance store volumes persist only for the life of the instance. When you stop, hibernate, or terminate
an instance, the applications and data in its instance store volumes are erased. We recommend that you
regularly back up or replicate important data in your instance store volumes. For more information, see
Amazon EC2 instance store (p. 1268) and SSD instance store volumes (p. 1280).

The following is a summary of the hardware specifications for storage optimized instances.

Instance type Default vCPUs Memory (GiB)

d2.xlarge 4 30.5

d2.2xlarge 8 61

d2.4xlarge 16 122

d2.8xlarge 36 244

d3.xlarge 4 32

d3.2xlarge 8 64

d3.4xlarge 16 128

d3.8xlarge 32 256

d3en.large 2 8

d3en.xlarge 4 16

d3en.2xlarge 8 32

d3en.4xlarge 16 64

284
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Storage optimized

Instance type Default vCPUs Memory (GiB)

d3en.6xlarge 24 96

d3en.8xlarge 32 128

d3en.12xlarge 48 192

h1.2xlarge 8 32

h1.4xlarge 16 64

h1.8xlarge 32 128

h1.16xlarge 64 256

i3.large 2 15.25

i3.xlarge 4 30.5

i3.2xlarge 8 61

i3.4xlarge 16 122

i3.8xlarge 32 244

i3.16xlarge 64 488

i3.metal 72 512

i3en.large 2 16

i3en.xlarge 4 32

i3en.2xlarge 8 64

i3en.3xlarge 12 96

i3en.6xlarge 24 192

i3en.12xlarge 48 384

i3en.24xlarge 96 768

i3en.metal 96 768

For more information about the hardware specifications for each Amazon EC2 instance type, see Amazon
EC2 Instance Types.

For more information about specifying CPU options, see Optimize CPU options (p. 668).

Instance performance
To ensure the best disk throughput performance from your instance on Linux, we recommend that you
use the most recent version of Amazon Linux 2 or the Amazon Linux AMI.

For instances with NVMe instance store volumes, you must use a Linux AMI with kernel version 4.4 or
later. Otherwise, your instance will not achieve the maximum IOPS performance available.

D2 instances provide the best disk performance when you use a Linux kernel that supports persistent
grants, an extension to the Xen block ring protocol that significantly improves disk throughput and
scalability. For more information about persistent grants, see this article in the Xen Project Blog.

285
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Storage optimized

EBS-optimized instances enable you to get consistently high performance for your EBS volumes by
eliminating contention between Amazon EBS I/O and other network traffic from your instance. Some
storage optimized instances are EBS-optimized by default at no additional cost. For more information,
see Amazon EBS–optimized instances (p. 1217).

Some storage optimized instance types provide the ability to control processor C-states and P-states on
Linux. C-states control the sleep levels that a core can enter when it is inactive, while P-states control the
desired performance (in CPU frequency) from a core. For more information, see Processor state control
for your EC2 instance (p. 658).

Network performance
You can enable enhanced networking on supported instance types to provide lower latencies, lower
network jitter, and higher packet-per-second (PPS) performance. Most applications do not consistently
need a high level of network performance, but can benefit from access to increased bandwidth when
they send or receive data. For more information, see Enhanced networking on Linux (p. 864).

The following is a summary of network performance for storage optimized instances that support
enhanced networking.

Instance type Network performance Enhanced networking

d2.xlarge Moderate Intel 82599 VF (p. 875)

d2.2xlarge | d2.4xlarge High Intel 82599 VF (p. 875)

d3.xlarge Up to 5 Gbps † ENA (p. 865)

d3.2xlarge 5 Gbps ENA (p. 865)

i3.4xlarge and smaller Up to 10 Gbps † ENA (p. 865)

d2.8xlarge 10 Gbps Intel 82599 VF (p. 875)

d3.4xlarge | i3.8xlarge | 10 Gbps ENA (p. 865)

h1.8xlarge

d3en.2xlarge and smaller | Up to 25 Gbps † ENA (p. 865)

i3en.3xlarge and smaller

d3.8xlarge | d3en.4xlarge 25 Gbps ENA (p. 865)

| i3.16xlarge | i3.metal
| i3en.6xlarge |
h1.16xlarge

d3en.6xlarge 40 Gbps ENA (p. 865)

d3.8xlarge | d3en.8xlarge | 50 Gbps ENA (p. 865)

i3en.12xlarge

d3en.12xlarge 75 Gbps ENA (p. 865)

i3en.24xlarge | i3en.metal 100 Gbps ENA (p. 865)

† These instances use a network I/O credit mechanism to allocate network bandwidth to instances based
on average bandwidth utilization. They accrue credits when their bandwidth is below their baseline
bandwidth, and can use these credits when they perform network data transfers. For more information,
open a support case and ask about baseline bandwidth for the specific instance types that you are
interested in.

286
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Storage optimized

SSD I/O performance

If you use a Linux AMI with kernel version 4.4 or later and use all the SSD-based instance store volumes
available to your instance, you get the IOPS (4,096 byte block size) performance listed in the following
table (at queue depth saturation). Otherwise, you get lower IOPS performance.

Instance Size 100% Random Read IOPS Write IOPS

i3.large * 100,125 35,000

i3.xlarge * 206,250 70,000

i3.2xlarge 412,500 180,000

i3.4xlarge 825,000 360,000

i3.8xlarge 1.65 million 720,000

i3.16xlarge 3.3 million 1.4 million

i3.metal 3.3 million 1.4 million

i3en.large * 42,500 32,500

i3en.xlarge * 85,000 65,000

i3en.2xlarge * 170,000 130,000

i3en.3xlarge 250,000 200,000

i3en.6xlarge 500,000 400,000

i3en.12xlarge 1 million 800,000

i3en.24xlarge 2 million 1.6 million

i3en.metal 2 million 1.6 million

* For these instances, you can get up to the specified performance.

As you fill your SSD-based instance store volumes, the I/O performance that you get decreases. This is
due to the extra work that the SSD controller must do to find available space, rewrite existing data, and
erase unused space so that it can be rewritten. This process of garbage collection results in internal write
amplification to the SSD, expressed as the ratio of SSD write operations to user write operations. This
decrease in performance is even larger if the write operations are not in multiples of 4,096 bytes or not
aligned to a 4,096-byte boundary. If you write a smaller amount of bytes or bytes that are not aligned,
the SSD controller must read the surrounding data and store the result in a new location. This pattern
results in significantly increased write amplification, increased latency, and dramatically reduced I/O
performance.

SSD controllers can use several strategies to reduce the impact of write amplification. One such strategy
is to reserve space in the SSD instance storage so that the controller can more efficiently manage the
space available for write operations. This is called over-provisioning. The SSD-based instance store
volumes provided to an instance don't have any space reserved for over-provisioning. To reduce write
amplification, we recommend that you leave 10% of the volume unpartitioned so that the SSD controller
can use it for over-provisioning. This decreases the storage that you can use, but increases performance
even if the disk is close to full capacity.

For instance store volumes that support TRIM, you can use the TRIM command to notify the SSD
controller whenever you no longer need data that you've written. This provides the controller with more

287
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Storage optimized

free space, which can reduce write amplification and increase performance. For more information, see
Instance store volume TRIM support (p. 1282).

Instance features
The following is a summary of features for storage optimized instances:

  EBS only Instance store Placement group

D2 No HDD Yes

D3 No HDD * Yes

D3en No HDD * Yes

H1 No HDD * Yes

I3 No NVMe * Yes

I3en No NVMe * Yes

* The root device volume must be an Amazon EBS volume.

For more information, see the following:

• Amazon EBS and NVMe on Linux instances (p. 1214)

• Amazon EC2 instance store (p. 1268)
• Placement groups (p. 927)

Support for vCPUs

The d2.8xlarge instance type provides 36 vCPUs, which might cause launch issues in some Linux
operating systems that have a vCPU limit of 32. We strongly recommend that you use the latest AMIs
when you launch d2.8xlarge instances.

The following Linux AMIs support launching d2.8xlarge instances with 36 vCPUs:

• Amazon Linux 2 (HVM)

• Amazon Linux AMI 2018.03 (HVM)
• Ubuntu Server 14.04 LTS (HVM) or later
• Red Hat Enterprise Linux 7.1 (HVM)
• SUSE Linux Enterprise Server 12 (HVM)

If you must use a different AMI for your application, and your d2.8xlarge instance launch does not
complete successfully (for example, if your instance status changes to stopped during launch with a
Client.InstanceInitiatedShutdown state transition reason), modify your instance as described in
the following procedure to support more than 32 vCPUs so that you can use the d2.8xlarge instance
type.

To update an instance to support more than 32 vCPUs

1. Launch a D2 instance using your AMI, choosing any D2 instance type other than d2.8xlarge.

288
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Storage optimized

2. Update the kernel to the latest version by following your operating system-specific instructions. For
example, for RHEL 6, use the following command:

sudo yum update -y kernel

3. Stop the instance.

4. (Optional) Create an AMI from the instance that you can use to launch any additional d2.8xlarge
instances that you need in the future.
5. Change the instance type of your stopped instance to d2.8xlarge (choose Actions, Instance
settings, Change instance type, and then follow the directions).
6. Start the instance. If the instance launches properly, you are done. If the instance still does not boot
properly, proceed to the next step.
7. (Optional) If the instance still does not boot properly, the kernel on your instance may not support
more than 32 vCPUs. However, you may be able to boot the instance if you limit the vCPUs.

a. Change the instance type of your stopped instance to any D2 instance type other than
d2.8xlarge (choose Actions, Instance settings, Change instance type, and then follow the
directions).
b. Add the maxcpus=32 option to your boot kernel parameters by following your operating
system-specific instructions. For example, for RHEL 6, edit the /boot/grub/menu.lst file and
add the following option to the most recent and active kernel entry:

default=0
timeout=1
splashimage=(hd0,0)/boot/grub/splash.xpm.gz
hiddenmenu
title Red Hat Enterprise Linux Server (2.6.32-504.3.3.el6.x86_64)
root (hd0,0)
kernel /boot/vmlinuz-2.6.32-504.3.3.el6.x86_64 maxcpus=32 console=ttyS0 ro
root=UUID=9996863e-b964-47d3-a33b-3920974fdbd9 rd_NO_LUKS KEYBOARDTYPE=pc
KEYTABLE=us LANG=en_US.UTF-8 xen_blkfront.sda_is_xvda=1 console=ttyS0,115200n8
console=tty0 rd_NO_MD SYSFONT=latarcyrheb-sun16 crashkernel=auto rd_NO_LVM
rd_NO_DM
initrd /boot/initramfs-2.6.32-504.3.3.el6.x86_64.img

c. Stop the instance.

d. (Optional) Create an AMI from the instance that you can use to launch any additional
d2.8xlarge instances that you need in the future.
e. Change the instance type of your stopped instance to d2.8xlarge (choose Actions, Instance
Settings, Change Instance Type, and then follow the directions).
f. Start the instance.

Release notes
• You must launch storage optimized instances using an HVM AMI. For more information, see Linux AMI
virtualization types (p. 104).
• Instances built on the Nitro System (p. 209) have the following requirements:
• NVMe drivers (p. 1214) must be installed
• Elastic Network Adapter (ENA) drivers (p. 865) must be installed

The following Linux AMIs meet these requirements:

• Amazon Linux 2
• Amazon Linux AMI 2018.03
• Ubuntu 14.04 (with linux-aws kernel) or later

289
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

• Red Hat Enterprise Linux 7.4 or later

• SUSE Linux Enterprise Server 12 SP2 or later
• CentOS 7.4.1708 or later
• FreeBSD 11.1 or later
• Debian GNU/Linux 9 or later
• Launching a bare metal instance boots the underlying server, which includes verifying all hardware and
firmware components. This means that it can take 20 minutes from the time the instance enters the
running state until it becomes available over the network.
• To attach or detach EBS volumes or secondary network interfaces from a bare metal instance requires
PCIe native hotplug support. Amazon Linux 2 and the latest versions of the Amazon Linux AMI
support PCIe native hotplug, but earlier versions do not. You must enable the following Linux kernel
configuration options:

CONFIG_HOTPLUG_PCI_PCIE=y
CONFIG_PCIEASPM=y

• Bare metal instances use a PCI-based serial device rather than an I/O port-based serial device. The
upstream Linux kernel and the latest Amazon Linux AMIs support this device. Bare metal instances also
provide an ACPI SPCR table to enable the system to automatically use the PCI-based serial device. The
latest Windows AMIs automatically use the PCI-based serial device.
• With FreeBSD AMIs, bare metal instances take nearly an hour to boot and I/O to the local NVMe
storage does not complete. As a workaround, add the following line to /boot/loader.conf and
reboot:

hw.nvme.per_cpu_io_queues="0"

• The d2.8xlarge instance type has 36 vCPUs, which might cause launch issues in some
Linux operating systems that have a vCPU limit of 32. For more information, see Support for
vCPUs (p. 288).
• The d3.8xlarge and d3en.12xlarge instances support a maximum of three attachments, including
the root volume. If you exceed the attachment limit when you add a network interface or EBS volume,
this causes attachment issues on your instance.
• There is a limit on the total number of instances that you can launch in a Region, and there are
additional limits on some instance types. For more information, see How many instances can I run in
Amazon EC2? in the Amazon EC2 FAQ.

Linux accelerated computing instances

Accelerated computing instances use hardware accelerators, or co-processors, to perform some
functions, such as floating point number calculations, graphics processing, or data pattern matching,
more efficiently than is possible in software running on CPUs. These instances enable more parallelism
for higher throughput on compute-intensive workloads.

If you require high processing capability, you'll benefit from using accelerated computing instances,
which provide access to hardware-based compute accelerators such as Graphics Processing Units (GPUs),
Field Programmable Gate Arrays (FPGAs), or AWS Inferentia.

Contents
• GPU instances (p. 291)
• Instances with AWS Inferentia (p. 292)
• FPGA instances (p. 293)
• Hardware specifications (p. 293)

290
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

• Instance performance (p. 294)

• Network performance (p. 295)
• Instance features (p. 295)
• Release notes (p. 296)
• Install NVIDIA drivers on Linux instances (p. 297)
• Install AMD drivers on Linux instances (p. 304)
• Activate NVIDIA GRID Virtual Applications (p. 308)
• Optimize GPU settings (p. 309)

GPU instances
GPU-based instances provide access to NVIDIA GPUs with thousands of compute cores. You can use
these instances to accelerate scientific, engineering, and rendering applications by leveraging the
CUDA or Open Computing Language (OpenCL) parallel computing frameworks. You can also use them
for graphics applications, including game streaming, 3-D application streaming, and other graphics
workloads.

G4ad and G4dn instances

G4ad instances use AMD Radeon Pro V520 GPUs and 2nd generation AMD EPYC processors, and are well-
suited for graphics applications such as remote graphics workstations, game streaming, and rendering
that leverage industry-standard APIs such as OpenGL, DirectX, and Vulkan. They provide up to 4 AMD
Radeon Pro V520 GPUs, 64 vCPUs, 25 Gbps networking, and 2.4 TB local NVMe-based SSD storage.

G4dn instances use NVIDIA Tesla GPUs and provide a cost-effective, high-performance platform for
general purpose GPU computing using the CUDA or machine learning frameworks along with graphics
applications using DirectX or OpenGL. These instances provide high- bandwidth networking, powerful
half and single-precision floating-point capabilities, along with INT8 and INT4 precisions. Each GPU has
16 GiB of GDDR6 memory, making G4dn instances well-suited for machine learning inference, video
transcoding, and graphics applications like remote graphics workstations and game streaming in the
cloud.

For more information, see Amazon EC2 G4 Instances.

G4dn instances support NVIDIA GRID Virtual Workstation. For more information, see NVIDIA Marketplace
offerings.

G3 instances

These instances use NVIDIA Tesla M60 GPUs and provide a cost-effective, high-performance platform
for graphics applications using DirectX or OpenGL. G3 instances also provide NVIDIA GRID Virtual
Workstation features, such as support for four monitors with resolutions up to 4096x2160, and NVIDIA
GRID Virtual Applications. G3 instances are well-suited for applications such as 3D visualizations,
graphics-intensive remote workstations, 3D rendering, video encoding, virtual reality, and other server-
side graphics workloads requiring massively parallel processing power.

For more information, see Amazon EC2 G3 Instances.

G3 instances support NVIDIA GRID Virtual Workstation and NVIDIA GRID Virtual Applications. To activate
either of these features, see Activate NVIDIA GRID Virtual Applications (p. 308).

G2 instances

These instances use NVIDIA GRID K520 GPUs and provide a cost-effective, high-performance platform
for graphics applications using DirectX or OpenGL. NVIDIA GRID GPUs also support NVIDIA’s fast capture
and encode API operations. Example applications include video creation services, 3D visualizations,
streaming graphics-intensive applications, and other server-side graphics workloads.

291
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

P4d instances

These instances use NVIDIA A100 GPUs and provide a high-performance platform for machine learning
and HPC workloads. P4d instances offer 400 Gbps of aggregate network bandwidth throughput and
support, Elastic Fabric Adapter (EFA). They are the first EC2 instances to provide multiple network cards.

For more information, see Amazon EC2 P4d Instances.

P4d instances support NVIDIA NVSwitch GPU interconnect and NVIDIA GPUDirect RDMA.

P3 instances

These instances use NVIDIA Tesla V100 GPUs and are designed for general purpose GPU computing
using the CUDA or OpenCL programming models or through a machine learning framework. P3
instances provide high-bandwidth networking, powerful half, single, and double-precision floating-
point capabilities, and up to 32 GiB of memory per GPU, which makes them ideal for deep learning,
computational fluid dynamics, computational finance, seismic analysis, molecular modeling, genomics,
rendering, and other server-side GPU compute workloads. Tesla V100 GPUs do not support graphics
mode.

For more information, see Amazon EC2 P3 Instances.

P3 instances support NVIDIA NVLink peer to peer transfers. For more information, see NVIDIA NVLink.

P2 instances

P2 instances use NVIDIA Tesla K80 GPUs and are designed for general purpose GPU computing using
the CUDA or OpenCL programming models. P2 instances provide high-bandwidth networking, powerful
single and double precision floating-point capabilities, and 12 GiB of memory per GPU, which makes
them ideal for deep learning, graph databases, high-performance databases, computational fluid
dynamics, computational finance, seismic analysis, molecular modeling, genomics, rendering, and other
server-side GPU compute workloads.

P2 instances support NVIDIA GPUDirect peer to peer transfers. For more information, see NVIDIA
GPUDirect.

Instances with AWS Inferentia

These instances are designed to accelerate machine learning using AWS Inferentia, a custom AI/ML
chip from Amazon that provides high performance and low latency machine learning inference. These
instances are optimized for deploying Deep Learning (DL) models for applications, such as natural
language processing, object detection and classification, content personalization and filtering, and
speech recognition.

There are a variety of ways that you can get started:

• Use SageMaker, a fully-managed service that is the easiest way to get started with machine learning
models. For more information, see Compile and deploy a TensorFlow model on Inf1 instances on
github.
• Launch an Inf1 instance using the Deep Learning AMI. For more information, see AWS Inferentia with
DLAMI in the AWS Deep Learning AMI Developer Guide.
• Launch an Inf1 instance using your own AMI and install the AWS Neuron SDK, which enables you to
compile, run, and profile deep learning models for AWS Inferentia.
• Launch a container instance using an Inf1 instance and an Amazon ECS-optimized AMI. For more
information, see Amazon Linux 2 (Inferentia) AMIs in the Amazon Elastic Container Service Developer
Guide.
• Create an Amazon EKS cluster with nodes running Inf1 instances. For more information, see Inferentia
support in the Amazon EKS User Guide.

292
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

For more information, see Machine Learning on AWS.

Inf1 instances

Inf1 instances use AWS Inferentia machine learning inference chips. Inferentia was developed to enable
highly cost-effective low latency inference performance at any scale.

For more information, see Amazon EC2 Inf1 Instances.

FPGA instances
FPGA-based instances provide access to large FPGAs with millions of parallel system logic cells. You can
use FPGA-based accelerated computing instances to accelerate workloads such as genomics, financial
analysis, real-time video processing, big data analysis, and security workloads by leveraging custom
hardware accelerations. You can develop these accelerations using hardware description languages such
as Verilog or VHDL, or by using higher-level languages such as OpenCL parallel computing frameworks.
You can either develop your own hardware acceleration code or purchase hardware accelerations through
the AWS Marketplace.

The FPGA Developer AMI provides the tools for developing, testing, and building AFIs. You can use the
FPGA Developer AMI on any EC2 instance with at least 32 GB of system memory (for example, C5, M4,
and R4 instances).

For more information, see the documentation for the AWS FPGA Hardware Development Kit.

F1 instances

F1 instances use Xilinx UltraScale+ VU9P FPGAs and are designed to accelerate computationally
intensive algorithms, such as data-flow or highly parallel operations not suited to general purpose
CPUs. Each FPGA in an F1 instance contains approximately 2.5 million logic elements and approximately
6,800 Digital Signal Processing (DSP) engines, along with 64 GiB of local DDR ECC protected memory,
connected to the instance by a dedicated PCIe Gen3 x16 connection. F1 instances provide local NVMe
SSD volumes.

Developers can use the FPGA Developer AMI and AWS Hardware Developer Kit to create custom
hardware accelerations for use on F1 instances. The FPGA Developer AMI includes development tools for
full-cycle FPGA development in the cloud. Using these tools, developers can create and share Amazon
FPGA Images (AFIs) that can be loaded onto the FPGA of an F1 instance.

For more information, see Amazon EC2 F1 Instances.

Hardware specifications
The following is a summary of the hardware specifications for accelerated computing instances.

Instance type Default vCPUs Memory (GiB) Accelerators

p2.xlarge 4 61 1

p2.8xlarge 32 488 8

p2.16xlarge 64 732 16

p3.2xlarge 8 61 1

p3.8xlarge 32 244 4

p3.16xlarge 64 488 8

p3dn.24xlarge 96 768 8

293
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

Instance type Default vCPUs Memory (GiB) Accelerators

p4d.24xlarge 96 1,152 8

g2.2xlarge 8 15 1

g2.8xlarge 32 60 4

g3s.xlarge 4 30.5 1

g3.4xlarge 16 122 1

g3.8xlarge 32 244 2

g3.16xlarge 64 488 4

g4ad.4xlarge 16 64 1

g4ad.8xlarge 32 128 2

g4ad.16xlarge 64 256 4

g4dn.xlarge 4 16 1

g4dn.2xlarge 8 32 1

g4dn.4xlarge 16 64 1

g4dn.8xlarge 32 128 1

g4dn.12xlarge 48 192 4

g4dn.16xlarge 64 256 1

g4dn.metal 96 384 8

f1.2xlarge 8 122 1

f1.4xlarge 16 244 2

f1.16xlarge 64 976 8

inf1.xlarge 4 8 1

inf1.2xlarge 8 16 1

inf1.6xlarge 24 48 4

inf1.24xlarge 96 192 16

For more information about the hardware specifications for each Amazon EC2 instance type, see Amazon
EC2 Instance Types.

For more information about specifying CPU options, see Optimize CPU options (p. 668).

Instance performance
There are several GPU setting optimizations that you can perform to achieve the best performance on
your instances. For more information, see Optimize GPU settings (p. 309).

EBS-optimized instances enable you to get consistently high performance for your EBS volumes
by eliminating contention between Amazon EBS I/O and other network traffic from your instance.

294
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

Some accelerated computing instances are EBS-optimized by default at no additional cost. For more
information, see Amazon EBS–optimized instances (p. 1217).

Some accelerated computing instance types provide the ability to control processor C-states and P-
states on Linux. C-states control the sleep levels that a core can enter when it is inactive, while P-states
control the desired performance (in CPU frequency) from a core. For more information, see Processor
state control for your EC2 instance (p. 658).

Network performance
You can enable enhanced networking on supported instance types to provide lower latencies, lower
network jitter, and higher packet-per-second (PPS) performance. Most applications do not consistently
need a high level of network performance, but can benefit from access to increased bandwidth when
they send or receive data. For more information, see Enhanced networking on Linux (p. 864).

The following is a summary of network performance for accelerated computing instances that support
enhanced networking.

Instance type Network performance Enhanced networking

f1.4xlarge and smaller | Up to 10 Gbps † ENA (p. 865)

g3.4xlarge | g4ad.4xlarge |
p3.2xlarge

g3s.xlarge | g3.8xlarge | 10 Gbps ENA (p. 865)

p2.8xlarge | p3.8xlarge

g4ad.8xlarge 15 Gbps ENA (p. 865)

g4dn.4xlarge and smaller | Up to 25 Gbps † ENA (p. 865)

inf1.2xlarge and smaller

f1.16xlarge | g3.16xlarge 25 Gbps ENA (p. 865)

| g4ad.16xlarge |
inf1.6xlarge | p2.16xlarge
| p3.16xlarge

g4dn.8xlarge | 50 Gbps ENA (p. 865)

g4dn.12xlarge |
g4dn.16xlarge

g4dn.metal | inf1.24xlarge 100 Gbps ENA (p. 865)

| p3dn.24xlarge

p4d.24xlarge 4x100 Gbps ENA (p. 865)

† These instances use a network I/O credit mechanism to allocate network bandwidth to instances based
on average bandwidth utilization. They accrue credits when their bandwidth is below their baseline
bandwidth, and can use these credits when they perform network data transfers. For more information,
open a support case and ask about baseline bandwidth for the specific instance types that you are
interested in.

Instance features
The following is a summary of features for accelerated computing instances.

295
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

  EBS only NVMe EBS Instance store Placement group

G2 No No SSD Yes

G3 Yes No No Yes

G4ad No Yes NVMe * Yes

G4dn No Yes NVMe * Yes

Inf1 Yes No No Yes

P2 Yes No No Yes

P3 24xlarge: No 24xlarge: Yes 24xlarge: NVMe * Yes

All other sizes: Yes All other sizes: No

P4d No Yes NVMe * Yes

F1 No No NVMe * Yes

* The root device volume must be an Amazon EBS volume.

For more information, see the following:

• Amazon EBS and NVMe on Linux instances (p. 1214)

• Amazon EC2 instance store (p. 1268)
• Placement groups (p. 927)

Release notes
• You must launch the instance using an HVM AMI.
• Instances built on the Nitro System (p. 209) have the following requirements:
• NVMe drivers (p. 1214) must be installed
• Elastic Network Adapter (ENA) drivers (p. 865) must be installed

The following Linux AMIs meet these requirements:

• Amazon Linux 2
• Amazon Linux AMI 2018.03
• Ubuntu 14.04 (with linux-aws kernel) or later
• Red Hat Enterprise Linux 7.4 or later
• SUSE Linux Enterprise Server 12 SP2 or later
• CentOS 7.4.1708 or later
• FreeBSD 11.1 or later
• Debian GNU/Linux 9 or later
• GPU-based instances can't access the GPU unless the NVIDIA drivers are installed. For more
information, see Install NVIDIA drivers on Linux instances (p. 297).
• Launching a bare metal instance boots the underlying server, which includes verifying all hardware and
firmware components. This means that it can take 20 minutes from the time the instance enters the
running state until it becomes available over the network.
• To attach or detach EBS volumes or secondary network interfaces from a bare metal instance requires
PCIe native hotplug support. Amazon Linux 2 and the latest versions of the Amazon Linux AMI

296
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

support PCIe native hotplug, but earlier versions do not. You must enable the following Linux kernel
configuration options:

CONFIG_HOTPLUG_PCI_PCIE=y
CONFIG_PCIEASPM=y

• Bare metal instances use a PCI-based serial device rather than an I/O port-based serial device. The
upstream Linux kernel and the latest Amazon Linux AMIs support this device. Bare metal instances also
provide an ACPI SPCR table to enable the system to automatically use the PCI-based serial device. The
latest Windows AMIs automatically use the PCI-based serial device.
• There is a limit of 100 AFIs per Region.
• There is a limit on the total number of instances that you can launch in a Region, and there are
additional limits on some instance types. For more information, see How many instances can I run in
Amazon EC2? in the Amazon EC2 FAQ.

Install NVIDIA drivers on Linux instances

An instance with an attached NVIDIA GPU, such as a P3 or G4dn instance, must have the appropriate
NVIDIA driver installed. Depending on the instance type, you can either download a public NVIDIA driver,
download a driver from Amazon S3 that is available only to AWS customers, or use an AMI with the
driver pre-installed.

To install AMD drivers on an instance with an attached AMD GPU, such as a G4ad instance, see Install
AMD drivers on Linux instances (p. 304) instead.

Contents
• Types of NVIDIA drivers (p. 297)
• Available drivers by instance type (p. 298)
• Installation options (p. 298)
• Option 1: AMIs with the NVIDIA drivers installed (p. 298)
• Option 2: Public NVIDIA drivers (p. 299)
• Option 3: GRID drivers (G3 and G4dn instances) (p. 299)
• Option 4: NVIDIA gaming drivers (G4dn instances) (p. 301)
• Install an additional version of CUDA (p. 304)

Types of NVIDIA drivers

The following are the main types of NVIDIA drivers that can be used with GPU-based instances.

Tesla drivers

These drivers are intended primarily for compute workloads, which use GPUs for computational
tasks such as parallelized floating-point calculations for machine learning and fast Fourier
transforms for high performance computing applications.
GRID drivers

These drivers are certified to provide optimal performance for professional visualization applications
that render content such as 3D models or high-resolution videos. You can configure GRID drivers to
support two modes. Quadro Virtual Workstations provide access to four 4K displays per GPU. GRID
vApps provide RDSH App hosting capabilities.
Gaming drivers

These drivers contain optimizations for gaming and are updated frequently to provide performance
enhancements. They support a single 4K display per GPU.

297
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

NVIDIA control panel

The NVIDIA control panel is supported with GRID and Gaming drivers. It is not supported with Tesla
drivers.

Supported APIs for Tesla, GRID, and gaming drivers

• OpenCL, OpenGL, and Vulkan

• NVIDIA CUDA and related libraries (for example, cuDNN, TensorRT, nvJPEG, and cuBLAS)
• NVENC for video encoding and NVDEC for video decoding

Available drivers by instance type

The following table summarizes the supported NVIDIA drivers for each GPU instance type.

Instance type Tesla driver GRID driver Gaming driver

G2 No Yes No

G3 Yes Yes No

G4dn Yes Yes Yes

P2 Yes No No

P3 Yes Yes † No

P4d Yes No No

† Using Marketplace AMIs only

Installation options
Use one of the following options to get the NVIDIA drivers required for your GPU instance.

Options
• Option 1: AMIs with the NVIDIA drivers installed (p. 298)
• Option 2: Public NVIDIA drivers (p. 299)
• Option 3: GRID drivers (G3 and G4dn instances) (p. 299)
• Option 4: NVIDIA gaming drivers (G4dn instances) (p. 301)

Option 1: AMIs with the NVIDIA drivers installed

AWS and NVIDIA offer different Amazon Machine Images (AMI) that come with the NVIDIA drivers
installed.

• Marketplace offerings with the Tesla driver

• Marketplace offerings with the GRID driver
• Marketplace offerings with the Gaming driver

To update the driver version installed using one of these AMIs, you must uninstall the NVIDIA packages
from your instance to avoid version conflicts. Use this command to uninstall the NVIDIA packages:

298
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

[ec2-user ~]$ sudo yum erase nvidia cuda

The CUDA toolkit package has dependencies on the NVIDIA drivers. Uninstalling the NVIDIA packages
erases the CUDA toolkit. You must reinstall the CUDA toolkit after installing the NVIDIA driver.

Option 2: Public NVIDIA drivers

The options offered by AWS come with the necessary license for the driver. Alternatively, you can install
the public drivers and bring your own license. To install a public driver, download it from the NVIDIA site
as described here.

Alternatively, you can use the options offered by AWS instead of the public drivers. To use a GRID driver
on a P3 instance, use the AWS Marketplace AMIs as described in Option 1 (p. 298). To use a GRID driver
on a G3 or G4dn instance, use the AWS Marketplace AMIs, as described in Option 1 or install the NVIDIA
drivers provided by AWS as described in Option 3 (p. 299).

To download a public NVIDIA driver

Log on to your Linux instance and download the 64-bit NVIDIA driver appropriate for the instance type
from http://www.nvidia.com/Download/Find.aspx. For Product Type, Product Series, and Product, use
the options in the following table.

Instance Product Type Product Series Product

G2 GRID GRID Series GRID K520

G3 Tesla M-Class M60

G4dn † Tesla T-Series T4

P2 Tesla K-Series K80

P3 Tesla V-Series V100

P4d Tesla A-Series A100

† G4dn instances require driver version 418.87 or later.

To install the NVIDIA driver on Linux

For more information about installing and configuring the driver, see the NVIDIA Driver Installation
Quickstart Guide.

Option 3: GRID drivers (G3 and G4dn instances)

These downloads are available to AWS customers only. By downloading, you agree to use the
downloaded software only to develop AMIs for use with the NVIDIA Tesla T4 or NVIDIA Tesla M60
hardware. Upon installation of the software, you are bound by the terms of the NVIDIA GRID Cloud End
User License Agreement.

Prerequisites

• Install the AWS CLI on your Linux instance and configure default credentials. For more information, see
Installing the AWS CLI in the AWS Command Line Interface User Guide.
• IAM users must have the permissions granted by the AmazonS3ReadOnlyAccess policy.

299
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

To install the NVIDIA GRID driver on your Linux instance

1. Connect to your Linux instance. Install gcc and make, if they are not already installed.
2. Update your package cache and get the package updates for your instance.

• For Amazon Linux, CentOS, and Red Hat Enterprise Linux:

[ec2-user ~]$ sudo yum update -y

• For Ubuntu and Debian:

$ sudo apt-get update -y

3. (Ubuntu 16.04 and later, with the linux-aws package) Upgrade the linux-aws package to receive
the latest version.

$ sudo apt-get upgrade -y linux-aws

4. Reboot your instance to load the latest kernel version.

[ec2-user ~]$ sudo reboot

5. Reconnect to your instance after it has rebooted.

6. Install the gcc compiler and the kernel headers package for the version of the kernel you are
currently running.

• For Amazon Linux, CentOS, and Red Hat Enterprise Linux:

[ec2-user ~]$ sudo yum install -y gcc kernel-devel-$(uname -r)

• For Ubuntu and Debian:

$ sudo apt-get install -y gcc make linux-headers-$(uname -r)

7. [CentOS, Red Hat Enterprise Linux, Ubuntu, Debian] Disable the nouveau open source driver for
NVIDIA graphics cards.

a. Add nouveau to the /etc/modprobe.d/blacklist.conf blacklist file. Copy the following

code block and paste it into a terminal.

[ec2-user ~]$ cat << EOF | sudo tee --append /etc/modprobe.d/blacklist.conf

blacklist vga16fb
blacklist nouveau
blacklist rivafb
blacklist nvidiafb
blacklist rivatv
EOF

b. Edit the /etc/default/grub file and add the following line:

GRUB_CMDLINE_LINUX="rdblacklist=nouveau"

c. Rebuild the Grub configuration.

• For CentOS and Red Hat Enterprise Linux:

[ec2-user ~]$ sudo grub2-mkconfig -o /boot/grub2/grub.cfg

300
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

• For Ubuntu and Debian:

$ sudo update-grub

8. Download the GRID driver installation utility using the following command:

[ec2-user ~]$ aws s3 cp --recursive s3://ec2-linux-nvidia-drivers/latest/ .

Multiple versions of the GRID driver are stored in this bucket. You can see all of the available
versions using the following command.

[ec2-user ~]$ aws s3 ls --recursive s3://ec2-linux-nvidia-drivers/

Starting with GRID version 11.0, you can use the driver packages under latest for both G3 and
G4dn instances. We will not add versions later than 11.0 to g4/latest, but will keep version 11.0
and the earlier versions specific to G4dn under g4/latest.
9. Add permissions to run the driver installation utility using the following command.

[ec2-user ~]$ chmod +x NVIDIA-Linux-x86_64*.run

10. Run the self-install script as follows to install the GRID driver that you downloaded. For example:

[ec2-user ~]$ sudo /bin/sh ./NVIDIA-Linux-x86_64*.run

When prompted, accept the license agreement and specify the installation options as required (you
can accept the default options).
11. Reboot the instance.

[ec2-user ~]$ sudo reboot

12. Confirm that the driver is functional. The response for the following command lists the installed
version of the NVIDIA driver and details about the GPUs.

[ec2-user ~]$ nvidia-smi -q | head

13. (Optional) Depending on your use case, you might complete the following optional steps. If you do
not require this functionality, do not complete these steps.

a. To help take advantage of the four displays of up to 4K resolution, set up the high-performance
display protocol NICE DCV.
b. NVIDIA Quadro Virtual Workstation mode is enabled by default. To activate GRID Virtual
Applications for RDSH Application hosting capabilities, complete the GRID Virtual Application
activation steps in Activate NVIDIA GRID Virtual Applications (p. 308).

Option 4: NVIDIA gaming drivers (G4dn instances)

These drivers are available to AWS customers only. By downloading them, you agree to use the
downloaded software only to develop AMIs for use with the NVIDIA Tesla T4 hardware. Upon installation
of the software, you are bound by the terms of the NVIDIA GRID Cloud End User License Agreement.

Prerequisites

• Install the AWS CLI on your Linux instance and configure default credentials. For more information, see
Installing the AWS CLI in the AWS Command Line Interface User Guide.

301
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

• IAM users must have the permissions granted by the AmazonS3ReadOnlyAccess policy.

To install the NVIDIA gaming driver on your Linux instance

1. Connect to your Linux instance. Install gcc and make, if they are not already installed.
2. Update your package cache and get the package updates for your instance.

• For Amazon Linux, CentOS, and Red Hat Enterprise Linux:

[ec2-user ~]$ sudo yum update -y

• For Ubuntu and Debian:

$ sudo apt-get update -y

3. (Ubuntu 16.04 and later, with the linux-aws package) Upgrade the linux-aws package to receive
the latest version.

$ sudo apt-get upgrade -y linux-aws

4. Reboot your instance to load the latest kernel version.

[ec2-user ~]$ sudo reboot

5. Reconnect to your instance after it has rebooted.

6. Install the gcc compiler and the kernel headers package for the version of the kernel you are
currently running.

• For Amazon Linux, CentOS, and Red Hat Enterprise Linux:

[ec2-user ~]$ sudo yum install -y gcc kernel-devel-$(uname -r)

• For Ubuntu and Debian:

$ sudo apt-get install -y gcc make linux-headers-$(uname -r)

7. [CentOS, Red Hat Enterprise Linux, Ubuntu, Debian] Disable the nouveau open source driver for
NVIDIA graphics cards.

a. Add nouveau to the /etc/modprobe.d/blacklist.conf blacklist file. Copy the following

code block and paste it into a terminal.

[ec2-user ~]$ cat << EOF | sudo tee --append /etc/modprobe.d/blacklist.conf

blacklist vga16fb
blacklist nouveau
blacklist rivafb
blacklist nvidiafb
blacklist rivatv
EOF

b. Edit the /etc/default/grub file and add the following line:

GRUB_CMDLINE_LINUX="rdblacklist=nouveau"

c. Rebuild the Grub configuration.

• For CentOS and Red Hat Enterprise Linux:

302
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

[ec2-user ~]$ sudo grub2-mkconfig -o /boot/grub2/grub.cfg

• For Ubuntu and Debian:

$ sudo update-grub

8. Download the gaming driver installation utility using the following command:

[ec2-user ~]$ aws s3 cp --recursive s3://nvidia-gaming/linux/latest/ .

Multiple versions of the gaming driver are stored in this bucket. You can see all of the available
versions using the following command:

[ec2-user ~]$ aws s3 ls --recursive s3://nvidia-gaming/linux/

9. Add permissions to run the driver installation utility using the following command.

[ec2-user ~]$ chmod +x NVIDIA-Linux-x86_64*.run

10. Run the installer using the following command:

[ec2-user ~]$ sudo ./NVIDIA-Linux-x86_64*.run

When prompted, accept the license agreement and specify the installation options as required (you
can accept the default options).
11. Use the following command to create the required configuration file.

[ec2-user ~]$ cat << EOF | sudo tee -a /etc/nvidia/gridd.conf

vGamingMarketplace=2
EOF

12. Use the following command to download and rename the certification file.
• For version 440.68 or later:

[ec2-user ~]$ sudo curl -o /etc/nvidia/GridSwCert.txt "https://nvidia-

gaming.s3.amazonaws.com/GridSwCert-Archive/GridSwCert-Linux_2020_04.cert"

• For earlier versions:

[ec2-user ~]$ sudo curl -o /etc/nvidia/GridSwCert.txt "https://nvidia-

gaming.s3.amazonaws.com/GridSwCert-Archive/GridSwCert-Linux_2019_09.cert"

13. Reboot the instance.

[ec2-user ~]$ sudo reboot

14. (Optional) To help take advantage of a single display of up to 4K resolution, set up the high-
performance display protocol NICE DCV. If you do not require this functionality, do not complete this
step.

303
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

Install an additional version of CUDA

After you install an NVIDIA graphics driver on your instance, you can install a version of CUDA other
than the version that is bundled with the graphics driver. The following procedure demonstrates how to
configure multiple versions of CUDA on the instance.

To install the CUDA toolkit

1. Connect to your Linux instance.

2. Open the NVIDIA website and select the version of CUDA that you need.
3. Select the architecture, distribution, and version for the operating system on your instance. For
Installer Type, select runfile (local).
4. Follow the instructions to download the install script.
5. Add run permissions to the install script that you downloaded using the following command.

[ec2-user ~]$ chmod +x downloaded_installer_file

6. Run the install script as follows to install the CUDA toolkit and add the CUDA version number to the
toolkit path.

[ec2-user ~]$ sudo downloaded_installer_file --silent --override --toolkit --samples --

toolkitpath=/usr/local/cuda-version --samplespath=/usr/local/cuda --no-opengl-lib

7. (Optional) Set the default CUDA version as follows.

[ec2-user ~]$ ln -s /usr/local/cuda-version /usr/local/cuda

Install AMD drivers on Linux instances

An instance with an attached AMD GPU, such as a G4ad instance, must have the appropriate AMD driver
installed. Depending on your requirements, you can either use an AMI with the driver preinstalled or
download a driver from Amazon S3.

To install NVIDIA drivers on an instance with an attached NVIDIA GPU, such as a G4dn instance, see
Install NVIDIA drivers on Linux instances (p. 297) instead.

Contents
• AMD Radeon Pro Software for Enterprise Driver (p. 304)
• AMIs with the AMD driver installed (p. 305)
• AMD driver download (p. 305)
• Set up an interactive desktop (p. 306)

AMD Radeon Pro Software for Enterprise Driver

The AMD Radeon Pro Software for Enterprise Driver is built to deliver support for professional-grade
graphics use cases. Using the driver, you can configure your instances with two 4K displays per GPU.

Supported APIs

• OpenGL, OpenCL
• Vulkan

304
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

• AMD Advanced Media Framework

• Video Acceleration API

AMIs with the AMD driver installed

AWS offers different Amazon Machine Images (AMI) that come with the AMD drivers installed. Open
Marketplace offerings with the AMD driver.

AMD driver download

If you aren't using an AMI with the AMD driver installed, you can download the AMD driver and install it
on your instance.

These downloads are available to AWS customers only. By downloading, you agree to use the
downloaded software only to develop AMIs for use with the AMD Radeon Pro V520 hardware. Upon
installation of the software, you are bound by the terms of the AMD Software End User License
Agreement.

Prerequisites

• Install the AWS CLI on your Linux instance and configure default credentials. For more information, see
Installing the AWS CLI in the AWS Command Line Interface User Guide.
• IAM users must have the permissions granted by the AmazonS3ReadOnlyAccess policy.

To install the AMD driver on your Linux instance

1. Connect to your Linux instance. Install gcc and make, if they are not already installed.
2. Update your package cache and get the package updates for your instance.

• For Amazon Linux 2:

$ sudo amazon-linux-extras install epel -y

$ sudo yum update -y

• For Ubuntu:

$ sudo dpkg --add-architecture i386

$ sudo apt-get update -y && sudo apt upgrade -y

• For CentOS:

$ sudo yum install epel-release -y

$ sudo yum update -y

3. Reboot the instance.

$ sudo reboot

4. Reconnect to the instance after it reboots.

5. Download the latest AMD driver.

$ aws s3 cp --recursive s3://ec2-amd-linux-drivers/latest/ .

6. Extract the file.

305
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

• For Amazon Linux 2 and CentOS:

$ tar -xf amdgpu-pro-*rhel*.tar.xz

• For Ubuntu:

$ tar -xf amdgpu-pro-*ubuntu*.tar-1.xz

7. Change to the folder for the extracted driver.

8. Add the GPG keys for the driver installation.

• For Amazon Linux 2 and CentOS:

$ sudo rpm --import RPM-GPG-KEY-amdgpu

• For Ubuntu:

$ cat RPM-GPG-KEY-amdgpu | sudo apt-key add -

9. Run the self install script to install the full graphics stack.

$ ./amdgpu-pro-install -y --opencl=pal,legacy

10. Reboot the instance.

$ sudo reboot

11. Confirm that the driver is functional.

$ dmesg | grep amdgpu

The response should look like the following:

Initialized amdgpu

Set up an interactive desktop

After you confirm that your instance has the AMD GPU driver installed and amdgpu is in use, you can
install an interactive desktop manager. We recommend the MATE desktop environment for the best
compatibility and performance.

Prerequisite

Open a text editor and save the following as a file named xorg.conf. You'll need this file on your
instance.

Section "ServerLayout"
Identifier "Layout0"
Screen 0 "Screen0"
InputDevice "Keyboard0" "CoreKeyboard"
InputDevice "Mouse0" "CorePointer"
EndSection
Section "Files"
ModulePath "/opt/amdgpu/lib64/xorg/modules/drivers"
ModulePath "/opt/amdgpu/lib/xorg/modules"

306
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

ModulePath "/opt/amdgpu-pro/lib/xorg/modules/extensions"
ModulePath "/opt/amdgpu-pro/lib64/xorg/modules/extensions"
ModulePath "/usr/lib64/xorg/modules"
ModulePath "/usr/lib/xorg/modules"
EndSection
Section "InputDevice"
# generated from default
Identifier "Mouse0"
Driver "mouse"
Option "Protocol" "auto"
Option "Device" "/dev/psaux"
Option "Emulate3Buttons" "no"
Option "ZAxisMapping" "4 5"
EndSection
Section "InputDevice"
# generated from default
Identifier "Keyboard0"
Driver "kbd"
EndSection
Section "Monitor"
Identifier "Monitor0"
VendorName "Unknown"
ModelName "Unknown"
EndSection
Section "Device"
Identifier "Device0"
Driver "amdgpu"
VendorName "AMD"
BoardName "Radeon MxGPU V520"
BusID "PCI:0:30:0"
EndSection
Section "Extensions"
Option "DPMS" "Disable"
EndSection
Section "Screen"
Identifier "Screen0"
Device "Device0"
Monitor "Monitor0"
DefaultDepth 24
Option "AllowEmptyInitialConfiguration" "True"
SubSection "Display"
Virtual 3840 2160
Depth 32
EndSubSection
EndSection

To set up an interactive desktop on Amazon Linux 2

1. Install the EPEL repository.

$ sudo amazon-linux-extras install epel -y

2. Install the MATE desktop.

$ sudo amazon-linux-extras install mate-desktop1.x -y

$ sudo yum groupinstall "MATE Desktop" -y
$ sudo systemctl disable firewalld

3. Copy the xorg.conf file to /etc/X11/xorg.conf.

4. Reboot the instance.

$ sudo reboot

307
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

5. (Optional) Install the NICE DCV server to use NICE DCV as a high-performance display protocol, and
then connect to a NICE DCV session using your preferred client.

To set up an interactive desktop on Ubuntu

1. Install the MATE desktop.

$ sudo apt install xorg-dev ubuntu-mate-desktop -y

$ sudo apt purge ifupdown -y

2. Copy the xorg.conf file to /etc/X11/xorg.conf.

3. Reboot the instance.

$ sudo reboot

4. (Optional) Install the NICE DCV server to use NICE DCV as a high-performance display protocol, and
then connect to a NICE DCV session using your preferred client.

To set up an interactive desktop on CentOS

1. Install the EPEL repository.

$ sudo yum update -y

$ sudo yum install epel-release -y

2. Install the MATE desktop.

$ sudo yum groupinstall "MATE Desktop" -y

$ sudo systemctl disable firewalld

3. Copy the xorg.conf file to /etc/X11/xorg.conf.

4. Reboot the instance.

$ sudo reboot

5. (Optional) Install the NICE DCV server to use NICE DCV as a high-performance display protocol, and
then connect to a NICE DCV session using your preferred client.

Activate NVIDIA GRID Virtual Applications

To activate the GRID Virtual Applications on G3 and G4dn instances (NVIDIA GRID Virtual Workstation is
enabled by default), you must define the product type for the driver in the /etc/nvidia/gridd.conf
file.

To activate GRID Virtual Applications on Linux instances

1. Create the /etc/nvidia/gridd.conf file from the provided template file.

[ec2-user ~]$ sudo cp /etc/nvidia/gridd.conf.template /etc/nvidia/gridd.conf

2. Open the /etc/nvidia/gridd.conf file in your favorite text editor.

3. Find the FeatureType line, and set it equal to 0. Then add a line with IgnoreSP=TRUE.

FeatureType=0

308
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Accelerated computing

IgnoreSP=TRUE

4. Save the file and exit.

5. Reboot the instance to pick up the new configuration.

[ec2-user ~]$ sudo reboot

Optimize GPU settings

There are several GPU setting optimizations that you can perform to achieve the best performance on
G3, G4dn, P2, P3, P3dn, and P4d instances. With some of these instance types, the NVIDIA driver uses an
autoboost feature, which varies the GPU clock speeds. By disabling autoboost and setting the GPU clock
speeds to their maximum frequency, you can consistently achieve the maximum performance with your
GPU instances. The following procedure helps you to configure the GPU settings to be persistent, disable
the autoboost feature if needed, and set the GPU clock speeds to their maximum frequency.

To optimize GPU settings

1. Configure the GPU settings to be persistent. This command can take several minutes to run.

[ec2-user ~]$ sudo nvidia-persistenced

2. G2, G3, and P2 instances: Disable the autoboost feature for all GPUs on the instance.
Note
GPUs on G4dn, P3, P3dn, and P4d instances do not support autoboost.

[ec2-user ~]$ sudo nvidia-smi --auto-boost-default=0

3. Set all GPU clock speeds to their maximum frequency. Use the memory and graphics clock speeds
specified in the following commands.

Some versions of the NVIDIA driver do not support setting the application clock speed, and display
the error "Setting applications clocks is not supported for GPU...", which you can
ignore.

• G3 instances:

[ec2-user ~]$ sudo nvidia-smi -ac 2505,1177

• G4dn instances:

[ec2-user ~]$ sudo nvidia-smi -ac 5001,1590

• P2 instances:

[ec2-user ~]$ sudo nvidia-smi -ac 2505,875

• P3 and P3dn instances:

[ec2-user ~]$ sudo nvidia-smi -ac 877,1530

• P4d instances:

[ec2-user ~]$ sudo nvidia-smi -ac 1215,1410

309
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Find an instance type

Find an Amazon EC2 instance type

Before you can launch an instance, you must select an instance type to use. The instance type that you
choose might depend on your requirements for the instances that you'll launch. For example, you might
choose an instance type based on the following requirements:

• Availability Zone or Region

• Compute
• Memory
• Networking
• Pricing
• Storage

Find an instance type using the console

You can find an instance type that meets your needs using the Amazon EC2 console.

To find an instance type using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region in which to launch your instances. You can select any
Region that's available to you, regardless of your location.
3. In the navigation pane, choose Instance Types.
4. (Optional) Choose the preferences (gear) icon to select which instance type attributes to display,
such as On-Demand Linux pricing, and then choose Confirm. Alternatively, select an instance type
and view all attributes using the Details pane.
5. Use the instance type attributes to filter the list of displayed instance types to only the instance
types that meet your needs. For example, you can list all instance types that have more than eight
vCPUs and also support hibernation.
6. (Optional) Select multiple instance types to see a side-by-side comparison across all attributes in the
Details pane.
7. (Optional) To save the list of instance types to a comma-separated values (.csv) file for further
review, choose Download list CSV. The file includes all instance types that match the filters you set.
8. After locating instance types that meet your needs, you can use them to launch instances. For more
information, see Launch an instance using the Launch Instance Wizard (p. 526).

Find an instance type using the AWS CLI

You can use AWS CLI commands for Amazon EC2 to find an instance type that meet your needs.

To find an instance type using the AWS CLI

1. If you have not done so already, install the AWS CLI For more information, see the AWS Command
Line Interface User Guide.
2. Use the describe-instance-types command to filter instance types based on instance attributes. For
example, you can use the following command to display only instance types with 48 vCPUs.

aws ec2 describe-instance-types --filters "Name=vcpu-info.default-vcpus,Values=48"

310
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the instance type

3. Use the describe-instance-type-offerings command to filter instance types offered by location

(Region or Availability Zone). For example, you can use the following command to display the
instance types offered in the specified Availability Zone.

aws ec2 describe-instance-type-offerings --location-type "availability-zone" --filters

Name=location,Values=us-east-2a --region us-east-2

4. After locating instance types that meet your needs, make note of them so that you can use these
instance types when you launch instances. For more information, see Launching an Instance Using
the AWS CLI in the AWS Command Line Interface User Guide.

Change the instance type

As your needs change, you might find that your instance is over-utilized (the instance type is too small)
or under-utilized (the instance type is too large). If this is the case, you can change the size of your
instance. For example, if your t2.micro instance is too small for its workload, you can change it to
another instance type that is appropriate for the workload.

You might also want to migrate from a previous generation instance type to a current generation
instance type to take advantage of some features; for example, support for IPv6.

If the root device for your instance is an EBS volume, you can change the size of the instance simply by
changing its instance type, which is known as resizing it. If the root device for your instance is an instance
store volume, you must migrate your application to a new instance with the instance type that you need.
For more information about root device volumes, see Storage for the root device (p. 102).

Requirements

• You must select an instance type that is compatible with the configuration of the instance. If the
instance type that you want is not compatible with the instance configuration you have, then you must
migrate your application to a new instance with the instance type that you need.
• You cannot resize an instance if hibernation is enabled.

Contents
• Compatibility for resizing instances (p. 311)
• Resize an Amazon EBS–backed instance (p. 312)
• Migrate an instance store-backed instance (p. 314)
• Migrate to a new instance configuration (p. 315)

Compatibility for resizing instances

You can resize an instance only if its current instance type and the new instance type that you want are
compatible in the following ways:

• Virtualization type: Linux AMIs use one of two types of virtualization: paravirtual (PV) or hardware
virtual machine (HVM). You can't resize an instance that was launched from a PV AMI to an instance
type that is HVM only. For more information, see Linux AMI virtualization types (p. 104). To check the
virtualization type of your instance, see the Virtualization field on the details pane of the Instances
screen in the Amazon EC2 console.
• Architecture: AMIs are specific to the architecture of the processor, so you must select an instance type
with the same processor architecture as the current instance type. For example:
• If you are resizing an instance type with a processor based on the Arm architecture, you are limited
to the instance types that support a processor based on the Arm architecture, such as C6g and M6g.

311
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the instance type

• The following instance types are the only instance types that support 32-bit AMIs: t2.nano,
t2.micro, t2.small, t2.medium, c3.large, t1.micro, m1.small, m1.medium, and
c1.medium. If you are resizing a 32-bit instance, you are limited to these instance types.
• Network: Newer instance types must be launched in a VPC. Therefore, you can't resize an instance in
the EC2-Classic platform to a instance type that is available only in a VPC unless you have a nondefault
VPC. To check whether your instance is in a VPC, check the VPC ID value on the details pane of the
Instances screen in the Amazon EC2 console. For more information, see Migrate from EC2-Classic to a
VPC (p. 962).
• Enhanced networking: Instance types that support enhanced networking (p. 864) require the
necessary drivers installed. For example, instances based on the Nitro System (p. 209) require EBS-
backed AMIs with the Elastic Network Adapter (ENA) drivers installed. To resize an instance from a type
that does not support enhanced networking to a type that supports enhanced networking, you must
install the ENA drivers (p. 865) or ixgbevf drivers (p. 875) on the instance, as appropriate.
• NVMe: EBS volumes are exposed as NVMe block devices on instances built on the Nitro
System (p. 209). If you resize an instance from an instance type that does not support NVMe to an
instance type that supports NVMe, you must first install the NVMe drivers (p. 1214) on your instance.
Also, the device names for devices that you specify in the block device mapping are renamed using
NVMe device names (/dev/nvme[0-26]n1). Therefore, to mount file systems at boot time using /
etc/fstab, you must use UUID/Label instead of device names.
• AMI: For information about the AMIs required by instance types that support enhanced networking
and NVMe, see the Release Notes in the following documentation:
• General purpose instances (p. 213)
• Compute optimized instances (p. 264)
• Memory optimized instances (p. 272)
• Storage optimized instances (p. 283)

Resize an Amazon EBS–backed instance

You must stop your Amazon EBS–backed instance before you can change its instance type. When you
stop and start an instance, be aware of the following:

• We move the instance to new hardware; however, the instance ID does not change.
• If your instance has a public IPv4 address, we release the address and give it a new public IPv4 address.
The instance retains its private IPv4 addresses, any Elastic IP addresses, and any IPv6 addresses.
• When you resize an instance, the resized instance usually has the same number of instance store
volumes that you specified when you launched the original instance. With instance types that support
NVMe instance store volumes (which are available by default), the resized instance might have
additional instance store volumes, depending on the AMI. Otherwise, you can migrate your application
to an instance with a new instance type manually, specifying the number of instance store volumes
that you need when you launch the new instance.
• If your instance is in an Auto Scaling group, the Amazon EC2 Auto Scaling service marks the stopped
instance as unhealthy, and may terminate it and launch a replacement instance. To prevent this,
you can suspend the scaling processes for the group while you're resizing your instance. For more
information, see Suspending and Resuming Scaling Processes in the Amazon EC2 Auto Scaling User
Guide.
• If your instance is in a cluster placement group (p. 928) and, after changing the instance type, the
instance start fails, try the following: stop all the instances in the cluster placement group, change
the instance type for the affected instance, and then restart all the instances in the cluster placement
group.
• Ensure that you plan for downtime while your instance is stopped. Stopping and resizing an instance
may take a few minutes, and restarting your instance may take a variable amount of time depending
on your application's startup scripts.

312
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the instance type

For more information, see Stop and start your instance (p. 623).

Use the following procedure to resize an Amazon EBS–backed instance using the AWS Management
Console.

New console

To resize an Amazon EBS–backed instance

1. (Optional) If the new instance type requires drivers that are not installed on the existing
instance, you must connect to your instance and install the drivers first. For more information,
see Compatibility for resizing instances (p. 311).
2. Open the Amazon EC2 console.
3. In the navigation pane, choose Instances.
4. Select the instance and choose Actions, Instance state, Stop instance.
5. In the confirmation dialog box, choose Stop. It can take a few minutes for the instance to stop.
6. With the instance still selected, choose Actions, Instance settings, Change instance type. This
action is disabled if the instance state is not stopped.
7. In the Change instance type dialog box, do the following:

a. From Instance type, select the instance type that you want. If the instance type that
you want does not appear in the list, then it is not compatible with the configuration of
your instance (for example, because of virtualization type). For more information, see
Compatibility for resizing instances (p. 311).
b. (Optional) If the instance type that you selected supports EBS–optimization, select
EBS-optimized to enable EBS–optimization or deselect EBS-optimized to disable EBS–
optimization. If the instance type that you selected is EBS–optimized by default, EBS-
optimized is selected and you can't deselect it.
c. Choose Apply to accept the new settings.
8. To restart the stopped instance, select the instance and choose Instance state, Start instance. It
can take a few minutes for the instance to enter the running state.
9. (Troubleshooting) If your instance won't boot, it is possible that one of the requirements for the
new instance type was not met. For more information, see Why is my Linux instance not booting
after I changed its type?

Old console

To resize an Amazon EBS–backed instance

1. (Optional) If the new instance type requires drivers that are not installed on the existing
instance, you must connect to your instance and install the drivers first. For more information,
see Compatibility for resizing instances (p. 311).
2. Open the Amazon EC2 console.
3. In the navigation pane, choose Instances.
4. Select the instance and choose Actions, Instance State, Stop.
5. In the confirmation dialog box, choose Yes, Stop. It can take a few minutes for the instance to
stop.
6. With the instance still selected, choose Actions, Instance Settings, Change Instance Type. This
action is disabled if the instance state is not stopped.
7. In the Change Instance Type dialog box, do the following:

a. From Instance Type, select the instance type that you want. If the instance type that
you want does not appear in the list, then it is not compatible with the configuration of

313
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the instance type

your instance (for example, because of virtualization type). For more information, see
Compatibility for resizing instances (p. 311).
b. (Optional) If the instance type that you selected supports EBS–optimization, select
EBS-optimized to enable EBS–optimization or deselect EBS-optimized to disable EBS–
optimization. If the instance type that you selected is EBS–optimized by default, EBS-
optimized is selected and you can't deselect it.
c. Choose Apply to accept the new settings.
8. To restart the stopped instance, select the instance and choose Actions, Instance State, Start.
9. In the confirmation dialog box, choose Yes, Start. It can take a few minutes for the instance to
enter the running state.
10. (Troubleshooting) If your instance won't boot, it is possible that one of the requirements for the
new instance type was not met. For more information, see Why is my Linux instance not booting
after I changed its type?

Migrate an instance store-backed instance

When you want to move your application from one instance store-backed instance to an instance store-
backed instance with a different instance type, you must migrate it by creating an image from your
instance, and then launching a new instance from this image with the instance type that you need.
To ensure that your users can continue to use the applications that you're hosting on your instance
uninterrupted, you must take any Elastic IP address that you've associated with your original instance
and associate it with the new instance. Then you can terminate the original instance.

New console

To migrate an instance store-backed instance

1. Back up any data on your instance store volumes that you need to keep to persistent storage.
To migrate data on your EBS volumes that you need to keep, take a snapshot of the volumes
(see Create Amazon EBS snapshots (p. 1129)) or detach the volume from the instance so that
you can attach it to the new instance later (see Detach an Amazon EBS volume from a Linux
instance (p. 1123)).
2. Create an AMI from your instance store-backed instance by satisfying the prerequisites and
following the procedures in Create an instance store-backed Linux AMI (p. 130). When you are
finished creating an AMI from your instance, return to this procedure.
3. Open the Amazon EC2 console and in the navigation pane, choose AMIs. From the filter lists,
choose Owned by me, and select the image that you created in the previous step. Notice that
AMI Name is the name that you specified when you registered the image and Source is your
Amazon S3 bucket.
Note
If you do not see the AMI that you created in the previous step, make sure that you
have selected the Region in which you created your AMI.
4. Choose Launch. When you specify options for the instance, be sure to select the new
instance type that you want. If the instance type that you want can't be selected, then it
is not compatible with configuration of the AMI that you created (for example, because of
virtualization type). You can also specify any EBS volumes that you detached from the original
instance.

It can take a few minutes for the instance to enter the running state.
5. (Optional) You can terminate the instance that you started with, if it's no longer needed. Select
the instance and verify that you are about to terminate the original instance, not the new
instance (for example, check the name or launch time). Choose Instance state, Terminate
instance.

314
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the instance type

Old console

To migrate an instance store-backed instance

1. Back up any data on your instance store volumes that you need to keep to persistent storage.
To migrate data on your EBS volumes that you need to keep, take a snapshot of the volumes
(see Create Amazon EBS snapshots (p. 1129)) or detach the volume from the instance so that
you can attach it to the new instance later (see Detach an Amazon EBS volume from a Linux
instance (p. 1123)).
2. Create an AMI from your instance store-backed instance by satisfying the prerequisites and
following the procedures in Create an instance store-backed Linux AMI (p. 130). When you are
finished creating an AMI from your instance, return to this procedure.
3. Open the Amazon EC2 console and in the navigation pane, choose AMIs. From the filter lists,
choose Owned by me, and choose the image that you created in the previous step. Notice that
AMI Name is the name that you specified when you registered the image and Source is your
Amazon S3 bucket.
Note
If you do not see the AMI that you created in the previous step, make sure that you
have selected the Region in which you created your AMI.
4. Choose Launch. When you specify options for the instance, be sure to select the new
instance type that you want. If the instance type that you want can't be selected, then it
is not compatible with configuration of the AMI that you created (for example, because of
virtualization type). You can also specify any EBS volumes that you detached from the original
instance.

It can take a few minutes for the instance to enter the running state.
5. (Optional) You can terminate the instance that you started with, if it's no longer needed.
Select the instance and verify that you are about to terminate the original instance, not the
new instance (for example, check the name or launch time). Choose Actions, Instance State,
Terminate.

Migrate to a new instance configuration

If the current configuration of your instance is incompatible with the new instance type that you want,
then you can't resize the instance to that instance type. Instead, you can migrate your application to a
new instance with a configuration that is compatible with the new instance type that you want.

If you want to move from an instance launched from a PV AMI to an instance type that is HVM only, the
general process is as follows:

New console

To migrate your application to a compatible instance

1. Back up any data on your instance store volumes that you need to keep to persistent storage.
To migrate data on your EBS volumes that you need to keep, create a snapshot of the volumes
(see Create Amazon EBS snapshots (p. 1129)) or detach the volume from the instance so that
you can attach it to the new instance later (see Detach an Amazon EBS volume from a Linux
instance (p. 1123)).
2. Launch a new instance, selecting the following:

• An HVM AMI.
• The HVM only instance type.
• If you are using an Elastic IP address, select the VPC that the original instance is currently
running in.

315
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the instance type

• Any EBS volumes that you detached from the original instance and want to attach to the new
instance, or new EBS volumes based on the snapshots that you created.
• If you want to allow the same traffic to reach the new instance, select the security group that
is associated with the original instance.
3. Install your application and any required software on the instance.
4. Restore any data that you backed up from the instance store volumes of the original instance.
5. If you are using an Elastic IP address, assign it to the newly launched instance as follows:

a. In the navigation pane, choose Elastic IPs.

b. Select the Elastic IP address that is associated with the original instance and choose
Actions, Disassociate Elastic IP address. When prompted for confirmation, choose
Disassociate.
c. With the Elastic IP address still selected, choose Actions, Associate Elastic IP address.
d. For Resource type, choose Instance.
e. For Instance, choose the instance with which to associate the Elastic IP address. You can
also enter text to search for a specific instance.
f. (Optional) For Private IP address, specify a private IP address with which to associate the
Elastic IP address.
g. Choose Associate.
6. (Optional) You can terminate the original instance if it's no longer needed. Select the instance
and verify that you are about to terminate the original instance, not the new instance (for
example, check the name or launch time). Choose Instance state, Terminate instance.

Old console

To migrate your application to a compatible instance

1. Back up any data on your instance store volumes that you need to keep to persistent storage.
To migrate data on your EBS volumes that you need to keep, create a snapshot of the volumes
(see Create Amazon EBS snapshots (p. 1129)) or detach the volume from the instance so that
you can attach it to the new instance later (see Detach an Amazon EBS volume from a Linux
instance (p. 1123)).
2. Launch a new instance, selecting the following:

• An HVM AMI.
• The HVM only instance type.
• If you are using an Elastic IP address, select the VPC that the original instance is currently
running in.
• Any EBS volumes that you detached from the original instance and want to attach to the new
instance, or new EBS volumes based on the snapshots that you created.
• If you want to allow the same traffic to reach the new instance, select the security group that
is associated with the original instance.
3. Install your application and any required software on the instance.
4. Restore any data that you backed up from the instance store volumes of the original instance.
5. If you are using an Elastic IP address, assign it to the newly launched instance as follows:

a. In the navigation pane, choose Elastic IPs.

b. Select the Elastic IP address that is associated with the original instance and choose
Actions, Disassociate address. When prompted for confirmation, choose Disassociate
address.
c. With the Elastic IP address still selected, choose Actions, Associate address.

316
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get recommendations

d. From Instance, select the new instance, and then choose Associate.
6. (Optional) You can terminate the original instance if it's no longer needed. Select the instance
and verify that you are about to terminate the original instance, not the new instance (for
example, check the name or launch time). Choose Actions, Instance State, Terminate.

Get recommendations for an instance type

AWS Compute Optimizer provides Amazon EC2 instance recommendations to help you improve
performance, save money, or both. You can use these recommendations to decide whether to move to a
new instance type.

To make recommendations, Compute Optimizer analyzes your existing instance specifications and
utilization metrics. The compiled data is then used to recommend which Amazon EC2 instance types are
best able to handle the existing workload. Recommendations are returned along with per-hour instance
pricing.

This topic outlines how to view recommendations through the Amazon EC2 console. For more
information, see the AWS Compute Optimizer User Guide.
Note
To get recommendations from Compute Optimizer, you must first opt in to Compute Optimizer.
For more information, see Getting Started with AWS Compute Optimizer in the AWS Compute
Optimizer User Guide.

Contents
• Limitations (p. 317)
• Findings (p. 317)
• View recommendations (p. 318)
• Considerations for evaluating recommendations (p. 319)

Limitations
Compute Optimizer currently generates recommendations for M, C, R, T, and X instance types. Other
instance types are not considered by Compute Optimizer. If you're using other instance types, they will
not be listed in the Compute Optimizer recommendations view. For information about these and other
instance types, see Instance types (p. 203).

Findings
Compute Optimizer classifies its findings for EC2 instances as follows:

• Under-provisioned – An EC2 instance is considered under-provisioned when at least one specification

of your instance, such as CPU, memory, or network, does not meet the performance requirements of
your workload. Under-provisioned EC2 instances might lead to poor application performance.
• Over-provisioned – An EC2 instance is considered over-provisioned when at least one specification
of your instance, such as CPU, memory, or network, can be sized down while still meeting the
performance requirements of your workload, and when no specification is under-provisioned. Over-
provisioned EC2 instances might lead to unnecessary infrastructure cost.
• Optimized – An EC2 instance is considered optimized when all specifications of your instance, such as
CPU, memory, and network, meet the performance requirements of your workload, and the instance is
not over-provisioned. An optimized EC2 instance runs your workloads with optimal performance and
infrastructure cost. For optimized instances, Compute Optimizer might sometimes recommend a new
generation instance type.

317
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get recommendations

• None – There are no recommendations for this instance. This might occur if you've been opted in to
Compute Optimizer for less than 12 hours, or when the instance has been running for less than 30
hours, or when the instance type is not supported by Compute Optimizer. For more information, see
Limitations (p. 317) in the previous section.

View recommendations
After you opt in to Compute Optimizer, you can view the findings that Compute Optimizer generates for
your EC2 instances in the EC2 console. You can then access the Compute Optimizer console to view the
recommendations. If you recently opted in, findings might not be reflected in the EC2 console for up to
12 hours.

New console

To view a recommendation for an EC2 instance through the EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and then choose the instance ID.
3. On the instance summary page, in the AWS Compute Optimizer banner near the bottom of the
page, choose View detail.

The instance opens in Compute Optimizer, where it is labeled as the Current instance. Up to
three different instance type recommendations, labeled Option 1, Option 2, and Option 3, are
provided. The bottom half of the window shows recent CloudWatch metric data for the current
instance: CPU utilization, Memory utilization, Network in, and Network out.
4.
(Optional) In the Compute Optimizer console, choose the settings ( ) icon to change the
visible columns in the table, or to view the public pricing information for a different purchasing
option for the current and recommended instance types.
Note
If you’ve purchased a Reserved Instance, your On-Demand Instance might be billed as
a Reserved Instance. Before you change your current instance type, first evaluate the
impact on Reserved Instance utilization and coverage.

Old console

To view a recommendation for an EC2 instance through the EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select an instance, and on the Description tab, inspect the Finding field. Choose View detail.

The instance opens in Compute Optimizer, where it is labeled as the Current instance. Up to
three different instance type recommendations, labeled Option 1, Option 2, and Option 3, are
provided. The bottom half of the window shows recent CloudWatch metric data for the current
instance: CPU utilization, Memory utilization, Network in, and Network out.
4.
(Optional) In the Compute Optimizer console, choose the settings ( ) icon to change the
visible columns in the table, or to view the public pricing information for a different purchasing
option for the current and recommended instance types.
Note
If you’ve purchased a Reserved Instance, your On-Demand Instance might be billed as
a Reserved Instance. Before you change your current instance type, first evaluate the
impact on Reserved Instance utilization and coverage.

318
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get recommendations

Determine whether you want to use one of the recommendations. Decide whether to optimize for
performance improvement, for cost reduction, or for a combination of the two. For more information,
see Viewing Resource Recommendations in the AWS Compute Optimizer User Guide.

To view recommendations for all EC2 instances across all Regions through the Compute
Optimizer console

1. Open the Compute Optimizer console at https://console.aws.amazon.com/compute-optimizer/.

2. Choose View recommendations for all EC2 instances.
3. You can perform the following actions on the recommendations page:

a. To filter recommendations to one or more AWS Regions, enter the name of the Region in the
Filter by one or more Regions text box, or choose one or more Regions in the drop-down list
that appears.
b. To view recommendations for resources in another account, choose Account, and then select a
different account ID.

This option is available only if you are signed in to a management account of an organization,
and you opted in all member accounts within the organization.
c. To clear the selected filters, choose Clear filters.
d. To change the purchasing option that is displayed for the current and recommended instance

types, choose the settings ( ) icon , and then choose On-Demand Instances, Reserved
Instances, standard 1-year no upfront, or Reserved Instances, standard 3-year no upfront.
e. To view details, such as additional recommendations and a comparison of utilization metrics,
choose the finding (Under-provisioned, Over-provisioned, or Optimized) listed next to the
desired instance. For more information, see Viewing Resource Details in the AWS Compute
Optimizer User Guide.

Considerations for evaluating recommendations

Before changing an instance type, consider the following:

• The recommendations don’t forecast your usage. Recommendations are based on your historical usage
over the most recent 14-day time period. Be sure to choose an instance type that is expected to meet
your future resource needs.
• Focus on the graphed metrics to determine whether actual usage is lower than instance capacity.
You can also view metric data (average, peak, percentile) in CloudWatch to further evaluate your EC2
instance recommendations. For example, notice how CPU percentage metrics change during the day
and whether there are peaks that need to be accommodated. For more information, see Viewing
Available Metrics in the Amazon CloudWatch User Guide.
• Compute Optimizer might supply recommendations for burstable performance instances, which are
T3, T3a, and T2 instances. If you periodically burst above the baseline, make sure that you can continue
to do so based on the vCPUs of the new instance type. For more information, see CPU credits and
baseline utilization for burstable performance instances (p. 225).
• If you’ve purchased a Reserved Instance, your On-Demand Instance might be billed as a Reserved
Instance. Before you change your current instance type, first evaluate the impact on Reserved Instance
utilization and coverage.
• Consider conversions to newer generation instances, where possible.
• When migrating to a different instance family, make sure the current instance type and the new
instance type are compatible, for example, in terms of virtualization, architecture, or network type. For
more information, see Compatibility for resizing instances (p. 311).
• Finally, consider the performance risk rating that's provided for each recommendation. Performance
risk indicates the amount of effort you might need to spend in order to validate whether the

319
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance purchasing options

recommended instance type meets the performance requirements of your workload. We also
recommend rigorous load and performance testing before and after making any changes.

There are other considerations when resizing an EC2 instance. For more information, see Change the
instance type (p. 311).

Additional resources

• Instance types (p. 203)

• AWS Compute Optimizer User Guide

Instance purchasing options

Amazon EC2 provides the following purchasing options to enable you to optimize your costs based on
your needs:

• On-Demand Instances – Pay, by the second, for the instances that you launch.
• Savings Plans – Reduce your Amazon EC2 costs by making a commitment to a consistent amount of
usage, in USD per hour, for a term of 1 or 3 years.
• Reserved Instances – Reduce your Amazon EC2 costs by making a commitment to a consistent
instance configuration, including instance type and Region, for a term of 1 or 3 years.
• Spot Instances – Request unused EC2 instances, which can reduce your Amazon EC2 costs significantly.
• Dedicated Hosts – Pay for a physical host that is fully dedicated to running your instances, and bring
your existing per-socket, per-core, or per-VM software licenses to reduce costs.
• Dedicated Instances – Pay, by the hour, for instances that run on single-tenant hardware.
• Capacity Reservations – Reserve capacity for your EC2 instances in a specific Availability Zone for any
duration.

If you require a capacity reservation, purchase Reserved Instances or Capacity Reservations for a specific
Availability Zone. Spot Instances are a cost-effective choice if you can be flexible about when your
applications run and if they can be interrupted. Dedicated Hosts or Dedicated Instances can help you
address compliance requirements and reduce costs by using your existing server-bound software
licenses. For more information, see Amazon EC2 Pricing.

For more information about Savings Plans, see the Savings Plans User Guide.

Contents
• Determine the instance lifecycle (p. 320)
• On-Demand Instances (p. 322)
• Reserved Instances (p. 325)
• Scheduled Reserved Instances (p. 363)
• Spot Instances (p. 364)
• Dedicated Hosts (p. 465)
• Dedicated Instances (p. 496)
• On-Demand Capacity Reservations (p. 501)

Determine the instance lifecycle

The lifecycle of an instance starts when it is launched and ends when it is terminated. The purchasing
option that you choose affects the lifecycle of the instance. For example, an On-Demand Instance runs

320
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Determine the instance lifecycle

when you launch it and ends when you terminate it. A Spot Instance runs as long as capacity is available
and your maximum price is higher than the Spot price.

Use the following procedure to determine the lifecycle of an instance.

New console

To determine the instance lifecycle using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance.
4. On the Details tab, under Instance details, find Lifecycle. If the value is spot, the instance
is a Spot Instance. If the value is normal, the instance is either an On-Demand Instance or a
Reserved Instance.
5. On the Details tab, under Host and placement group, find Tenancy. If the value is host, the
instance is running on a Dedicated Host. If the value is dedicated, the instance is a Dedicated
Instance.
6. (Optional) If you have purchased a Reserved Instance and want to verify that it is being applied,
you can check the usage reports for Amazon EC2. For more information, see Amazon EC2 usage
reports (p. 1325).

Old console

To determine the instance lifecycle using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance.
4. On the Description tab, find Tenancy. If the value is host, the instance is running on a
Dedicated Host. If the value is dedicated, the instance is a Dedicated Instance.
5. On the Description tab, find Lifecycle. If the value is spot, the instance is a Spot Instance. If the
value is normal, the instance is either an On-Demand Instance or a Reserved Instance.
6. (Optional) If you have purchased a Reserved Instance and want to verify that it is being applied,
you can check the usage reports for Amazon EC2. For more information, see Amazon EC2 usage
reports (p. 1325).

To determine the instance lifecycle using the AWS CLI

Use the following describe-instances command:

aws ec2 describe-instances --instance-ids i-1234567890abcdef0

If the instance is running on a Dedicated Host, the output contains the following information:

"Tenancy": "host"

If the instance is a Dedicated Instance, the output contains the following information:

"Tenancy": "dedicated"

If the instance is a Spot Instance, the output contains the following information:

321
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Instances

"InstanceLifecycle": "spot"

Otherwise, the output does not contain InstanceLifecycle.

On-Demand Instances
With On-Demand Instances, you pay for compute capacity by the second with no long-term
commitments. You have full control over its lifecycle—you decide when to launch, stop, hibernate, start,
reboot, or terminate it.

There is no long-term commitment required when you purchase On-Demand Instances. You pay only
for the seconds that your On-Demand Instances are in the running state. The price per second for a
running On-Demand Instance is fixed, and is listed on the Amazon EC2 Pricing, On-Demand Pricing page.

We recommend that you use On-Demand Instances for applications with short-term, irregular workloads
that cannot be interrupted.

For significant savings over On-Demand Instances, use AWS Savings Plans, Spot Instances (p. 364), or
Reserved Instances (p. 325).

Contents
• Work with On-Demand Instances (p. 322)
• On-Demand Instance limits (p. 322)
• Calculate how many vCPUs you need (p. 323)
• Request a limit increase (p. 325)
• Monitor On-Demand Instance limits and usage (p. 325)
• Query the prices of AWS services (p. 325)

Work with On-Demand Instances

You can work with On-Demand Instances in the following ways:

• Launch your instance (p. 525)

• Connect to your Linux instance (p. 598)
• Stop and start your instance (p. 623)
• Hibernate your Linux instance (p. 627)
• Reboot your instance (p. 639)
• Instance retirement (p. 640)
• Terminate your instance (p. 643)
• Recover your instance (p. 648)
• Configure your Amazon Linux instance (p. 649)
• Identify EC2 Linux instances (p. 738)

If you're new to Amazon EC2, see How to get started with Amazon EC2 (p. 2).

On-Demand Instance limits

There is a limit on the number of running On-Demand Instances per AWS account per Region. On-
Demand Instance limits are managed in terms of the number of virtual central processing units (vCPUs)
that your running On-Demand Instances are using, regardless of the instance type.

322
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Instances

There are six On-Demand Instance limits, listed in the following table. Each limit specifies the vCPU limit
for one or more instance families. For information about the different instance families, generations, and
sizes, see Amazon EC2 Instance Types.

On-Demand Instance limit name Default vCPU limit

Running On-Demand All Standard (A, C, D, H, I, M, 1152 vCPUs

R, T, Z) instances

Running On-Demand All F instances 128 vCPUs

Running On-Demand All G instances 128 vCPUs

Running On-Demand All Inf instances 128 vCPUs

Running On-Demand All P instances 128 vCPUs

Running On-Demand All X instances 128 vCPUs

Note
New AWS accounts might start with limits that are lower than the limits described here.

With vCPU limits, you can use your limit in terms of the number of vCPUs required to launch any
combination of instance types that meet your changing application needs. For example, with a
Standard instance limit of 256 vCPUs, you could launch 32 m5.2xlarge instances (32 x 8 vCPUs) or 16
c5.4xlarge instances (16 x 16 vCPUs), or a combination of any Standard instance types and sizes that
total 256 vCPUs. For more information, see EC2 On-Demand Instance limits.

Calculate how many vCPUs you need

You can use the vCPU limits calculator to determine the number of vCPUs that you require for your
application needs.

When using the calculator, keep the following in mind: The calculator assumes that you have reached
your current limit. The value that you enter for Instance count is the number of instances that you need
to launch in addition to what is permitted by your current limit. The calculator adds your current limit to
the Instance count to arrive at a new limit.

The following screenshot shows the vCPU limits calculator.

323
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Instances

You can view and use the following controls and information:

• Instance type – The instance types that you add to the vCPU limits calculator.
• Instance count – The number of instances that you require for the selected instance type.
• vCPU count – The number of vCPUs that corresponds to the Instance count.
• Current limit – Your current limit for the limit type to which the instance type belongs. The limit
applies to all instance types of the same limit type. For example, in the preceding screenshot, the
current limit for m5.2xlarge and c5.4xlarge is 1,920 vCPUs, which is the limit for all the instance
types that belong to the All Standard instances limit.
• New limit – The new limit, in number of vCPUs, which is calculated by adding vCPU count and Current
limit.
• X – Choose the X to remove the row.
• Add instance type – Choose Add instance type to add another instance type to the calculator.
• Limits calculation – Displays the current limit, vCPUs needed, and new limit for the limit types.
• Instance limit name – The limit type for the instance types that you selected.
• Current limit – The current limit for the limit type.
• vCPUs needed – The number of vCPUs that corresponds to the number of instances that you
specified in Instance count. For the All Standard instances limit type, the vCPUs needed is calculated
by adding the values for vCPU count for all the instance types of this limit type.
• New limit – The new limit is calculated by adding Current limit and vCPUs needed.
• Options – Choose Request limit increase to request a limit increase for the corresponding limit
type.

To calculate the number of required vCPUs

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select a Region.
3. From the left navigator, choose Limits.
4. Choose Calculate vCPU limit.

324
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

5. Choose Add instance type, choose the required instance type, and specify the required number of
instances. To add more instance types, choose Add instance type again.
6. View Limits calculation for the required new limit.
7. When you've finished using the calculator, choose Close.

Request a limit increase

You can request a limit increase for each On-Demand Instance limit type from the Limits page or the
vCPU limits calculator in the Amazon EC2 console. Complete the required fields on the AWS Support
Center limit increase form with your use case. For Primary Instance Type, select the limit type that
corresponds to the Instance limit name in the vCPU limits calculator. For the new limit value, use the
value that appears in the New limit column in the vCPU limits calculator. For more information about
requesting a limit increase, see Amazon EC2 service quotas (p. 1323).

Monitor On-Demand Instance limits and usage

You can view and manage your On-Demand Instance limits using the following:

• The Limits page in the Amazon EC2 console

• The Amazon EC2 Services quotas page in the Service Quotas console
• The get-service-quota AWS CLI
• The Service Limits page in the AWS Trusted Advisor console

For more information, see Amazon EC2 service quotas (p. 1323) in the Amazon EC2 User Guide for Linux
Instances, Viewing a Service Quota in the Service Quotas User Guide, and AWS Trusted Advisor.

With Amazon CloudWatch metrics integration, you can monitor EC2 usage against limits. You can
also configure alarms to warn about approaching limits. For more information, see Using Amazon
CloudWatch Alarms in the Service Quotas User Guide.

Query the prices of AWS services

You can use the Price List Service API or the AWS Price List API to query the prices of On-Demand
Instances. For more information, see Using the AWS Price List API in the AWS Billing and Cost
Management User Guide.

Reserved Instances
Reserved Instances provide you with significant savings on your Amazon EC2 costs compared to On-
Demand Instance pricing. Reserved Instances are not physical instances, but rather a billing discount
applied to the use of On-Demand Instances in your account. These On-Demand Instances must match
certain attributes, such as instance type and Region, in order to benefit from the billing discount.

Savings Plans also offer significant savings on your Amazon EC2 costs compared to On-Demand Instance
pricing. With Savings Plans, you make a commitment to a consistent usage amount, measured in USD per
hour. This provides you with the flexibility to use the instance configurations that best meet your needs
and continue to save money, instead of making a commitment to a specific instance configuration. For
more information, see the AWS Savings Plans User Guide.

Reserved Instances topics

• Reserved Instance overview (p. 326)
• Key variables that determine Reserved Instance pricing (p. 326)
• Reserved Instance limits (p. 328)
• Regional and zonal Reserved Instances (scope) (p. 328)

325
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

• Types of Reserved Instances (offering classes) (p. 329)

• How Reserved Instances are applied (p. 330)
• Use your Reserved Instances (p. 335)
• How you are billed (p. 336)
• Buying Reserved Instances (p. 340)
• Sell in the Reserved Instance Marketplace (p. 346)
• Modify Reserved Instances (p. 352)
• Exchange Convertible Reserved Instances (p. 359)

Reserved Instance overview

The following diagram shows a basic overview of purchasing and using Reserved Instances.

In this scenario, you have a running On-Demand Instance (T2) in your account, for which you're currently
paying On-Demand rates. You purchase a Reserved Instance that matches the attributes of your running
instance, and the billing benefit is immediately applied. Next, you purchase a Reserved Instance for
a C4 instance. You do not have any running instances in your account that match the attributes of
this Reserved Instance. In the final step, you launch an instance that matches the attributes of the C4
Reserved Instance, and the billing benefit is immediately applied.

Key variables that determine Reserved Instance pricing

The Reserved Instance pricing is determined by the following key variables.

Instance attributes
A Reserved Instance has four instance attributes that determine its price.

• Instance type: For example, m4.large. This is composed of the instance family (for example, m4) and
the instance size (for example, large).
• Region: The Region in which the Reserved Instance is purchased.

326
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

• Tenancy: Whether your instance runs on shared (default) or single-tenant (dedicated) hardware. For
more information, see Dedicated Instances (p. 496).
• Platform: The operating system; for example, Windows or Linux/Unix. For more information, see
Choosing a platform (p. 340).

Term commitment
You can purchase a Reserved Instance for a one-year or three-year commitment, with the three-year
commitment offering a bigger discount.

• One-year: A year is defined as 31536000 seconds (365 days).

• Three-year: Three years is defined as 94608000 seconds (1095 days).

Reserved Instances do not renew automatically; when they expire, you can continue using the EC2
instance without interruption, but you are charged On-Demand rates. In the above example, when the
Reserved Instances that cover the T2 and C4 instances expire, you go back to paying the On-Demand
rates until you terminate the instances or purchase new Reserved Instances that match the instance
attributes.

Payment options
The following payment options are available for Reserved Instances:

• All Upfront: Full payment is made at the start of the term, with no other costs or additional hourly
charges incurred for the remainder of the term, regardless of hours used.
• Partial Upfront: A portion of the cost must be paid upfront and the remaining hours in the term are
billed at a discounted hourly rate, regardless of whether the Reserved Instance is being used.
• No Upfront: You are billed a discounted hourly rate for every hour within the term, regardless of
whether the Reserved Instance is being used. No upfront payment is required.
Note
No Upfront Reserved Instances are based on a contractual obligation to pay monthly for the
entire term of the reservation. For this reason, a successful billing history is required before
you can purchase No Upfront Reserved Instances.

Generally speaking, you can save more money making a higher upfront payment for Reserved Instances.
You can also find Reserved Instances offered by third-party sellers at lower prices and shorter term
lengths on the Reserved Instance Marketplace. For more information, see Sell in the Reserved Instance
Marketplace (p. 346).

Offering class
If your computing needs change, you might be able to modify or exchange your Reserved Instance,
depending on the offering class.

• Standard: These provide the most significant discount, but can only be modified. Standard Reserved
Instances can't be exchanged.
• Convertible: These provide a lower discount than Standard Reserved Instances, but can be exchanged
for another Convertible Reserved Instance with different instance attributes. Convertible Reserved
Instances can also be modified.

For more information, see Types of Reserved Instances (offering classes) (p. 329).

After you purchase a Reserved Instance, you cannot cancel your purchase. However, you might be able to
modify (p. 352), exchange (p. 359), or sell (p. 346) your Reserved Instance if your needs change.

327
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

For more information, see the Amazon EC2 Reserved Instances Pricing page.

Reserved Instance limits

There is a limit to the number of Reserved Instances that you can purchase per month. For each
Region you can purchase 20 regional (p. 330) Reserved Instances per month plus an additional 20
zonal (p. 330) Reserved Instances per month for each Availability Zone.

For example, in a Region with three Availability Zones, the limit is 80 Reserved Instances per month:
20 regional Reserved Instances for the Region plus 20 zonal Reserved Instances for each of the three
Availability Zones (20x3=60).

A regional Reserved Instance applies a discount to a running On-Demand Instance. The default On-
Demand Instance limit is 20. You cannot exceed your running On-Demand Instance limit by purchasing
regional Reserved Instances. For example, if you already have 20 running On-Demand Instances, and you
purchase 20 regional Reserved Instances, the 20 regional Reserved Instances are used to apply a discount
to the 20 running On-Demand Instances. If you purchase more regional Reserved Instances, you will not
be able to launch more instances because you have reached your On-Demand Instance limit.

Before purchasing regional Reserved Instances, make sure your On-Demand Instance limit matches or
exceeds the number of regional Reserved Instances you intend to own. If required, make sure you request
an increase to your On-Demand Instance limit before purchasing more regional Reserved Instances.

A zonal Reserved Instance—a Reserved Instance that is purchased for a specific Availability Zone—
provides capacity reservation as well as a discount. You can exceed your running On-Demand Instance
limit by purchasing zonal Reserved Instances. For example, if you already have 20 running On-Demand
Instances, and you purchase 20 zonal Reserved Instances, you can launch a further 20 On-Demand
Instances that match the specifications of your zonal Reserved Instances, giving you a total of 40 running
instances.

The Amazon EC2 console provides limit information. For more information, see View your current
limits (p. 1323).

Regional and zonal Reserved Instances (scope)

When you purchase a Reserved Instance, you determine the scope of the Reserved Instance. The scope is
either regional or zonal.

• Regional: When you purchase a Reserved Instance for a Region, it's referred to as a regional Reserved
Instance.
• Zonal: When you purchase a Reserved Instance for a specific Availability Zone, it's referred to as a
zonal Reserved Instance.

The scope does not affect the price. You pay the same price for a regional or zonal Reserved Instance. For
more information about Reserved Instance pricing, see Key variables that determine Reserved Instance
pricing (p. 326) and Amazon EC2 Reserved Instances Pricing.

Differences between regional and zonal Reserved Instances

The following table highlights some key differences between regional Reserved Instances and zonal
Reserved Instances:

  Regional Reserved Instances Zonal Reserved Instances

Ability to reserve capacity A regional Reserved Instance A zonal Reserved Instance

does not reserve capacity. reserves capacity in the specified
Availability Zone.

328
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

  Regional Reserved Instances Zonal Reserved Instances

Availability Zone flexibility The Reserved Instance discount No Availability Zone flexibility—
applies to instance usage in any the Reserved Instance discount
Availability Zone in the specified applies to instance usage in the
Region. specified Availability Zone only.

Instance size flexibility The Reserved Instance discount No instance size flexibility—
applies to instance usage within the Reserved Instance discount
the instance family, regardless of applies to instance usage for the
size. Only supported on Amazon specified instance type and size
Linux/Unix Reserved Instances only.
with default tenancy. For more
information, see Instance
size flexibility determined by
normalization factor (p. 330).

Queuing a purchase You can queue purchases for You can't queue purchases for
regional Reserved Instances. zonal Reserved Instances.

For more information and examples, see How Reserved Instances are applied (p. 330).

Types of Reserved Instances (offering classes)

The offering class of a Reserved Instance is either Standard or Convertible. A Standard Reserved Instance
provides a more significant discount than a Convertible Reserved Instance, but you can't exchange a
Standard Reserved Instance. You can exchange Convertible Reserved Instances. You can modify Standard
and Convertible Reserved Instances.

The configuration of a Reserved Instance comprises a single instance type, platform, scope, and tenancy
over a term. If your computing needs change, you might be able to modify or exchange your Reserved
Instance.

Differences between Standard and Convertible Reserved Instances

The following are the differences between Standard and Convertible Reserved Instances.

  Standard Reserved Instance Convertible Reserved Instance

Modifying Reserved Instances Some attributes can be Some attributes can be

modified. For more information, modified. For more information,
see Modify Reserved see Modify Reserved
Instances (p. 352). Instances (p. 352).

Exchanging Reserved Instances Can't be exchanged. Can be exchanged during the

term for another Convertible
Reserved Instance with
new attributes, including
instance family, instance type,
platform, scope, or tenancy.
For more information, see
Exchange Convertible Reserved
Instances (p. 359).

Selling in the Reserved Instance Can be sold in the Reserved Can't be sold in the Reserved
Marketplace Instance Marketplace. Instance Marketplace.

329
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

  Standard Reserved Instance Convertible Reserved Instance

Buying in the Reserved Instance Can be bought in the Reserved Can't be bought in the Reserved
Marketplace Instance Marketplace. Instance Marketplace.

How Reserved Instances are applied

If you purchase a Reserved Instance and you already have a running instance that matches the
specifications of the Reserved Instance, the billing benefit is immediately applied. You do not have to
restart your instances. If you do not have an eligible running instance, launch an instance and ensure that
you match the same criteria that you specified for your Reserved Instance. For more information, see Use
your Reserved Instances (p. 335).

Reserved Instances apply to usage in the same manner, irrespective of the offering type (Standard or
Convertible), and are automatically applied to running On-Demand Instances with matching attributes.

How zonal Reserved Instances are applied

Reserved Instances assigned to a specific Availability Zone provide the Reserved Instance discount
to matching instance usage in that Availability Zone. For example, if you purchase two c4.xlarge
default tenancy Linux/Unix Standard Reserved Instances in Availability Zone us-east-1a, then up to two
c4.xlarge default tenancy Linux/Unix instances running in the Availability Zone us-east-1a can benefit
from the Reserved Instance discount. The attributes (tenancy, platform, Availability Zone, instance type,
and instance size) of the running instances must match that of the Reserved Instances.

How regional Reserved Instances are applied

Regional Reserved Instances are purchased for a Region and provide Availability Zone flexibility. The
Reserved Instance discount applies to instance usage in any Availability Zone in that Region.

Regional Reserved Instances also provide instance size flexibility where the Reserved Instance discount
applies to instance usage within the instance family, regardless of size.

Limitations for instance size flexibility

Instance size flexibility does not apply to the following Reserved Instances:

• Reserved Instances that are purchased for a specific Availability Zone (zonal Reserved Instances)
• Reserved Instances with dedicated tenancy
• Reserved Instances for Windows Server, Windows Server with SQL Standard, Windows Server with SQL
Server Enterprise, Windows Server with SQL Server Web, RHEL, and SUSE Linux Enterprise Server
• Reserved Instances for G4dn instances

Instance size flexibility determined by normalization factor

Instance size flexibility is determined by the normalization factor of the instance size. The discount
applies either fully or partially to running instances of the same instance family, depending on the
instance size of the reservation, in any Availability Zone in the Region. The only attributes that must be
matched are the instance family, tenancy, and platform.

Instance size flexibility is applied from the smallest to the largest instance size within the instance family
based on the normalization factor.

The following table lists the different sizes within an instance family, and the corresponding
normalization factor per hour. This scale is used to apply the discounted rate of Reserved Instances to
the normalized usage of the instance family.

330
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

Instance size Normalization factor

nano 0.25

micro 0.5

small 1

medium 2

large 4

xlarge 8

2xlarge 16

3xlarge 24

4xlarge 32

6xlarge 48

8xlarge 64

9xlarge 72

10xlarge 80

12xlarge 96

16xlarge 128

18xlarge 144

24xlarge 192

32xlarge 256

For example, a t2.medium instance has a normalization factor of 2. If you purchase a t2.medium
default tenancy Amazon Linux/Unix Reserved Instance in the US East (N. Virginia) and you have two
running t2.small instances in your account in that Region, the billing benefit is applied in full to both
instances.

331
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

Or, if you have one t2.large instance running in your account in the US East (N. Virginia) Region, the
billing benefit is applied to 50% of the usage of the instance.

The normalization factor is also applied when modifying Reserved Instances. For more information, see
Modify Reserved Instances (p. 352).

Normalization factor for bare metal instances

Instance size flexibility also applies to bare metal instances within the instance family. If you have
regional Amazon Linux/Unix Reserved Instances with shared tenancy on bare metal instances, you can
benefit from the Reserved Instance savings within the same instance family. The opposite is also true: if
you have regional Amazon Linux/Unix Reserved Instances with shared tenancy on instances in the same
family as a bare metal instance, you can benefit from the Reserved Instance savings on the bare metal
instance.

A bare metal instance is the same size as the largest instance within the same instance family. For
example, an i3.metal is the same size as an i3.16xlarge, so they have the same normalization factor.
Note
The .metal instance sizes do not have a single normalization factor. They vary based on the
specific instance family.

Bare metal instance size Normalization factor

a1.metal 32

c5.metal 192

c5d.metal 192

c5n.metal 144

c6g.metal 128

c6gd.metal 128

g4dn.metal 128

i3.metal 128

i3en.metal 192

m5.metal 192

332
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

Bare metal instance size Normalization factor

m5d.metal 192

m6g.metal 128

m6gd.metal 128

r5.metal 192

r5d.metal 192

r6g.metal 128

r6gd.metal 128

z1d.metal 96

For example, an i3.metal instance has a normalization factor of 128. If you purchase an i3.metal
default tenancy Amazon Linux/Unix Reserved Instance in the US East (N. Virginia), the billing benefit can
apply as follows:

• If you have one running i3.16xlarge in your account in that Region, the billing benefit is applied in
full to the i3.16xlarge instance (i3.16xlarge normalization factor = 128).
• Or, if you have two running i3.8xlarge instances in your account in that Region, the billing benefit is
applied in full to both i3.8xlarge instances (i3.8xlarge normalization factor = 64).
• Or, if you have four running i3.4xlarge instances in your account in that Region, the billing benefit
is applied in full to all four i3.4xlarge instances (i3.4xlarge normalization factor = 32).

The opposite is also true. For example, if you purchase two i3.8xlarge default tenancy Amazon Linux/
Unix Reserved Instances in the US East (N. Virginia), and you have one running i3.metal instance in
that Region, the billing benefit is applied in full to the i3.metal instance.

Examples of applying Reserved Instances

The following scenarios cover the ways in which Reserved Instances are applied.

Example Scenario 1: Reserved Instances in a single account

You are running the following On-Demand Instances in account A:

• 4 x m3.large Linux, default tenancy instances in Availability Zone us-east-1a

• 2 x m4.xlarge Amazon Linux, default tenancy instances in Availability Zone us-east-1b
• 1 x c4.xlarge Amazon Linux, default tenancy instances in Availability Zone us-east-1c

You purchase the following Reserved Instances in account A:

• 4 x m3.large Linux, default tenancy Reserved Instances in Availability Zone us-east-1a (capacity is
reserved)
• 4 x m4.large Amazon Linux, default tenancy Reserved Instances in Region us-east-1
• 1 x c4.large Amazon Linux, default tenancy Reserved Instances in Region us-east-1

The Reserved Instance benefits are applied in the following way:

333
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

• The discount and capacity reservation of the four m3.large zonal Reserved Instances is used by the
four m3.large instances because the attributes (instance size, Region, platform, tenancy) between
them match.
• The m4.large regional Reserved Instances provide Availability Zone and instance size flexibility,
because they are regional Amazon Linux Reserved Instances with default tenancy.

An m4.large is equivalent to 4 normalized units/hour.

You've purchased four m4.large regional Reserved Instances, and in total, they are equal to 16
normalized units/hour (4x4). Account A has two m4.xlarge instances running, which is equivalent to
16 normalized units/hour (2x8). In this case, the four m4.large regional Reserved Instances provide
the billing benefit to an entire hour of usage of the two m4.xlarge instances.
• The c4.large regional Reserved Instance in us-east-1 provides Availability Zone and instance size
flexibility, because it is a regional Amazon Linux Reserved Instance with default tenancy, and applies
to the c4.xlarge instance. A c4.large instance is equivalent to 4 normalized units/hour and a
c4.xlarge is equivalent to 8 normalized units/hour.

In this case, the c4.large regional Reserved Instance provides partial benefit to c4.xlarge usage.
This is because the c4.large Reserved Instance is equivalent to 4 normalized units/hour of usage,
but the c4.xlarge instance requires 8 normalized units/hour. Therefore, the c4.large Reserved
Instance billing discount applies to 50% of c4.xlarge usage. The remaining c4.xlarge usage is
charged at the On-Demand rate.

Example Scenario 2: Regional Reserved Instances in linked accounts

Reserved Instances are first applied to usage within the purchasing account, followed by qualifying usage
in any other account in the organization. For more information, see Reserved Instances and consolidated
billing (p. 338). For regional Reserved Instances that offer instance size flexibility, the benefit is applied
from the smallest to the largest instance size within the instance family.

You're running the following On-Demand Instances in account A (the purchasing account):

• 2 x m4.xlarge Linux, default tenancy instances in Availability Zone us-east-1a

• 1 x m4.2xlarge Linux, default tenancy instances in Availability Zone us-east-1b
• 2 x c4.xlarge Linux, default tenancy instances in Availability Zone us-east-1a
• 1 x c4.2xlarge Linux, default tenancy instances in Availability Zone us-east-1b

Another customer is running the following On-Demand Instances in account B—a linked account:

• 2 x m4.xlarge Linux, default tenancy instances in Availability Zone us-east-1a

You purchase the following regional Reserved Instances in account A:

• 4 x m4.xlarge Linux, default tenancy Reserved Instances in Region us-east-1

• 2 x c4.xlarge Linux, default tenancy Reserved Instances in Region us-east-1

The regional Reserved Instance benefits are applied in the following way:

• The discount of the four m4.xlarge Reserved Instances is used by the two m4.xlarge instances
and the single m4.2xlarge instance in account A (purchasing account). All three instances match
the attributes (instance family, Region, platform, tenancy). The discount is applied to instances in the
purchasing account (account A) first, even though account B (linked account) has two m4.xlarge that
also match the Reserved Instances. There is no capacity reservation because the Reserved Instances are
regional Reserved Instances.

334
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

• The discount of the two c4.xlarge Reserved Instances applies to the two c4.xlarge instances,
because they are a smaller instance size than the c4.2xlarge instance. There is no capacity
reservation because the Reserved Instances are regional Reserved Instances.

Example Scenario 3: Zonal Reserved Instances in a linked account

In general, Reserved Instances that are owned by an account are applied first to usage in that account.
However, if there are qualifying, unused Reserved Instances for a specific Availability Zone (zonal
Reserved Instances) in other accounts in the organization, they are applied to the account before regional
Reserved Instances owned by the account. This is done to ensure maximum Reserved Instance utilization
and a lower bill. For billing purposes, all the accounts in the organization are treated as one account. The
following example might help explain this.

You're running the following On-Demand Instance in account A (the purchasing account):

• 1 x m4.xlarge Linux, default tenancy instance in Availability Zone us-east-1a

A customer is running the following On-Demand Instance in linked account B:

• 1 x m4.xlarge Linux, default tenancy instance in Availability Zone us-east-1b

You purchase the following regional Reserved Instances in account A:

• 1 x m4.xlarge Linux, default tenancy Reserved Instance in Region us-east-1

A customer also purchases the following zonal Reserved Instances in linked account C:

• 1 x m4.xlarge Linux, default tenancy Reserved Instances in Availability Zone us-east-1a

The Reserved Instance benefits are applied in the following way:

• The discount of the m4.xlarge zonal Reserved Instance owned by account C is applied to the
m4.xlarge usage in account A.
• The discount of the m4.xlarge regional Reserved Instance owned by account A is applied to the
m4.xlarge usage in account B.
• If the regional Reserved Instance owned by account A was first applied to the usage in account A, the
zonal Reserved Instance owned by account C remains unused and usage in account B is charged at On-
Demand rates.

For more information, see Reserved Instances in the Billing and Cost Management Report.

Use your Reserved Instances

Reserved Instances are automatically applied to running On-Demand Instances provided that the
specifications match. If you have no running On-Demand Instances that match the specifications of
your Reserved Instance, the Reserved Instance is unused until you launch an instance with the required
specifications.

If you're launching an instance to take advantage of the billing benefit of a Reserved Instance, ensure
that you specify the following information during launch:

• Platform: You must choose an Amazon Machine Image (AMI) that matches the platform (product
description) of your Reserved Instance. For example, if you specified Linux/UNIX, you can launch an
instance from an Amazon Linux AMI or an Ubuntu AMI.

335
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

• Instance type: Specify the same instance type as your Reserved Instance; for example, t2.large.
• Availability Zone: If you purchased a zonal Reserved Instance for a specific Availability Zone, you must
launch the instance into the same Availability Zone. If you purchased a regional Reserved Instance, you
can launch your instance into any Availability Zone.
• Tenancy: The tenancy of your instance must match the tenancy of the Reserved Instance; for example,
dedicated or shared. For more information, see Dedicated Instances (p. 496).

For more information, see Launch an instance using the Launch Instance Wizard (p. 526). For examples
of how Reserved Instances are applied to your running instances, see How Reserved Instances are
applied (p. 330).

You can use Amazon EC2 Auto Scaling or other AWS services to launch the On-Demand Instances that
use your Reserved Instance benefits. For more information, see the Amazon EC2 Auto Scaling User Guide.

How you are billed

All Reserved Instances provide you with a discount compared to On-Demand pricing. With Reserved
Instances, you pay for the entire term regardless of actual use. You can choose to pay for your Reserved
Instance upfront, partially upfront, or monthly, depending on the payment option (p. 327) specified for
the Reserved Instance.

When Reserved Instances expire, you are charged On-Demand rates for EC2 instance usage. You can
queue a Reserved Instance for purchase up to three years in advance. This can help you ensure that you
have uninterrupted coverage. For more information, see Queue your purchase (p. 341).

The AWS Free Tier is available for new AWS accounts. If you are using the AWS Free Tier to run Amazon
EC2 instances, and you purchase a Reserved Instance, you are charged under standard pricing guidelines.
For information, see AWS Free Tier.

Contents
• Usage billing (p. 336)
• Viewing your bill (p. 337)
• Reserved Instances and consolidated billing (p. 338)
• Reserved Instance discount pricing tiers (p. 338)

Usage billing
Reserved Instances are billed for every clock-hour during the term that you select, regardless of whether
an instance is running. Each clock-hour starts on the hour (zero minutes and zero seconds past the hour)
of a standard 24-hour clock. For example, 1:00:00 to 1:59:59 is one clock-hour. For more information
about instance states, see Instance lifecycle (p. 520).

A Reserved Instance billing benefit can be applied to a running instance on a per-second basis. Per-
second billing is available for instances using an open-source Linux distribution, such as Amazon Linux
and Ubuntu. Per-hour billing is used for commercial Linux distributions, such as Red Hat Enterprise Linux
and SUSE Linux Enterprise Server.

A Reserved Instance billing benefit can apply to a maximum of 3600 seconds (one hour) of instance
usage per clock-hour. You can run multiple instances concurrently, but can only receive the benefit of the
Reserved Instance discount for a total of 3600 seconds per clock-hour; instance usage that exceeds 3600
seconds in a clock-hour is billed at the On-Demand rate.

For example, if you purchase one m4.xlarge Reserved Instance and run four m4.xlarge instances
concurrently for one hour, one instance is charged at one hour of Reserved Instance usage and the other
three instances are charged at three hours of On-Demand usage.

336
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

However, if you purchase one m4.xlarge Reserved Instance and run four m4.xlarge instances for 15
minutes (900 seconds) each within the same hour, the total running time for the instances is one hour,
which results in one hour of Reserved Instance usage and 0 hours of On-Demand usage.

If multiple eligible instances are running concurrently, the Reserved Instance billing benefit is applied
to all the instances at the same time up to a maximum of 3600 seconds in a clock-hour; thereafter, On-
Demand rates apply.

Cost Explorer on the Billing and Cost Management console enables you to analyze the savings against
running On-Demand Instances. The Reserved Instances FAQ includes an example of a list value
calculation.

If you close your AWS account, On-Demand billing for your resources stops. However, if you have any
Reserved Instances in your account, you continue to receive a bill for these until they expire.

Viewing your bill

You can find out about the charges and fees to your account by viewing the AWS Billing and Cost
Management console.

• The Dashboard displays a spend summary for your account.

• On the Bills page, under Details expand the Elastic Compute Cloud section and the Region to get
billing information about your Reserved Instances.

You can view the charges online, or you can download a CSV file.

337
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

You can also track your Reserved Instance utilization using the AWS Cost and Usage Report. For
more information, see Reserved Instances under Cost and Usage Report in the AWS Billing and Cost
Management User Guide.

Reserved Instances and consolidated billing

The pricing benefits of Reserved Instances are shared when the purchasing account is part of a set of
accounts billed under one consolidated billing payer account. The instance usage across all member
accounts is aggregated in the payer account every month. This is typically useful for companies in which
there are different functional teams or groups; then, the normal Reserved Instance logic is applied to
calculate the bill. For more information, see Consolidated Billing and AWS Organizations in the AWS
Organizations User Guide.

If you close the account that purchased the Reserved Instance, the payer account will continue being
charged for the Reserved Instance until either the Reserved Instance expires or the closed account is
permanently deleted. The closed account is permanently deleted after 90 days. After it is deleted, the
member accounts will stop benefitting from the Reserved Instance billing discount. For more information
about closing an account, see Closing an AWS Account in the AWS Organizations User Guide.

Reserved Instance discount pricing tiers

If your account qualifies for a discount pricing tier, it automatically receives discounts on upfront and
instance usage fees for Reserved Instance purchases that you make within that tier level from that point
on. To qualify for a discount, the list value of your Reserved Instances in the Region must be $500,000
USD or more.

The following rules apply:

• Pricing tiers and related discounts apply only to purchases of Amazon EC2 Standard Reserved
Instances.
• Pricing tiers do not apply to Reserved Instances for Windows with SQL Server Standard, SQL Server
Web, and SQL Server Enterprise.
• Pricing tiers do not apply to Reserved Instances for Linux with SQL Server Standard, SQL Server Web,
and SQL Server Enterprise.
• Pricing tier discounts only apply to purchases made from AWS. They do not apply to purchases of
third-party Reserved Instances.
• Discount pricing tiers are currently not applicable to Convertible Reserved Instance purchases.

Topics
• Calculate Reserved Instance pricing discounts (p. 338)
• Buy with a discount tier (p. 339)
• Crossing pricing tiers (p. 339)
• Consolidated billing for pricing tiers (p. 340)

Calculate Reserved Instance pricing discounts

You can determine the pricing tier for your account by calculating the list value for all of your Reserved
Instances in a Region. Multiply the hourly recurring price for each reservation by the total number of
hours for the term and add the undiscounted upfront price (also known as the fixed price) at the time of
purchase. Because the list value is based on undiscounted (public) pricing, it is not affected if you qualify
for a volume discount or if the price drops after you buy your Reserved Instances.

List value = fixed price + (undiscounted recurring hourly price * hours in term)

338
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

For example, for a 1-year Partial Upfront t2.small Reserved Instance, assume the upfront price is
$60.00 and the hourly rate is $0.007. This provides a list value of $121.32.

121.32 = 60.00 + (0.007 * 8760)

To view the fixed price values for Reserved Instances using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Reserved Instances.
3. Display the Upfront Price column by choosing Show/Hide Columns (the gear-shaped icon) in the
top right corner.

To view the fixed price values for Reserved Instances using the command line

• describe-reserved-instances (AWS CLI)

• Get-EC2ReservedInstance (AWS Tools for Windows PowerShell)
• DescribeReservedInstances (Amazon EC2 API)

Buy with a discount tier

When you buy Reserved Instances, Amazon EC2 automatically applies any discounts to the part of your
purchase that falls within a discount pricing tier. You don't need to do anything differently, and you can
buy Reserved Instances using any of the Amazon EC2 tools. For more information, see Buying Reserved
Instances (p. 340).

After the list value of your active Reserved Instances in a Region crosses into a discount pricing tier,
any future purchase of Reserved Instances in that Region are charged at a discounted rate. If a single
purchase of Reserved Instances in a Region takes you over the threshold of a discount tier, then the
portion of the purchase that is above the price threshold is charged at the discounted rate. For more
information about the temporary Reserved Instance IDs that are created during the purchase process,
see Crossing pricing tiers (p. 339).

If your list value falls below the price point for that discount pricing tier—for example, if some of your
Reserved Instances expire—future purchases of Reserved Instances in the Region are not discounted.
However, you continue to get the discount applied against any Reserved Instances that were originally
purchased within the discount pricing tier.

When you buy Reserved Instances, one of four possible scenarios occurs:

• No discount—Your purchase within a Region is still below the discount threshold.

• Partial discount—Your purchase within a Region crosses the threshold of the first discount tier. No
discount is applied to one or more reservations and the discounted rate is applied to the remaining
reservations.
• Full discount—Your entire purchase within a Region falls within one discount tier and is discounted
appropriately.
• Two discount rates—Your purchase within a Region crosses from a lower discount tier to a higher
discount tier. You are charged two different rates: one or more reservations at the lower discounted
rate, and the remaining reservations at the higher discounted rate.

Crossing pricing tiers

If your purchase crosses into a discounted pricing tier, you see multiple entries for that purchase: one for
that part of the purchase charged at the regular price, and another for that part of the purchase charged
at the applicable discounted rate.

339
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

The Reserved Instance service generates several Reserved Instance IDs because your purchase crossed
from an undiscounted tier, or from one discounted tier to another. There is an ID for each set of
reservations in a tier. Consequently, the ID returned by your purchase CLI command or API action is
different from the actual ID of the new Reserved Instances.

Consolidated billing for pricing tiers

A consolidated billing account aggregates the list value of member accounts within a Region. When
the list value of all active Reserved Instances for the consolidated billing account reaches a discount
pricing tier, any Reserved Instances purchased after this point by any member of the consolidated
billing account are charged at the discounted rate (as long as the list value for that consolidated account
stays above the discount pricing tier threshold). For more information, see Reserved Instances and
consolidated billing (p. 338).

Buying Reserved Instances

To purchase a Reserved Instance, search for Reserved Instance offerings from AWS and third-party sellers,
adjusting your search parameters until you find the exact match that you're looking for.

When you search for Reserved Instances to buy, you receive a quote on the cost of the returned offerings.
When you proceed with the purchase, AWS automatically places a limit price on the purchase price. The
total cost of your Reserved Instances won't exceed the amount that you were quoted.

If the price rises or changes for any reason, the purchase is not completed. If, at the time of purchase,
there are offerings similar to your choice but at a lower price, AWS sells you the offerings at the lower
price.

Before you confirm your purchase, review the details of the Reserved Instance that you plan to buy, and
make sure that all the parameters are accurate. After you purchase a Reserved Instance (either from a
third-party seller in the Reserved Instance Marketplace or from AWS), you cannot cancel your purchase.
Note
To purchase and modify Reserved Instances, ensure that your IAM user account has the
appropriate permissions, such as the ability to describe Availability Zones. For information,
see Example Policies for Working With the AWS CLI or an AWS SDK and Example Policies for
Working in the Amazon EC2 Console.

Topics
• Choosing a platform (p. 340)
• Queue your purchase (p. 341)
• Buy Standard Reserved Instances (p. 341)
• Buy Convertible Reserved Instances (p. 343)
• Buy from the Reserved Instance Marketplace (p. 345)
• View your Reserved Instances (p. 345)
• Cancel a queued purchase (p. 345)
• Renew a Reserved Instance (p. 346)

Choosing a platform
Amazon EC2 supports the following Linux platforms for Reserved Instances:

• Linux/UNIX
• Linux with SQL Server Standard
• Linux with SQL Server Web

340
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

• Linux with SQL Server Enterprise

• SUSE Linux
• Red Hat Enterprise Linux

When you purchase a Reserved Instance, you must choose an offering for a platform that represents the
operating system for your instance.

• For SUSE Linux and RHEL distributions, you must choose offerings for those specific platforms, i.e., for
the SUSE Linux or Red Hat Enterprise Linux platforms.
• For all other Linux distributions (including Ubuntu), choose an offering for the Linux/UNIX platform.
• If you bring your existing RHEL subscription, you must choose an offering for the Linux/UNIX
platform, not an offering for the Red Hat Enterprise Linux platform.

Important
If you plan to purchase a Reserved Instance to apply to an On-Demand Instance that was
launched from an AWS Marketplace AMI, first check the PlatformDetails field of the AMI.
The PlatformDetails field indicates which Reserved Instance to purchase. The platform
details of the AMI must match the platform of the Reserved Instance, otherwise the Reserved
Instance will not be applied to the On-Demand Instance. For information about how to view the
platform details of the AMI, see Understanding AMI billing information (p. 172).

For information about the supported platforms for Windows, see Choosing a platform in the Amazon
EC2 User Guide for Windows Instances.

Queue your purchase

By default, when you purchase a Reserved Instance, the purchase is made immediately. Alternatively,
you can queue your purchases for a future date and time. For example, you can queue a purchase for
around the time that an existing Reserved Instance expires. This can help you ensure that you have
uninterrupted coverage.

You can queue purchases for regional Reserved Instances, but not zonal Reserved Instances or Reserved
Instances from other sellers. You can queue a purchase up to three years in advance. On the scheduled
date and time, the purchase is made using the default payment method. After the payment is successful,
the billing benefit is applied.

You can view your queued purchases in the Amazon EC2 console. The status of a queued purchase is
queued. You can cancel a queued purchase any time before its scheduled time. For details, see Cancel a
queued purchase (p. 345).

Buy Standard Reserved Instances

You can buy Standard Reserved Instances in a specific Availability Zone and get a capacity reservation.
Alternatively, you can forego the capacity reservation and purchase a regional Standard Reserved
Instance.

To buy Standard Reserved Instances using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Reserved Instances, and then choose Purchase Reserved Instances.
3. For Offering Class, choose Standard to display Standard Reserved Instances.
4. To purchase a capacity reservation, choose Only show offerings that reserve capacity in the top-
right corner of the purchase screen. To purchase a regional Reserved Instance, leave the check box
unselected.

341
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

5. Select other configurations as needed and choose Search.

To purchase a Standard Reserved Instance from the Reserved Instance Marketplace, look for 3rd
Party in the Seller column in the search results. The Term column displays non-standard terms.
6. For each Reserved Instance that you want to purchase, enter the quantity, and choose Add to Cart.
7. To see a summary of the Reserved Instances that you selected, choose View Cart.
8. If Order On is Now, the purchase is completed immediately. To queue a purchase, choose Now and
select a date. You can select a different date for each eligible offering in the cart. The purchase is
queued until 00:00 UTC on the selected date.
9. To complete the order, choose Order.

If, at the time of placing the order, there are offerings similar to your choice but with a lower price,
AWS sells you the offerings at the lower price.
10. Choose Close.

The status of your order is listed in the State column. When your order is complete, the State value
changes from payment-pending to active. When the Reserved Instance is active, it is ready to
use.

Note
If the status goes to retired, AWS might not have received your payment.

To buy a Standard Reserved Instance using the AWS CLI

1. Find available Reserved Instances using the describe-reserved-instances-offerings command. Specify

standard for the --offering-class parameter to return only Standard Reserved Instances.
You can apply additional parameters to narrow your results. For example, if you want to purchase
a regional t2.large Reserved Instance with a default tenancy for Linux/UNIX for a 1-year term
only:

aws ec2 describe-reserved-instances-offerings \

--instance-type t2.large \
--offering-class standard \
--product-description "Linux/UNIX" \
--instance-tenancy default \
--filters Name=duration,Values=31536000 Name=scope,Values=Region

To find Reserved Instances on the Reserved Instance Marketplace only, use the marketplace filter
and do not specify a duration in the request, as the term might be shorter than a 1– or 3-year term.

aws ec2 describe-reserved-instances-offerings \

--instance-type t2.large \
--offering-class standard \
--product-description "Linux/UNIX" \
--instance-tenancy default \
--filters Name=marketplace,Values=true

When you find a Reserved Instance that meets your needs, take note of the offering ID. For example:

"ReservedInstancesOfferingId": "bec624df-a8cc-4aad-a72f-4f8abc34caf2"

2. Use the purchase-reserved-instances-offering command to buy your Reserved Instance. You must
specify the Reserved Instance offering ID you obtained the previous step and you must specify the
number of instances for the reservation.

aws ec2 purchase-reserved-instances-offering \

342
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

--reserved-instances-offering-id bec624df-a8cc-4aad-a72f-4f8abc34caf2 \
--instance-count 1

By default, the purchase is completed immediately. Alternatively, to queue the purchase, add the
following parameter to the previous call.

--purchase-time "2020-12-01T00:00:00Z"

3. Use the describe-reserved-instances command to get the status of your Reserved Instance.

aws ec2 describe-reserved-instances

Alternatively, use the following AWS Tools for Windows PowerShell commands:

• Get-EC2ReservedInstancesOffering
• New-EC2ReservedInstance
• Get-EC2ReservedInstance

After the purchase is complete, if you already have a running instance that matches the specifications
of the Reserved Instance, the billing benefit is immediately applied. You do not have to restart your
instances. If you do not have a suitable running instance, launch an instance and ensure that you match
the same criteria that you specified for your Reserved Instance. For more information, see Use your
Reserved Instances (p. 335).

For examples of how Reserved Instances are applied to your running instances, see How Reserved
Instances are applied (p. 330).

Buy Convertible Reserved Instances

You can buy Convertible Reserved Instances in a specific Availability Zone and get a capacity reservation.
Alternatively, you can forego the capacity reservation and purchase a regional Convertible Reserved
Instance.

To buy Convertible Reserved Instances using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Reserved Instances, and then choose Purchase Reserved Instances.
3. For Offering Class, choose Convertible to display Convertible Reserved Instances.
4. To purchase a capacity reservation, choose Only show offerings that reserve capacity in the top-
right corner of the purchase screen. To purchase a regional Reserved Instance, leave the check box
unselected.
5. Select other configurations as needed and choose Search.
6. For each Convertible Reserved Instance that you want to purchase, enter the quantity, and choose
Add to Cart.
7. To see a summary of your selection, choose View Cart.
8. If Order On is Now, the purchase is completed immediately. To queue a purchase, choose Now and
select a date. You can select a different date for each eligible offering in the cart. The purchase is
queued until 00:00 UTC on the selected date.
9. To complete the order, choose Order.

If, at the time of placing the order, there are offerings similar to your choice but with a lower price,
AWS sells you the offerings at the lower price.

343
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

10. Choose Close.

The status of your order is listed in the State column. When your order is complete, the State value
changes from payment-pending to active. When the Reserved Instance is active, it is ready to
use.

Note
If the status goes to retired, AWS might not have received your payment.

To buy a Convertible Reserved Instance using the AWS CLI

1. Find available Reserved Instances using the describe-reserved-instances-offerings command.

Specify convertible for the --offering-class parameter to return only Convertible Reserved
Instances. You can apply additional parameters to narrow your results; for example, if you want to
purchase a regional t2.large Reserved Instance with a default tenancy for Linux/UNIX:

aws ec2 describe-reserved-instances-offerings \

--instance-type t2.large \
--offering-class convertible \
--product-description "Linux/UNIX" \
--instance-tenancy default \
--filters Name=scope,Values=Region

When you find a Reserved Instance that meets your needs, take note of the offering ID. For example:

"ReservedInstancesOfferingId": "bec624df-a8cc-4aad-a72f-4f8abc34caf2"

2. Use the purchase-reserved-instances-offering command to buy your Reserved Instance. You must
specify the Reserved Instance offering ID you obtained the previous step and you must specify the
number of instances for the reservation.

aws ec2 purchase-reserved-instances-offering \

--reserved-instances-offering-id bec624df-a8cc-4aad-a72f-4f8abc34caf2 \
--instance-count 1

By default, the purchase is completed immediately. Alternatively, to queue the purchase, add the
following parameter to the previous call.

--purchase-time "2020-12-01T00:00:00Z"

3. Use the describe-reserved-instances command to get the status of your Reserved Instance.

aws ec2 describe-reserved-instances

Alternatively, use the following AWS Tools for Windows PowerShell commands:

• Get-EC2ReservedInstancesOffering
• New-EC2ReservedInstance
• Get-EC2ReservedInstance

If you already have a running instance that matches the specifications of the Reserved Instance, the
billing benefit is immediately applied. You do not have to restart your instances. If you do not have
a suitable running instance, launch an instance and ensure that you match the same criteria that you
specified for your Reserved Instance. For more information, see Use your Reserved Instances (p. 335).

344
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

For examples of how Reserved Instances are applied to your running instances, see How Reserved
Instances are applied (p. 330).

Buy from the Reserved Instance Marketplace

You can purchase Reserved Instances from third-party sellers who own Reserved Instances that they no
longer need from the Reserved Instance Marketplace. You can do this using the Amazon EC2 console
or a command line tool. The process is similar to purchasing Reserved Instances from AWS. For more
information, see Buy Standard Reserved Instances (p. 341).

There are a few differences between Reserved Instances purchased in the Reserved Instance Marketplace
and Reserved Instances purchased directly from AWS:

• Term – Reserved Instances that you purchase from third-party sellers have less than a full standard
term remaining. Full standard terms from AWS run for one year or three years.
• Upfront price – Third-party Reserved Instances can be sold at different upfront prices. The usage or
recurring fees remain the same as the fees set when the Reserved Instances were originally purchased
from AWS.
• Types of Reserved Instances – Only Amazon EC2 Standard Reserved Instances can be purchased
from the Reserved Instance Marketplace. Convertible Reserved Instances, Amazon RDS, and Amazon
ElastiCache Reserved Instances are not available for purchase on the Reserved Instance Marketplace.

Basic information about you is shared with the seller, for example, your ZIP code and country
information.

This information enables sellers to calculate any necessary transaction taxes that they have to remit to
the government (such as sales tax or value-added tax) and is provided as a disbursement report. In rare
circumstances, AWS might have to provide the seller with your email address, so that they can contact
you regarding questions related to the sale (for example, tax questions).

For similar reasons, AWS shares the legal entity name of the seller on the buyer's purchase invoice. If you
need additional information about the seller for tax or related reasons, contact AWS Support.

View your Reserved Instances

You can view the Reserved Instances you've purchased using the Amazon EC2 console, or a command line
tool.

To view your Reserved Instances in the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Reserved Instances.
3. Your active and retired Reserved Instances are listed. The State column displays the state.
4. If you are a seller in the Reserved Instance Marketplace the My Listings tab displays the status of a
reservation that's listed in the Reserved Instance Marketplace (p. 346). For more information, see
Reserved Instance listing states (p. 350).

To view your Reserved Instances using the command line

• describe-reserved-instances (AWS CLI)

• Get-EC2ReservedInstance (Tools for Windows PowerShell)

Cancel a queued purchase

You can queue a purchase up to three years in advance. You can cancel a queued purchase any time
before its scheduled time.

345
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

To cancel a queued purchase

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Reserved Instances.
3. Select one or more Reserved Instances.
4. Choose Actions, Delete Queued Reserved Instances.
5. When prompted for confirmation, choose Yes, Delete.

To cancel a queued purchase using the command line

• delete-queued-reserved-instances (AWS CLI)

• Remove-EC2QueuedReservedInstance (Tools for Windows PowerShell)

Renew a Reserved Instance

You can renew a Reserved Instance before it is scheduled to expire. Renewing a Reserved Instance queues
the purchase of a Reserved Instance with the same configuration until the current Reserved Instance
expires.

To renew an Reserved Instance using a queued purchase

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Reserved Instances.
3. Select one or more Reserved Instances.
4. Choose Actions, Renew Reserved Instances.
5. To complete the order, choose Order.

Sell in the Reserved Instance Marketplace

The Reserved Instance Marketplace is a platform that supports the sale of third-party and AWS
customers' unused Standard Reserved Instances, which vary in term lengths and pricing options. For
example, you might want to sell Reserved Instances after moving instances to a new AWS Region,
changing to a new instance type, ending projects before the term expiration, when your business needs
change, or if you have unneeded capacity.

As soon as you list your Reserved Instances in the Reserved Instance Marketplace, they are available
for potential buyers to find. All Reserved Instances are grouped according to the duration of the term
remaining and the hourly price.

To fulfill a buyer's request, AWS first sells the Reserved Instance with the lowest upfront price in the
specified grouping. Then, AWS sells the Reserved Instance with the next lowest price, until the buyer's
entire order is fulfilled. AWS then processes the transactions and transfers ownership of the Reserved
Instances to the buyer.

You own your Reserved Instance until it's sold. After the sale, you've given up the capacity reservation
and the discounted recurring fees. If you continue to use your instance, AWS charges you the On-
Demand price starting from the time that your Reserved Instance was sold.

If you want to sell your unused Reserved Instances on the Reserved Instance Marketplace, you must meet
certain eligibility criteria.

For information about buying Reserved Instances on the Reserved Instance Marketplace, see Buy from
the Reserved Instance Marketplace (p. 345).

346
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

Contents
• Restrictions and limitations (p. 347)
• Register as a seller (p. 347)
• Bank account for disbursement (p. 348)
• Tax information (p. 348)
• Price your Reserved Instances (p. 349)
• List your Reserved Instances (p. 349)
• Reserved Instance listing states (p. 350)
• Lifecycle of a listing (p. 351)
• After your Reserved Instance is sold (p. 351)
• Getting paid (p. 352)
• Information shared with the buyer (p. 352)

Restrictions and limitations

Before you can sell your unused reservations, you must register as a seller in the Reserved Instance
Marketplace. For information, see Register as a seller (p. 347).

The following limitations and restrictions apply when selling Reserved Instances:

• Only Amazon EC2 Standard Reserved Instances can be sold in the Reserved Instance Marketplace.
Amazon EC2 Convertible Reserved Instances cannot be sold. Reserved Instances for other AWS
services, such as Amazon RDS and Amazon ElastiCache, cannot be sold.
• There must be at least one month remaining in the term of the Standard Reserved Instance.
• You cannot sell a Standard Reserved Instance in a Region that is disabled by default.
• The minimum price allowed in the Reserved Instance Marketplace is $0.00.
• You can sell No Upfront, Partial Upfront, or All Upfront Reserved Instances in the Reserved Instance
Marketplace. If there is an upfront payment on a Reserved Instance, it can be sold only after AWS has
received the upfront payment and the reservation has been active (you've owned it) for at least 30
days.
• You cannot modify your listing in the Reserved Instance Marketplace directly. However, you can
change your listing by first canceling it and then creating another listing with new parameters. For
information, see Price your Reserved Instances (p. 349). You can also modify your Reserved Instances
before listing them. For information, see Modify Reserved Instances (p. 352).
• AWS charges a service fee of 12 percent of the total upfront price of each Standard Reserved Instance
you sell in the Reserved Instance Marketplace. The upfront price is the price the seller is charging for
the Standard Reserved Instance.
• When you register as a seller, the bank you specify must have a US address. For more information, see
Additional seller requirements for paid products in the AWS Marketplace Seller Guide.
• Amazon Internet Services Private Limited (AISPL) customers can't sell Reserved Instances in the
Reserved Instance Marketplace even if they have a US bank account. For more information, see What
are the differences between AWS accounts and AISPL accounts?

Register as a seller
Note
Only the AWS account root user can register an account as a seller.

To sell in the Reserved Instance Marketplace, you must first register as a seller. During registration, you
provide the following information:

347
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

• Bank information—AWS must have your bank information in order to disburse funds collected when
you sell your reservations. The bank you specify must have a US address. For more information, see
Bank account for disbursement (p. 348).
• Tax information—All sellers are required to complete a tax information interview to determine any
necessary tax reporting obligations. For more information, see Tax information (p. 348).

After AWS receives your completed seller registration, you receive an email confirming your registration
and informing you that you can get started selling in the Reserved Instance Marketplace.

Bank account for disbursement

AWS must have your bank information in order to disburse funds collected when you sell your Reserved
Instance. The bank you specify must have a US address. For more information, see Additional seller
requirements for paid products in the AWS Marketplace Seller Guide.

To register a default bank account for disbursements

1. Open the Reserved Instance Marketplace Seller Registration page and sign in using your AWS
credentials.
2. On the Manage Bank Account page, provide the following information about the bank through to
receive payment:

• Bank account holder name

• Routing number
• Account number
• Bank account type

Note
If you are using a corporate bank account, you are prompted to send the information about
the bank account via fax (1-206-765-3424).

After registration, the bank account provided is set as the default, pending verification with the
bank. It can take up to two weeks to verify a new bank account, during which time you can't receive
disbursements. For an established account, it usually takes about two days for disbursements to
complete.

To change the default bank account for disbursement

1. On the Reserved Instance Marketplace Seller Registration page, sign in with the account that you
used when you registered.
2. On the Manage Bank Account page, add a new bank account or modify the default bank account as
needed.

Tax information
Your sale of Reserved Instances might be subject to a transaction-based tax, such as sales tax or value-
added tax. You should check with your business's tax, legal, finance, or accounting department to
determine if transaction-based taxes are applicable. You are responsible for collecting and sending the
transaction-based taxes to the appropriate tax authority.

As part of the seller registration process, you must complete a tax interview in the Seller Registration
Portal. The interview collects your tax information and populates an IRS form W-9, W-8BEN, or W-8BEN-
E, which is used to determine any necessary tax reporting obligations.

348
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

The tax information you enter as part of the tax interview might differ depending on whether you
operate as an individual or business, and whether you or your business are a US or non-US person or
entity. As you fill out the tax interview, keep in mind the following:

• Information provided by AWS, including the information in this topic, does not constitute tax, legal, or
other professional advice. To find out how the IRS reporting requirements might affect your business,
or if you have other questions, contact your tax, legal, or other professional advisor.
• To fulfill the IRS reporting requirements as efficiently as possible, answer all questions and enter all
information requested during the interview.
• Check your answers. Avoid misspellings or entering incorrect tax identification numbers. They can
result in an invalidated tax form.

Based on your tax interview responses and IRS reporting thresholds, Amazon might file Form 1099-K.
Amazon mails a copy of your Form 1099-K on or before January 31 in the year following the year that
your tax account reaches the threshold levels. For example, if your account reaches the threshold in
2018, your Form 1099-K is mailed on or before January 31, 2019.

For more information about IRS requirements and Form 1099-K, see the IRS website.

Price your Reserved Instances

The upfront fee is the only fee that you can specify for the Reserved Instance that you're selling. The
upfront fee is the one-time fee that the buyer pays when they purchase a Reserved Instance.

The following are important limits to note:

• You can sell up to $50,000 in Reserved Instances. To increase this limit, complete the EC2 Reserved
Instance Sales form.
• You can sell up to 5,000 Reserved Instances. To increase this limit, complete the EC2 Reserved
Instance Sales form.
• The minimum price is $0. The minimum allowed price in the Reserved Instance Marketplace is $0.00.

You cannot modify your listing directly. However, you can change your listing by first canceling it and
then creating another listing with new parameters.

You can cancel your listing at any time, as long as it's in the active state. You cannot cancel the listing
if it's already matched or being processed for a sale. If some of the instances in your listing are matched
and you cancel the listing, only the remaining unmatched instances are removed from the listing.

Because the value of Reserved Instances decreases over time, by default, AWS can set prices to decrease
in equal increments month over month. However, you can set different upfront prices based on when
your reservation sells.

For example, if your Reserved Instance has nine months of its term remaining, you can specify the
amount that you would accept if a customer were to purchase that Reserved Instance with nine months
remaining. You could set another price with five months remaining, and yet another price with one
month remaining.

List your Reserved Instances

As a registered seller, you can choose to sell one or more of your Reserved Instances. You can choose
to sell all of them in one listing or in portions. In addition, you can list Reserved Instances with any
configuration of instance type, platform, and scope.

The console determines a suggested price. It checks for offerings that match your Reserved Instance and
matches the one with the lowest price. Otherwise, it calculates a suggested price based on the cost of

349
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

the Reserved Instance for its remaining time. If the calculated value is less than $1.01, the suggested
price is $1.01.

If you cancel your listing and a portion of that listing has already been sold, the cancellation is not
effective on the portion that has been sold. Only the unsold portion of the listing is no longer available
in the Reserved Instance Marketplace.

To list a Reserved Instance in the Reserved Instance Marketplace using the AWS Management
Console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Reserved Instances.
3. Select the Reserved Instances to list, and choose Actions, Sell Reserved Instances.
4. On the Configure Your Reserved Instance Listing page, set the number of instances to sell and the
upfront price for the remaining term in the relevant columns. See how the value of your reservation
changes over the remainder of the term by selecting the arrow next to the Months Remaining
column.
5. If you are an advanced user and you want to customize the pricing, you can enter different values for
the subsequent months. To return to the default linear price drop, choose Reset.
6. Choose Continue when you are finished configuring your listing.
7. Confirm the details of your listing, on the Confirm Your Reserved Instance Listing page and if
you're satisfied, choose List Reserved Instance.

To view your listings in the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Reserved Instances.
3. Select the Reserved Instance that you've listed and choose the My Listings tab near the bottom of
the page.

To manage Reserved Instances in the Reserved Instance Marketplace using the AWS CLI

1. Get a list of your Reserved Instances by using the describe-reserved-instances command.

2. Note the ID of the Reserved Instance you want to list and call create-reserved-instances-listing. You
must specify the ID of the Reserved Instance, the number of instances, and the pricing schedule.
3. To view your listing, use the describe-reserved-instances-listings command.
4. To cancel your listing, use the cancel-reserved-instances-listings command.

Reserved Instance listing states

Listing State on the My Listings tab of the Reserved Instances page displays the current status of your
listings:

The information displayed by Listing State is about the status of your listing in the Reserved Instance
Marketplace. It is different from the status information that is displayed by the State column in the
Reserved Instances page. This State information is about your reservation.

• active—The listing is available for purchase.

• canceled—The listing is canceled and isn't available for purchase in the Reserved Instance
Marketplace.
• closed—The Reserved Instance is not listed. A Reserved Instance might be closed because the sale of
the listing was completed.

350
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

Lifecycle of a listing
When all the instances in your listing are matched and sold, the My Listings tab shows that the Total
instance count matches the count listed under Sold. Also, there are no Available instances left for your
listing, and its Status is closed.

When only a portion of your listing is sold, AWS retires the Reserved Instances in the listing and creates
the number of Reserved Instances equal to the Reserved Instances remaining in the count. So, the listing
ID and the listing that it represents, which now has fewer reservations for sale, is still active.

Any future sales of Reserved Instances in this listing are processed this way. When all the Reserved
Instances in the listing are sold, AWS marks the listing as closed.

For example, you create a listing Reserved Instances listing ID 5ec28771-05ff-4b9b-aa31-9e57dexample

with a listing count of 5.

The My Listings tab in the Reserved Instance console page displays the listing this way:

Reserved Instance listing ID 5ec28771-05ff-4b9b-aa31-9e57dexample

• Total reservation count = 5

• Sold = 0
• Available = 5
• Status = active

A buyer purchases two of the reservations, which leaves a count of three reservations still available for
sale. Because of this partial sale, AWS creates a new reservation with a count of three to represent the
remaining reservations that are still for sale.

This is how your listing looks in the My Listings tab:

Reserved Instance listing ID 5ec28771-05ff-4b9b-aa31-9e57dexample

• Total reservation count = 5

• Sold = 2
• Available = 3
• Status = active

If you cancel your listing and a portion of that listing has already sold, the cancelation is not effective
on the portion that has been sold. Only the unsold portion of the listing is no longer available in the
Reserved Instance Marketplace.

After your Reserved Instance is sold

When your Reserved Instance is sold, AWS sends you an email notification. Each day that there is any
kind of activity, you receive one email notification capturing all the activities of the day. Activities can
include when you create or sell a listing, or when AWS sends funds to your account.

To track the status of a Reserved Instance listing in the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation page, choose Reserved Instances.
3. Choose the My Listings tab.

The My Listings tab contains the Listing State value. It also contains information about the term,
listing price, and a breakdown of how many instances in the listing are available, pending, sold, and
canceled.

351
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

You can also use the describe-reserved-instances-listings command with the appropriate filter to obtain
information about your listings.

Getting paid
As soon as AWS receives funds from the buyer, a message is sent to the registered owner account email
for the sold Reserved Instance.

AWS sends an Automated Clearing House (ACH) wire transfer to your specified bank account.
Typically, this transfer occurs between one to three days after your Reserved Instance has been sold.
Disbursements take place once a day. You will receive an email with a disbursement report after the
funds are released. Keep in mind that you can't receive disbursements until AWS receives verification
from your bank. This can take up to two weeks.

The Reserved Instance that you sold continues to appear when you describe your Reserved Instances.

You receive a cash disbursement for your Reserved Instances through a wire transfer directly into your
bank account. AWS charges a service fee of 12 percent of the total upfront price of each Reserved
Instance you sell in the Reserved Instance Marketplace.

Information shared with the buyer

When you sell in the Reserved Instance Marketplace, AWS shares your company’s legal name on the
buyer’s statement in accordance with US regulations. In addition, if the buyer calls AWS Support because
the buyer needs to contact you for an invoice or for some other tax-related reason, AWS might need to
provide the buyer with your email address so that the buyer can contact you directly.

For similar reasons, the buyer's ZIP code and country information are provided to the seller in the
disbursement report. As a seller, you might need this information to accompany any necessary
transaction taxes that you remit to the government (such as sales tax and value-added tax).

AWS cannot offer tax advice, but if your tax specialist determines that you need specific additional
information, contact AWS Support.

Modify Reserved Instances

When your needs change, you can modify your Standard or Convertible Reserved Instances and continue
to benefit from the billing benefit. You can modify attributes such as the Availability Zone, instance size
(within the same instance family), and scope of your Reserved Instance.
Note
You can also exchange a Convertible Reserved Instance for another Convertible Reserved
Instance with a different configuration. For more information, see Exchange Convertible
Reserved Instances (p. 359).

You can modify all or a subset of your Reserved Instances. You can separate your original Reserved
Instances into two or more new Reserved Instances. For example, if you have a reservation for 10
instances in us-east-1a and decide to move 5 instances to us-east-1b, the modification request
results in two new reservations: one for 5 instances in us-east-1a and the other for 5 instances in us-
east-1b.

You can also merge two or more Reserved Instances into a single Reserved Instance. For example, if
you have four t2.small Reserved Instances of one instance each, you can merge them to create one
t2.large Reserved Instance. For more information, see Support for modifying instance sizes (p. 354).

After modification, the benefit of the Reserved Instances is applied only to instances that match the new
parameters. For example, if you change the Availability Zone of a reservation, the capacity reservation
and pricing benefits are automatically applied to instance usage in the new Availability Zone. Instances

352
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

that no longer match the new parameters are charged at the On-Demand rate, unless your account has
other applicable reservations.

If your modification request succeeds:

• The modified reservation becomes effective immediately and the pricing benefit is applied to the new
instances beginning at the hour of the modification request. For example, if you successfully modify
your reservations at 9:15PM, the pricing benefit transfers to your new instance at 9:00PM. You can
get the effective date of the modified Reserved Instances by using the describe-reserved-instances
command.
• The original reservation is retired. Its end date is the start date of the new reservation, and the end
date of the new reservation is the same as the end date of the original Reserved Instance. If you
modify a three-year reservation that had 16 months left in its term, the resulting modified reservation
is a 16-month reservation with the same end date as the original one.
• The modified reservation lists a $0 fixed price and not the fixed price of the original reservation.
• The fixed price of the modified reservation does not affect the discount pricing tier calculations
applied to your account, which are based on the fixed price of the original reservation.

If your modification request fails, your Reserved Instances maintain their original configuration, and are
immediately available for another modification request.

There is no fee for modification, and you do not receive any new bills or invoices.

You can modify your reservations as frequently as you like, but you cannot change or cancel a pending
modification request after you submit it. After the modification has completed successfully, you can
submit another modification request to roll back any changes you made, if needed.

Contents
• Requirements and restrictions for modification (p. 353)
• Support for modifying instance sizes (p. 354)
• Submit modification requests (p. 357)
• Troubleshoot modification requests (p. 358)

Requirements and restrictions for modification

You can modify these attributes as follows.

Modifiable attribute Supported platforms Limitations

Change Availability Zones Linux and Windows -

within the same Region

Change the scope from Linux and Windows If you change the scope from
Availability Zone to Region and Availability Zone to Region, you
vice versa lose the capacity reservation
benefit.

If you change the scope from

Region to Availability Zone, you
lose Availability Zone flexibility
and instance size flexibility
(if applicable). For more
information, see How Reserved
Instances are applied (p. 330).

353
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

Modifiable attribute Supported platforms Limitations

Change the instance size within Linux/UNIX only The reservation must use
the same instance family default tenancy. Some instance
Instance size flexibility is not families are not supported,
available for Reserved Instances because there are no other sizes
on the other platforms, which available. For more information,
include Linux with SQL Server see Support for modifying
Standard, Linux with SQL Server instance sizes (p. 354).
Web, Linux with SQL Server
Enterprise, Red Hat Enterprise
Linux, SUSE Linux, Windows,
Windows with SQL Standard,
Windows with SQL Server
Enterprise, and Windows with
SQL Server Web.

Change the network from EC2- Linux and Windows The network platform must be
Classic to Amazon VPC and vice available in your AWS account.
versa If you created your AWS account
after 2013-12-04, it does not
support EC2-Classic.

Requirements

Amazon EC2 processes your modification request if there is sufficient capacity for your new configuration
(if applicable), and if the following conditions are met:

• The Reserved Instance cannot be modified before or at the same time that you purchase it
• The Reserved Instance must be active
• There cannot be a pending modification request
• The Reserved Instance is not listed in the Reserved Instance Marketplace
• There must be a match between the instance size footprint of the original reservation and the new
configuration. For more information, see Support for modifying instance sizes (p. 354).
• The original Reserved Instances are all Standard Reserved Instances or all Convertible Reserved
Instances, not some of each type
• The original Reserved Instances must expire within the same hour, if they are Standard Reserved
Instances
• The Reserved Instance is not a G4dn instance.

Support for modifying instance sizes

You can modify the instance size of a Reserved Instance if the following requirements are met.

Requirements

• The platform is Linux/UNIX.

• You must select another instance size in the same instance family. For example, you cannot modify an
Reserved Instance from t2 to t3, whether you use the same size or a different size.

You cannot modify the instance size of Reserved Instances for the following instances, because each of
these instance families has only one size:
• cc2.8xlarge
• cr1.8xlarge

354
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

• hs1.8xlarge
• t1.micro
• The original and new Reserved Instance must have the same instance size footprint.

Contents
• Instance size footprint (p. 355)
• Normalization factors for bare metal instances (p. 357)

Instance size footprint

Each Reserved Instance has an instance size footprint, which is determined by the normalization factor
of the instance size and the number of instances in the reservation. When you modify the instance
sizes in an Reserved Instance, the footprint of the new configuration must match that of the original
configuration, otherwise the modification request is not processed.

To calculate the instance size footprint of a Reserved Instance, multiply the number of instances by
the normalization factor. In the Amazon EC2 console, the normalization factor is measured in units.
The following table describes the normalization factor for the instance sizes in an instance family. For
example, t2.medium has a normalization factor of 2, so a reservation for four t2.medium instances has
a footprint of 8 units.

Instance size Normalization factor

nano 0.25

micro 0.5

small 1

medium 2

large 4

xlarge 8

2xlarge 16

4xlarge 32

8xlarge 64

9xlarge 72

10xlarge 80

12xlarge 96

16xlarge 128

18xlarge 144

24xlarge 192

32xlarge 256

You can allocate your reservations into different instance sizes across the same instance family as
long as the instance size footprint of your reservation remains the same. For example, you can divide

355
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

a reservation for one t2.large (1 @ 4 units) instance into four t2.small (4 @ 1 unit) instances.
Similarly, you can combine a reservation for four t2.small instances into one t2.large instance.
However, you cannot change your reservation for two t2.small instances into one t2.large instance
because the footprint of the new reservation (4 units) is larger than the footprint of the original
reservation (2 units).

In the following example, you have a reservation with two t2.micro instances (1 unit) and a reservation
with one t2.small instance (1 unit). If you merge both of these reservations to a single reservation
with one t2.medium instance (2 units), the footprint of the new reservation equals the footprint of the
combined reservations.

You can also modify a reservation to divide it into two or more reservations. In the following example,
you have a reservation with a t2.medium instance (2 units). You can divide the reservation into two
reservations, one with two t2.nano instances (.5 units) and the other with three t2.micro instances
(1.5 units).

356
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

Normalization factors for bare metal instances

You can modify a reservation with metal instances using other sizes within the same instance family.
Similarly, you can modify a reservation with instances other than bare metal instances using the metal
size within the same instance family. Generally, a bare metal instance is the same size as the largest
available instance size within the same instance family. For example, an i3.metal instance is the same
size as an i3.16xlarge instance, so they have the same normalization factor.

The following table describes the normalization factor for the bare metal instance sizes in the instance
families that have bare metal instances. The normalization factor for metal instances depends on the
instance family, unlike the other instance sizes.

Bare metal instance size Normalization factor

c5.metal 192

i3.metal 128

r5.metal 192

r5d.metal 192

z1d.metal 96

m5.metal 192

m5d.metal 192

For example, an i3.metal instance has a normalization factor of 128. If you purchase an i3.metal
default tenancy Amazon Linux/Unix Reserved Instance, you can divide the reservation as follows:

• An i3.16xlarge is the same size as an i3.metal instance, so its normalization factor is 128 (128/1).
The reservation for one i3.metal instance can be modified into one i3.16xlarge instance.
• An i3.8xlarge is half the size of an i3.metal instance, so its normalization factor is 64 (128/2). The
reservation for one i3.metal instance can be divided into two i3.8xlarge instances.
• An i3.4xlarge is a quarter the size of an i3.metal instance, so its normalization factor is 32
(128/4). The reservation for one i3.metal instance can be divided into four i3.4xlarge instances.

Submit modification requests

Before you modify your Reserved Instances, ensure that you have read the applicable
restrictions (p. 353). Before you modify the instance size, calculate the total instance size
footprint (p. 354) of the original reservations that you want to modify and ensure that it matches the
total instance size footprint of your new configurations.

To modify your Reserved Instances using the AWS Management Console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. On the Reserved Instances page, select one or more Reserved Instances to modify, and choose
Actions, Modify Reserved Instances.
Note
If your Reserved Instances are not in the active state or cannot be modified, Modify
Reserved Instances is disabled.
3. The first entry in the modification table displays attributes of selected Reserved Instances, and
at least one target configuration beneath it. The Units column displays the total instance size

357
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

footprint. Choose Add for each new configuration to add. Modify the attributes as needed for each
configuration, and then choose Continue:

• Scope: Choose whether the configuration applies to an Availability Zone or to the whole Region.
• Availability Zone: Choose the required Availability Zone. Not applicable for regional Reserved
Instances.
• Instance Type: Select the required instance type. The combined configurations must equal the
instance size footprint of your original configurations.
• Count: Specify the number of instances. To split the Reserved Instances into multiple
configurations, reduce the count, choose Add, and specify a count for the additional configuration.
For example, if you have a single configuration with a count of 10, you can change its count to 6
and add a configuration with a count of 4. This process retires the original Reserved Instance after
the new Reserved Instances are activated.
4. To confirm your modification choices when you finish specifying your target configurations, choose
Submit Modifications.
5. You can determine the status of your modification request by looking at the State column in the
Reserved Instances screen. The following are the possible states.

• active (pending modification) — Transition state for original Reserved Instances

• retired (pending modification) — Transition state for original Reserved Instances while new
Reserved Instances are being created
• retired — Reserved Instances successfully modified and replaced
• active — One of the following:
• New Reserved Instances created from a successful modification request
• Original Reserved Instances after a failed modification request

To modify your Reserved Instances using the command line

1. To modify your Reserved Instances, you can use one of the following commands:
• modify-reserved-instances (AWS CLI)
• Edit-EC2ReservedInstance (AWS Tools for Windows PowerShell)
2. To get the status of your modification request (processing, fulfilled, or failed), use one of
the following commands:
• describe-reserved-instances-modifications (AWS CLI)
• Get-EC2ReservedInstancesModification (AWS Tools for Windows PowerShell)

Troubleshoot modification requests

If the target configuration settings that you requested were unique, you receive a message that your
request is being processed. At this point, Amazon EC2 has only determined that the parameters of
your modification request are valid. Your modification request can still fail during processing due to
unavailable capacity.

In some situations, you might get a message indicating incomplete or failed modification requests
instead of a confirmation. Use the information in such messages as a starting point for resubmitting
another modification request. Ensure that you have read the applicable restrictions (p. 353) before
submitting the request.

Not all selected Reserved Instances can be processed for modification

Amazon EC2 identifies and lists the Reserved Instances that cannot be modified. If you receive a message
like this, go to the Reserved Instances page in the Amazon EC2 console and check the information for
the Reserved Instances.

358
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

Error in processing your modification request

You submitted one or more Reserved Instances for modification and none of your requests can be
processed. Depending on the number of reservations you are modifying, you can get different versions of
the message.

Amazon EC2 displays the reasons why your request cannot be processed. For example, you might have
specified the same target configuration—a combination of Availability Zone and platform—for one or
more subsets of the Reserved Instances you are modifying. Try submitting the modification requests
again, but ensure that the instance details of the reservations match, and that the target configurations
for all subsets being modified are unique.

Exchange Convertible Reserved Instances

You can exchange one or more Convertible Reserved Instances for another Convertible Reserved Instance
with a different configuration, including instance family, operating system, and tenancy. There are no
limits to how many times you perform an exchange, as long as the new Convertible Reserved Instance is
of an equal or higher value than the original Convertible Reserved Instances that you are exchanging.

When you exchange your Convertible Reserved Instance, the number of instances for your current
reservation is exchanged for a number of instances that cover the equal or higher value of the
configuration of the new Convertible Reserved Instance. Amazon EC2 calculates the number of Reserved
Instances that you can receive as a result of the exchange.

You can't exchange Standard Reserved Instances, but you can modify them. For more information, see
Modify Reserved Instances (p. 352) .

Contents
• Requirements for exchanging Convertible Reserved Instances (p. 359)
• Calculate Convertible Reserved Instances exchanges (p. 360)
• Merge Convertible Reserved Instances (p. 361)
• Exchange a portion of a Convertible Reserved Instance (p. 361)
• Submit exchange requests (p. 362)

Requirements for exchanging Convertible Reserved Instances

If the following conditions are met, Amazon EC2 processes your exchange request. Your Convertible
Reserved Instance must be:

• Active
• Not pending a previous exchange request

The following rules apply:

• Convertible Reserved Instances can only be exchanged for other Convertible Reserved Instances
currently offered by AWS.
• Convertible Reserved Instances are associated with a specific Region, which is fixed for the duration
of the reservation's term. You cannot exchange a Convertible Reserved Instance for a Convertible
Reserved Instance in a different Region.
• You can exchange one or more Convertible Reserved Instances at a time for one Convertible Reserved
Instance only.
• To exchange a portion of a Convertible Reserved Instance, you can modify it into two or
more reservations, and then exchange one or more of the reservations for a new Convertible
Reserved Instance. For more information, see Exchange a portion of a Convertible Reserved

359
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

Instance (p. 361). For more information about modifying your Reserved Instances, see Modify
Reserved Instances (p. 352).
• All Upfront Convertible Reserved Instances can be exchanged for Partial Upfront Convertible Reserved
Instances, and vice versa.
Note
If the total upfront payment required for the exchange (true-up cost) is less than $0.00, AWS
automatically gives you a quantity of instances in the Convertible Reserved Instance that
ensures that true-up cost is $0.00 or more.
Note
If the total value (upfront price + hourly price * number of remaining hours) of the new
Convertible Reserved Instance is less than the total value of the exchanged Convertible
Reserved Instance, AWS automatically gives you a quantity of instances in the Convertible
Reserved Instance that ensures that the total value is the same or higher than that of the
exchanged Convertible Reserved Instance.
• To benefit from better pricing, you can exchange a No Upfront Convertible Reserved Instance for an All
Upfront or Partial Upfront Convertible Reserved Instance.
• You cannot exchange All Upfront and Partial Upfront Convertible Reserved Instances for No Upfront
Convertible Reserved Instances.
• You can exchange a No Upfront Convertible Reserved Instance for another No Upfront Convertible
Reserved Instance only if the new Convertible Reserved Instance's hourly price is the same or higher
than the exchanged Convertible Reserved Instance's hourly price.
Note
If the total value (hourly price * number of remaining hours) of the new Convertible Reserved
Instance is less than the total value of the exchanged Convertible Reserved Instance, AWS
automatically gives you a quantity of instances in the Convertible Reserved Instance that
ensures that the total value is the same or higher than that of the exchanged Convertible
Reserved Instance.
• If you exchange multiple Convertible Reserved Instances that have different expiration dates, the
expiration date for the new Convertible Reserved Instance is the date that's furthest in the future.
• If you exchange a single Convertible Reserved Instance, it must have the same term (1-year or 3-
years) as the new Convertible Reserved Instance. If you merge multiple Convertible Reserved Instances
with different term lengths, the new Convertible Reserved Instance has a 3-year term. For more
information, see Merge Convertible Reserved Instances (p. 361).
• After you exchange a Convertible Reserved Instance, the original reservation is retired. Its end date is
the start date of the new reservation, and the end date of the new reservation is the same as the end
date of the original Convertible Reserved Instance. For example, if you modify a three-year reservation
that had 16 months left in its term, the resulting modified reservation is a 16-month reservation with
the same end date as the original one.

Calculate Convertible Reserved Instances exchanges

Exchanging Convertible Reserved Instances is free. However, you might be required to pay a true-up cost,
which is a prorated upfront cost of the difference between the original Convertible Reserved Instances
that you had and the new Convertible Reserved Instances that you receive from the exchange.

Each Convertible Reserved Instance has a list value. This list value is compared to the list value of the
Convertible Reserved Instances that you want in order to determine how many instance reservations you
can receive from the exchange.

For example: You have 1 x $35-list value Convertible Reserved Instance that you want to exchange for a
new instance type with a list value of $10.

$35/$10 = 3.5

360
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

You can exchange your Convertible Reserved Instance for three $10 Convertible Reserved Instances.
It's not possible to purchase half reservations; therefore you must purchase an additional Convertible
Reserved Instance to cover the remainder:

3.5 = 3 whole Convertible Reserved Instances + 1 additional Convertible Reserved Instance

The fourth Convertible Reserved Instance has the same end date as the other three. If you are
exchanging Partial or All Upfront Convertible Reserved Instances, you pay the true-up cost for the fourth
reservation. If the remaining upfront cost of your Convertible Reserved Instances is $500, and the new
reservation would normally cost $600 on a prorated basis, you are charged $100.

$600 prorated upfront cost of new reservations - $500 remaining upfront cost of original
reservations = $100 difference

Merge Convertible Reserved Instances

If you merge two or more Convertible Reserved Instances, the term of the new Convertible Reserved
Instance must be the same as the original Convertible Reserved Instances, or the highest of the original
Convertible Reserved Instances. The expiration date for the new Convertible Reserved Instance is the
expiration date that's furthest in the future.

For example, you have the following Convertible Reserved Instances in your account:

Reserved Instance ID Term Expiration date

aaaa1111 1-year 2018-12-31

bbbb2222 1-year 2018-07-31

cccc3333 3-year 2018-06-30

dddd4444 3-year 2019-12-31

• You can merge aaaa1111 and bbbb2222 and exchange them for a 1-year Convertible Reserved
Instance. You cannot exchange them for a 3-year Convertible Reserved Instance. The expiration date of
the new Convertible Reserved Instance is 2018-12-31.
• You can merge bbbb2222 and cccc3333 and exchange them for a 3-year Convertible Reserved
Instance. You cannot exchange them for a 1-year Convertible Reserved Instance. The expiration date of
the new Convertible Reserved Instance is 2018-07-31.
• You can merge cccc3333 and dddd4444 and exchange them for a 3-year Convertible Reserved
Instance. You cannot exchange them for a 1-year Convertible Reserved Instance. The expiration date of
the new Convertible Reserved Instance is 2019-12-31.

Exchange a portion of a Convertible Reserved Instance

You can use the modification process to split your Convertible Reserved Instance into smaller
reservations, and then exchange one or more of the new reservations for a new Convertible Reserved
Instance. The following examples demonstrate how you can do this.

Example Example: Convertible Reserved Instance with multiple instances

In this example, you have a t2.micro Convertible Reserved Instance with four instances in the
reservation. To exchange two t2.micro instances for an m4.xlarge instance:

1. Modify the t2.micro Convertible Reserved Instance by splitting it into two t2.micro Convertible
Reserved Instances with two instances each.

361
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reserved Instances

2. Exchange one of the new t2.micro Convertible Reserved Instances for an m4.xlarge Convertible
Reserved Instance.

Example Example: Convertible Reserved Instance with a single instance

In this example, you have a t2.large Convertible Reserved Instance. To change it to a smaller
t2.medium instance and a m3.medium instance:

1. Modify the t2.large Convertible Reserved Instance by splitting it into two t2.medium Convertible
Reserved Instances. A single t2.large instance has the same instance size footprint as two
t2.medium instances.
2. Exchange one of the new t2.medium Convertible Reserved Instances for an m3.medium Convertible
Reserved Instance.

For more information, see Support for modifying instance sizes (p. 354) and Submit exchange
requests (p. 362).

Submit exchange requests

You can exchange your Convertible Reserved Instances using the Amazon EC2 console or a command line
tool.

Exchange a Convertible Reserved Instance using the console

You can search for Convertible Reserved Instances offerings and select your new configuration from the
choices provided.

To exchange Convertible Reserved Instances using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

362
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled Instances

2. Choose Reserved Instances, select the Convertible Reserved Instances to exchange, and choose
Actions, Exchange Reserved Instance.
3. Select the attributes of the desired configuration using the drop-down menus, and choose Find
Offering.
4. Select a new Convertible Reserved Instance The Instance Count column displays the number
of Reserved Instances that you receive for the exchange. When you have selected a Convertible
Reserved Instance that meets your needs, choose Exchange.

The Reserved Instances that were exchanged are retired, and the new Reserved Instances are displayed in
the Amazon EC2 console. This process can take a few minutes to propagate.

Exchange a Convertible Reserved Instance using the command line interface

To exchange a Convertible Reserved Instance, first find a new Convertible Reserved Instance that meets
your needs:

• describe-reserved-instances-offerings (AWS CLI)

• Get-EC2ReservedInstancesOffering (Tools for Windows PowerShell)

Get a quote for the exchange, which includes the number of Reserved Instances you get from the
exchange, and the true-up cost for the exchange:

• get-reserved-instances-exchange-quote (AWS CLI)

• GetEC2-ReservedInstancesExchangeQuote (Tools for Windows PowerShell)

Finally, perform the exchange:

• accept-reserved-instances-exchange-quote (AWS CLI)

• Confirm-EC2ReservedInstancesExchangeQuote (Tools for Windows PowerShell)

Scheduled Reserved Instances

Important
We do not have any capacity for purchasing Scheduled Reserved Instances or any plans to make
it available in the future. To reserve capacity, use On-Demand Capacity Reservations (p. 501)
instead. For discounted rates, use Savings Plans.

363
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Spot Instances
A Spot Instance is an unused EC2 instance that is available for less than the On-Demand price. Because
Spot Instances enable you to request unused EC2 instances at steep discounts, you can lower your
Amazon EC2 costs significantly. The hourly price for a Spot Instance is called a Spot price. The Spot price
of each instance type in each Availability Zone is set by Amazon EC2, and is adjusted gradually based on
the long-term supply of and demand for Spot Instances. Your Spot Instance runs whenever capacity is
available and the maximum price per hour for your request exceeds the Spot price.

Spot Instances are a cost-effective choice if you can be flexible about when your applications run and if
your applications can be interrupted. For example, Spot Instances are well-suited for data analysis, batch
jobs, background processing, and optional tasks. For more information, see Amazon EC2 Spot Instances.

Topics

• Concepts (p. 364)

• How to get started (p. 365)
• Related services (p. 366)
• Pricing and savings (p. 367)

Concepts
Before you get started with Spot Instances, you should be familiar with the following concepts:

• Spot Instance pool – A set of unused EC2 instances with the same instance type (for example,
m5.large), operating system, Availability Zone, and network platform.
• Spot price – The current price of a Spot Instance per hour.
• Spot Instance request – Requests a Spot Instance. The request provides the maximum price per hour
that you are willing to pay for a Spot Instance. If you don't specify a maximum price, the default
maximum price is the On-Demand price. When the maximum price per hour for your request exceeds
the Spot price, Amazon EC2 fulfills your request if capacity is available. A Spot Instance request is
either one-time or persistent. Amazon EC2 automatically resubmits a persistent Spot Instance request
after the Spot Instance associated with the request is terminated. Your Spot Instance request can
optionally specify a duration for the Spot Instances.
• Spot Fleet – A set of Spot Instances that is launched based on criteria that you specify. The Spot
Fleet selects the Spot Instance pools that meet your needs and launches Spot Instances to meet the
target capacity for the fleet. By default, Spot Fleets are set to maintain target capacity by launching
replacement instances after Spot Instances in the fleet are terminated. You can submit a Spot Fleet as
a one-time request, which does not persist after the instances have been terminated. You can include
On-Demand Instance requests in a Spot Fleet request.
• EC2 instance rebalance recommendation - Amazon EC2 emits an instance rebalance recommendation
signal to notify you that a Spot Instance is at an elevated risk of interruption. This signal gives you the
opportunity to proactively rebalance your workloads across existing or new Spot Instances without
having to wait for the two-minute Spot Instance interruption notice.
• Spot Instance interruption – Amazon EC2 terminates, stops, or hibernates your Spot Instance when the
Spot price exceeds the maximum price for your request or capacity is no longer available. Amazon EC2
provides a Spot Instance interruption notice, which gives the instance a two-minute warning before it
is interrupted.

Key differences between Spot Instances and On-Demand Instances

The following table lists the key differences between Spot Instances and On-Demand Instances.

364
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

  Spot Instances On-Demand Instances

Launch time Can only be launched immediately if the Can only be launched immediately if
Spot Request is active and capacity is you make a manual launch request and
available. capacity is available.

Available If capacity is not available, the Spot If capacity is not available when you
capacity Request continues to automatically make a launch request, you get an
make the launch request until capacity insufficient capacity error (ICE).
becomes available.

Hourly price The hourly price for Spot Instances varies The hourly price for On-Demand
based on demand. Instances is static.

Rebalance The signal that Amazon EC2 emits for a You determine when an On-Demand
recommendation running Spot Instance when the instance Instance is interrupted (stopped,
is at an elevated risk of interruption. hibernated, or terminated).

Instance You can stop and start an Amazon You determine when an On-Demand
interruption EBS-backed Spot Instance. In addition, Instance is interrupted (stopped,
the Amazon EC2 Spot service can hibernated, or terminated).
interrupt (p. 452) an individual Spot
Instance if capacity is no longer available,
the Spot price exceeds your maximum
price, or demand for Spot Instances
increases.

Strategies for using Spot Instances

One strategy is to maintain a minimum level of guaranteed compute resources for your applications by
launching a core group of On-Demand Instances, and supplementing them with Spot Instances when the
opportunity arises.

Another strategy is to launch Spot Instances with a specified duration (also known as Spot blocks),
which are designed not to be interrupted and will run continuously for the duration you select. In
rare situations, Spot blocks may be interrupted due to Amazon EC2 capacity needs. In these cases,
we provide a two-minute warning before we terminate an instance, and you are not charged for the
terminated instances even if you used them. For more information, see Define a duration for your Spot
Instances (p. 384).

How to get started

The first thing you need to do is get set up to use Amazon EC2. It can also be helpful to have experience
launching On-Demand Instances before launching Spot Instances.

365
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Get up and running

• Set up to use Amazon EC2 (p. 28)

• Tutorial: Get started with Amazon EC2 Linux instances (p. 32)

Spot basics

• How Spot Instances work (p. 369)

• How Spot Fleet works (p. 371)

Working with Spot Instances

• Prepare for interruptions (p. 456)

• Create a Spot Instance request (p. 387)
• Get request status information (p. 447)

Working with Spot Fleets

• Spot Fleet permissions (p. 403)

• Create a Spot Fleet request (p. 407)

Related services
You can provision Spot Instances directly using Amazon EC2. You can also provision Spot Instances using
other services in AWS. For more information, see the following documentation.

Amazon EC2 Auto Scaling and Spot Instances

You can create launch templates or configurations with the maximum price that you are willing
to pay, so that Amazon EC2 Auto Scaling can launch Spot Instances. For more information, see
Launching Spot Instances in Your Auto Scaling Group and Using Multiple Instance Types and
Purchase Options in the Amazon EC2 Auto Scaling User Guide.
Amazon EMR and Spot Instances

There are scenarios where it can be useful to run Spot Instances in an Amazon EMR cluster. For
more information, see Spot Instances and When Should You Use Spot Instances in the Amazon EMR
Management Guide.
AWS CloudFormation templates

AWS CloudFormation enables you to create and manage a collection of AWS resources using
a template in JSON format. AWS CloudFormation templates can include the maximum price
you are willing to pay. For more information, see EC2 Spot Instance Updates - Auto Scaling and
CloudFormation Integration.
AWS SDK for Java

You can use the Java programming language to manage your Spot Instances. For more information,
see Tutorial: Amazon EC2 Spot Instances and Tutorial: Advanced Amazon EC2 Spot Request
Management.
AWS SDK for .NET

You can use the .NET programming environment to manage your Spot Instances. For more
information, see Tutorial: Amazon EC2 Spot Instances.

366
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Pricing and savings

You pay the Spot price for Spot Instances, which is set by Amazon EC2 and adjusted gradually based on
the long-term supply of and demand for Spot Instances. If the maximum price for your request exceeds
the current Spot price, Amazon EC2 fulfills your request if capacity is available. Your Spot Instances run
until you terminate them, capacity is no longer available, the Spot price exceeds your maximum price, or
your Amazon EC2 Auto Scaling group terminates them during scale in.

Spot Instances with a predefined duration use a fixed hourly price that remains in effect for the Spot
Instance while it runs.

If you or Amazon EC2 interrupts a running Spot Instance, you are charged for the seconds used or the full
hour, or you receive no charge, depending on the operating system used and who interrupted the Spot
Instance. For more information, see Billing for interrupted Spot Instances (p. 459).

View prices
To view the current (updated every five minutes) lowest Spot price per AWS Region and instance type,
see the Spot Instances Pricing page.

To view the Spot price history for the past three months, use the Amazon EC2 console or the
describe-spot-price-history command (AWS CLI). For more information, see Spot Instance pricing
history (p. 380).

We independently map Availability Zones to codes for each AWS account. Therefore, you can get
different results for the same Availability Zone code (for example, us-west-2a) between different
accounts.

View savings
You can view the savings made from using Spot Instances for a single Spot Fleet or for all Spot Instances.
You can view the savings made in the last hour or the last three days, and you can view the average cost
per vCPU hour and per memory (GiB) hour. Savings are estimated and may differ from actual savings
because they do not include the billing adjustments for your usage. For more information about viewing
savings information, see Savings from purchasing Spot Instances (p. 381).

View billing
Your bill provides details about your service usage. For more information, see Viewing your bill in the
AWS Billing and Cost Management User Guide.

Best practices for EC2 Spot

Amazon EC2 Spot Instances are spare EC2 compute capacity in the AWS Cloud that are available to you
at savings of up to 90% off compared to On-Demand prices. The only difference between On-Demand
Instances and Spot Instances is that Spot Instances can be interrupted by Amazon EC2, with two minutes
of notification, when Amazon EC2 needs the capacity back.

Spot Instances are recommended for stateless, fault-tolerant, flexible applications. For example,
Spot Instances work well for big data, containerized workloads, CI/CD, stateless web servers, high
performance computing (HPC), and rendering workloads.

While running, Spot Instances are exactly the same as On-Demand Instances. However, Spot does
not guarantee that you can keep your running instances long enough to finish your workloads. Spot
also does not guarantee that you can get immediate availability of the instances that you are looking
for, or that you can always get the aggregate capacity that you requested. Moreover, Spot Instance

367
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

interruptions and capacity can change over time because Spot Instance availability varies based on
supply and demand, and past performance isn’t a guarantee of future results.

Spot Instances are not suitable for workloads that are inflexible, stateful, fault-intolerant, or tightly
coupled between instance nodes. It's also not recommended for workloads that are intolerant of
occasional periods when the target capacity is not completely available. We strongly warn against
using Spot Instances for these workloads or attempting to fail-over to On-Demand Instances to handle
interruptions.

Regardless of whether you're an experienced Spot user or new to Spot Instances, if you are currently
experiencing issues with Spot Instance interruptions or availability, we recommend that you follow these
best practices to have the best experience using the Spot service.

Spot best practices

• Prepare individual instances for interruptions (p. 368)
• Be flexible about instance types and Availability Zones (p. 368)
• Use EC2 Auto Scaling groups or Spot Fleet to manage your aggregate capacity (p. 369)
• Use the capacity optimized allocation strategy (p. 369)
• Use proactive capacity rebalancing (p. 369)
• Use integrated AWS services to manage your Spot Instances (p. 369)

Prepare individual instances for interruptions

The best way for you to gracefully handle Spot Instance interruptions is to architect your application
to be fault-tolerant. To accomplish this, you can take advantage of EC2 instance rebalance
recommendations and Spot Instance interruption notices.

An EC2 Instance rebalance recommendation is a new signal that notifies you when a Spot Instance is
at elevated risk of interruption. The signal gives you the opportunity to proactively manage the Spot
Instance in advance of the two-minute Spot Instance interruption notice. You can decide to rebalance
your workload to new or existing Spot Instances that are not at an elevated risk of interruption. We've
made it easy for you to use this new signal by using the Capacity Rebalancing feature in Auto Scaling
groups and Spot Fleet. For more information, see Use proactive capacity rebalancing (p. 369).

A Spot Instance interruption notice is a warning that is issued two minutes before Amazon EC2 interrupts
a Spot Instance. If your workload is "time-flexible," you can configure your Spot Instances to be stopped
or hibernated, instead of being terminated, when they are interrupted. Amazon EC2 automatically stops
or hibernates your Spot Instances on interruption, and automatically resumes the instances when we
have available capacity.

We recommend that you create a rule in Amazon EventBridge that captures the rebalance
recommendations and interruption notifications, and then triggers a checkpoint for the progress of
your workload or gracefully handles the interruption. For more information, see Monitor rebalance
recommendation signals (p. 450). For a detailed example that walks you through how to create and use
event rules, see Taking Advantage of Amazon EC2 Spot Instance Interruption Notices.

For more information, see EC2 instance rebalance recommendations (p. 449) and Spot Instance
interruptions (p. 452).

Be flexible about instance types and Availability Zones

A Spot Instance pool is a set of unused EC2 instances with the same instance type (for example,
m5.large) and Availability Zone (for example, us-east-1a). You should be flexible about which instance
types you request and in which Availability Zones you can deploy your workload. This gives Spot a better

368
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

chance to find and allocate your required amount of compute capacity. For example, don't just ask for
c5.large if you'd be willing to use larges from the c4, m5, and m4 families.

Depending on your specific needs, you can evaluate which instance types you can be flexible across
to fulfill your compute requirements. If a workload can be vertically scaled, you should include larger
instance types (more vCPUs and memory) in your requests. If you can only scale horizontally, you should
include older generation instance types because they are less in demand from On-Demand customers.

A good rule of thumb is to be flexible across at least 10 instance types for each workload. In addition,
make sure that all Availability Zones are configured for use in your VPC and selected for your workload.

Use EC2 Auto Scaling groups or Spot Fleet to manage your aggregate capacity
Spot enables you to think in terms of aggregate capacity—in units that include vCPUs, memory, storage,
or network throughput—rather than thinking in terms of individual instances. Auto Scaling groups and
Spot Fleet enable you to launch and maintain a target capacity, and to automatically request resources
to replace any that are disrupted or manually terminated. When you configure an Auto Scaling group
or a Spot Fleet, you need only specify the instance types and target capacity based on your application
needs. For more information, see Auto Scaling Groups in the Amazon EC2 Auto Scaling User Guide and
Create a Spot Fleet request (p. 407) in this user guide.

Use the capacity optimized allocation strategy

Allocation strategies in Auto Scaling groups help you to provision your target capacity without the need
to manually look for the Spot Instance pools with spare capacity. We recommend using the capacity
optimized strategy because this strategy automatically provisions instances from the most-available
Spot Instance pools. You can also take advantage of the capacity optimized allocation strategy
in Spot Fleet. Because your Spot Instance capacity is sourced from pools with optimal capacity, this
decreases the possibility that your Spot Instances are reclaimed. For more information about allocation
strategies, see Spot Instances in the Amazon EC2 Auto Scaling User Guide and Configure Spot Fleet for
capacity optimization (p. 373) in this user guide.

Use proactive capacity rebalancing

Capacity Rebalancing helps you maintain workload availability by proactively augmenting your fleet with
a new Spot Instance before a running Spot Instance receives the two-minute Spot Instance interruption
notice. When Capacity Rebalancing is enabled, Auto Scaling or Spot Fleet attempts to proactively replace
Spot Instances that have received a rebalance recommendation, providing the opportunity to rebalance
your workload to new Spot Instances that are not at elevated risk of interruption.

Capacity Rebalancing complements the capacity optimized allocation strategy (which is designed to
help find the most optimal spare capacity) and the mixed instances policy (which is designed to enhance
availability by deploying instances across multiple instance types running in multiple Availability Zones).

For more information, see Capacity Rebalancing (p. 374).

Use integrated AWS services to manage your Spot Instances

Other AWS services integrate with Spot to reduce overall compute costs without the need to manage
the individual instances or fleets. We recommend that you consider the following solutions for your
applicable workloads: Amazon EMR, Amazon ECS, AWS Batch, Amazon EKS, SageMaker, AWS Elastic
Beanstalk, and Amazon GameLift. To learn more about Spot best practices with these services, see the
Amazon EC2 Spot Instances Workshops Website.

How Spot Instances work

To launch a Spot Instance, either you create a Spot Instance request, or Amazon EC2 creates a Spot
Instance request on your behalf. The Spot Instance launches when the Spot Instance request is fulfilled.

369
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

You can launch a Spot Instance using several different services. For more information, see Getting
Started with Amazon EC2 Spot Instances. In this user guide, we describe the following ways to launch a
Spot Instance using EC2:

• You can create a Spot Instance request. For more information, see Create a Spot Instance
request (p. 387).
• You can create a Spot Fleet request, in which you specify the desired number of Spot Instances.
Amazon EC2 creates a Spot Instance request on your behalf for every Spot Instance that is specified in
the Spot Fleet request. For more information, see Create a Spot Fleet request (p. 407).
• You can create an EC2 Fleet, in which you specify the desired number of Spot Instances. Amazon EC2
creates a Spot Instance request on your behalf for every Spot Instance that is specified in the EC2
Fleet. For more information, see Create an EC2 Fleet (p. 570).

The Spot Instance request must include the maximum price that you're willing to pay per hour per
instance. If you don't specify a price, the price defaults to the On-Demand price. The request can include
other constraints such as the instance type and Availability Zone.

Your Spot Instance launches if the maximum price that you're willing to pay exceeds the Spot price, and
if there is available capacity. If the maximum price you're willing to pay is lower than the Spot price, then
your instance does not launch. However, because Amazon EC2 gradually adjusts the Spot price based on
the long-term supply of and demand for Spot Instances, the maximum price you're willing to pay might
eventually exceed the Spot price, in which case your instance will launch.

Your Spot Instance runs until you stop or terminate it, or until Amazon EC2 interrupts it (known as a Spot
Instance interruption).

When you use Spot Instances, you must be prepared for interruptions. Amazon EC2 can interrupt
your Spot Instance when the Spot price exceeds your maximum price, when the demand for Spot
Instances rises, or when the supply of Spot Instances decreases. When Amazon EC2 interrupts a Spot
Instance, it provides a Spot Instance interruption notice, which gives the instance a two-minute warning
before Amazon EC2 interrupts it. You can't enable termination protection for Spot Instances. For more
information, see Spot Instance interruptions (p. 452).

You can stop, start, reboot, or terminate an Amazon EBS-backed Spot Instance. The Spot service can
stop, terminate, or hibernate a Spot Instance when it interrupts it.

Contents
• Launch Spot Instances in a launch group (p. 370)
• Launch Spot Instances in an Availability Zone group (p. 371)
• Launch Spot Instances in a VPC (p. 371)

Launch Spot Instances in a launch group

Specify a launch group in your Spot Instance request to tell Amazon EC2 to launch a set of Spot
Instances only if it can launch them all. In addition, if the Spot service must terminate one of the
instances in a launch group (for example, if the Spot price exceeds your maximum price), it must
terminate them all. However, if you terminate one or more of the instances in a launch group, Amazon
EC2 does not terminate the remaining instances in the launch group.

Although this option can be useful, adding this constraint can decrease the chances that your Spot
Instance request is fulfilled and increase the chances that your Spot Instances are terminated. For
example, your launch group includes instances in multiple Availability Zones. If capacity in one of these
Availability Zones decreases and is no longer available, then Amazon EC2 terminates all instances for the
launch group.

If you create another successful Spot Instance request that specifies the same (existing) launch group
as an earlier successful request, then the new instances are added to the launch group. Subsequently, if

370
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

an instance in this launch group is terminated, all instances in the launch group are terminated, which
includes instances launched by the first and second requests.

Launch Spot Instances in an Availability Zone group

Specify an Availability Zone group in your Spot Instance request to tell the Spot service to launch a
set of Spot Instances in the same Availability Zone. Amazon EC2 need not interrupt all instances in
an Availability Zone group at the same time. If Amazon EC2 must interrupt one of the instances in an
Availability Zone group, the others remain running.

Although this option can be useful, adding this constraint can lower the chances that your Spot Instance
request is fulfilled.

If you specify an Availability Zone group but don't specify an Availability Zone in the Spot Instance
request, the result depends on the network you specified.

Default VPC

Amazon EC2 uses the Availability Zone for the specified subnet. If you don't specify a subnet, it selects
an Availability Zone and its default subnet, but not necessarily the lowest-priced zone. If you deleted the
default subnet for an Availability Zone, then you must specify a different subnet.

Nondefault VPC

Amazon EC2 uses the Availability Zone for the specified subnet.

Launch Spot Instances in a VPC

You specify a subnet for your Spot Instances the same way that you specify a subnet for your On-
Demand Instances.

• You should use the default maximum price (the On-Demand price), or base your maximum price on the
Spot price history of Spot Instances in a VPC.
• [Default VPC] If you want your Spot Instance launched in a specific low-priced Availability Zone, you
must specify the corresponding subnet in your Spot Instance request. If you do not specify a subnet,
Amazon EC2 selects one for you, and the Availability Zone for this subnet might not have the lowest
Spot price.
• [Nondefault VPC] You must specify the subnet for your Spot Instance.

How Spot Fleet works

A Spot Fleet is a collection, or fleet, of Spot Instances, and optionally On-Demand Instances.

The Spot Fleet attempts to launch the number of Spot Instances and On-Demand Instances to meet the
target capacity that you specified in the Spot Fleet request. The request for Spot Instances is fulfilled
if there is available capacity and the maximum price you specified in the request exceeds the current
Spot price. The Spot Fleet also attempts to maintain its target capacity fleet if your Spot Instances are
interrupted.

You can also set a maximum amount per hour that you’re willing to pay for your fleet, and Spot Fleet
launches instances until it reaches the maximum amount. When the maximum amount you're willing to
pay is reached, the fleet stops launching instances even if it hasn’t met the target capacity.

A Spot Instance pool is a set of unused EC2 instances with the same instance type (for example,
m5.large), operating system, Availability Zone, and network platform. When you make a Spot Fleet
request, you can include multiple launch specifications, that vary by instance type, AMI, Availability Zone,
or subnet. The Spot Fleet selects the Spot Instance pools that are used to fulfill the request, based on

371
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

the launch specifications included in your Spot Fleet request, and the configuration of the Spot Fleet
request. The Spot Instances come from the selected pools.

Contents
• On-Demand in Spot Fleet (p. 372)
• Allocation strategy for Spot Instances (p. 372)
• Capacity Rebalancing (p. 374)
• Spot price overrides (p. 375)
• Control spending (p. 376)
• Spot Fleet instance weighting (p. 376)
• Walkthrough: Using Spot Fleet with instance weighting (p. 378)

On-Demand in Spot Fleet

To ensure that you always have instance capacity, you can include a request for On-Demand capacity
in your Spot Fleet request. In your Spot Fleet request, you specify your desired target capacity and how
much of that capacity must be On-Demand. The balance comprises Spot capacity, which is launched if
there is available Amazon EC2 capacity and availability. For example, if in your Spot Fleet request you
specify target capacity as 10 and On-Demand capacity as 8, Amazon EC2 launches 8 capacity units as
On-Demand, and 2 capacity units (10-8=2) as Spot.

Prioritize instance types for On-Demand capacity

When Spot Fleet attempts to fulfill your On-Demand capacity, it defaults to launching the lowest-priced
instance type first. If OnDemandAllocationStrategy is set to prioritized, Spot Fleet uses priority
to determine which instance type to use first in fulfilling On-Demand capacity. The priority is assigned to
the launch template override, and the highest priority is launched first.

For example, you have configured three launch template overrides, each with a different instance type:
c3.large, c4.large, and c5.large. The On-Demand price for c5.large is less than for c4.large.
c3.large is the cheapest. If you do not use priority to determine the order, the fleet fulfills On-Demand
capacity by starting with c3.large, and then c5.large. Because you often have unused Reserved
Instances for c4.large, you can set the launch template override priority so that the order is c4.large,
c3.large, and then c5.large.

Allocation strategy for Spot Instances

The allocation strategy for the Spot Instances in your Spot Fleet determines how it fulfills your Spot
Fleet request from the possible Spot Instance pools represented by its launch specifications. The
following are the allocation strategies that you can specify in your Spot Fleet request:

lowestPrice

The Spot Instances come from the pool with the lowest price. This is the default strategy.
diversified

The Spot Instances are distributed across all pools.

capacityOptimized

The Spot Instances come from the pool with optimal capacity for the number of instances that are
launching.
InstancePoolsToUseCount

The Spot Instances are distributed across the number of Spot pools that you specify. This parameter
is valid only when used in combination with lowestPrice.

372
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Maintain target capacity

After Spot Instances are terminated due to a change in the Spot price or available capacity of a Spot
Instance pool, a Spot Fleet of type maintain launches replacement Spot Instances. If the allocation
strategy is lowestPrice, the fleet launches replacement instances in the pool where the Spot price is
currently the lowest. If the allocation strategy is diversified, the fleet distributes the replacement
Spot Instances across the remaining pools. If the allocation strategy is lowestPrice in combination
with InstancePoolsToUseCount, the fleet selects the Spot pools with the lowest price and launches
Spot Instances across the number of Spot pools that you specify.

Configure Spot Fleet for cost optimization

To optimize the costs for your use of Spot Instances, specify the lowestPrice allocation strategy so
that Spot Fleet automatically deploys the least expensive combination of instance types and Availability
Zones based on the current Spot price.

For On-Demand Instance target capacity, Spot Fleet always selects the least expensive instance type
based on the public On-Demand price, while continuing to follow the allocation strategy (either
lowestPrice, capacityOptimized, or diversified) for Spot Instances.

Configure Spot Fleet for cost optimization and diversification

To create a fleet of Spot Instances that is both cheap and diversified, use the lowestPrice allocation
strategy in combination with InstancePoolsToUseCount. Spot Fleet automatically deploys the
cheapest combination of instance types and Availability Zones based on the current Spot price across the
number of Spot pools that you specify. This combination can be used to avoid the most expensive Spot
Instances.

Configure Spot Fleet for capacity optimization

With Spot Instances, pricing changes slowly over time based on long-term trends in supply and demand,
but capacity fluctuates in real time. The capacityOptimized strategy automatically launches Spot
Instances into the most available pools by looking at real-time capacity data and predicting which
are the most available. This works well for workloads such as big data and analytics, image and
media rendering, machine learning, and high performance computing that may have a higher cost of
interruption associated with restarting work and checkpointing. By offering the possibility of fewer
interruptions, the capacityOptimized strategy can lower the overall cost of your workload.

Choose an appropriate allocation strategy

You can optimize your Spot Fleets based on your use case.

If your fleet is small or runs for a short time, the probability that your Spot Instances may be interrupted
is low, even with all the instances in a single Spot Instance pool. Therefore, the lowestPrice strategy is
likely to meet your needs while providing the lowest cost.

If your fleet is large or runs for a long time, you can improve the availability of your fleet by distributing
the Spot Instances across multiple pools. For example, if your Spot Fleet request specifies 10 pools and
a target capacity of 100 instances, the fleet launches 10 Spot Instances in each pool. If the Spot price
for one pool exceeds your maximum price for this pool, only 10% of your fleet is affected. Using this
strategy also makes your fleet less sensitive to increases in the Spot price in any one pool over time.

With the diversified strategy, the Spot Fleet does not launch Spot Instances into any pools with a
Spot price that is equal to or higher than the On-Demand price.

To create a cheap and diversified fleet, use the lowestPrice strategy in combination with
InstancePoolsToUseCount. You can use a low or high number of Spot pools across which to allocate
your Spot Instances. For example, if you run batch processing, we recommend specifying a low number

373
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

of Spot pools (for example, InstancePoolsToUseCount=2) to ensure that your queue always has
compute capacity while maximizing savings. If you run a web service, we recommend specifying a high
number of Spot pools (for example, InstancePoolsToUseCount=10) to minimize the impact if a Spot
Instance pool becomes temporarily unavailable.

If your fleet runs workloads that may have a higher cost of interruption associated with restarting work
and checkpointing, then use the capacityOptimized strategy. This strategy offers the possibility of
fewer interruptions, which can lower the overall cost of your workload.

Capacity Rebalancing
You can configure Spot Fleet to launch a replacement Spot Instance when Amazon EC2 emits a rebalance
recommendation to notify you that a Spot Instance is at an elevated risk of interruption. Capacity
Rebalancing helps you maintain workload availability by proactively augmenting your fleet with a new
Spot Instance before a running instance is interrupted by Amazon EC2. For more information, see EC2
instance rebalance recommendations (p. 449).

To configure Spot Fleet to launch a replacement Spot Instance, you can use the Amazon EC2 console or
the AWS CLI.

• Amazon EC2 console: You must select the Capacity rebalance check box when you create the Spot
Fleet. For more information, see step 6.d. in Create a Spot Fleet request using defined parameters
(console) (p. 407).
• AWS CLI: Use the request-spot-fleet command and the relevant parameters in the
SpotMaintenanceStrategies structure. For more information, see the example launch
configuration (p. 427).

Limitations

• Only available for fleets of type maintain.

• When the fleet is running, you can't modify the Capacity Rebalancing setting. To change the Capacity
Rebalancing setting, you must delete the fleet and create a new fleet.

Considerations

If you configure a Spot Fleet for Capacity Rebalancing, consider the following:

Spot Fleet can launch new replacement Spot Instances until fulfilled capacity is double target
capacity

When a Spot Fleet is configured for Capacity Rebalancing, the fleet attempts to launch a new
replacement Spot Instance for every Spot Instance that receives a rebalance recommendation.
After a Spot Instance receives a rebalance recommendation, it is no longer counted as part of the
fulfilled capacity, and Spot Fleet does not automatically terminate the instance. This gives you the
opportunity to perform rebalancing actions (p. 450) on the instance. Thereafter, you can terminate
the instance, or you can leave it running.

If your fleet reaches double its target capacity, it stops launching new replacement instances even if
the replacement instances themselves receive a rebalance recommendation.

For example, you create a Spot Fleet with a target capacity of 100 Spot Instances. All the Spot
Instances receive a rebalance recommendation, which causes Spot Fleet to launch 100 replacement
Spot Instances. This raises the number of fulfilled Spot Instances to 200, which is double the target
capacity. Some of the replacement instances receive a rebalance recommendation, but no more
replacement instances are launched because the fleet cannot exceed double its target capacity.

Note that you are charged for all of the instances while they are running.

374
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

We recommend that you manually terminate Spot Instances that receive a rebalance
recommendation

If you configure your Spot Fleet for Capacity Rebalancing, we recommend that you monitor the
rebalance recommendation signal that is received by the Spot Instances in the fleet. By monitoring
the signal, you can quickly perform rebalancing actions (p. 450) on the affected instances
before Amazon EC2 interrupts them, and then you can manually terminate them. If you do not
terminate the instances, you continue paying for them while they are running. Spot Fleet does not
automatically terminate the instances that receive a rebalance recommendation.

You can set up notifications using Amazon EventBridge or instance metadata. For more information,
see Monitor rebalance recommendation signals (p. 450).
Spot Fleet does not count instances that receive a rebalance recommendation when calculating
fulfilled capacity during scale in or out

If your Spot Fleet is configured for Capacity Rebalancing, and you change the target capacity to
either scale in or scale out, the fleet does not count the instances that are marked for rebalance as
part of the fulfilled capacity, as follows:
• Scale in – If you decrease your desired target capacity, the fleet terminates instances that are
not marked for rebalance until the desired capacity is reached. The instances that are marked for
rebalance are not counted towards the fulfilled capacity.

For example, you create a Spot Fleet with a target capacity of 100 Spot Instances. 10 instances
receive a rebalance recommendation, so the fleet launches 10 new replacement instances,
resulting in a fulfilled capacity of 110 instances. You then reduce the target capacity to 50 (scale
in), but the fulfilled capacity is actually 60 instances because the 10 instances that are marked for
rebalance are not terminated by the fleet. You need to manually terminate these instances, or you
can leave them running.
• Scale out – If you increase your desired target capacity, the fleet launches new instances until the
desired capacity is reached. The instances that are marked for rebalance are not counted towards
the fulfilled capacity.

For example, you create a Spot Fleet with a target capacity of 100 Spot Instances. 10 instances
receive a rebalance recommendation, so the fleet launches 10 new replacement instances,
resulting in a fulfilled capacity of 110 instances. You then increase the target capacity to 200
(scale out), but the fullfilled capacity is actually 210 instances because the 10 instances that are
marked for rebalance are not counted by the fleet as part of the target capacity. You need to
manually terminate these instances, or you can leave them running.
Provide as many Spot Instance pools in the request as possible

Configure your Spot Fleet to use multiple instance types and Availability Zones. This provides the
flexibility to launch Spot Instances in various Spot Instance pools. For more information, see Be
flexible about instance types and Availability Zones (p. 368).
Configure your Spot Fleet to use the most optimal Spot Instance pools

Use the capacity-optimized allocation strategy to ensure that replacement Spot Instances
are launched in the most optimal Spot Instance pools. For more information, see Use the capacity
optimized allocation strategy (p. 369).

Spot price overrides

Each Spot Fleet request can include a global maximum price, or use the default (the On-Demand price).
Spot Fleet uses this as the default maximum price for each of its launch specifications.

You can optionally specify a maximum price in one or more launch specifications. This price is specific
to the launch specification. If a launch specification includes a specific price, the Spot Fleet uses this

375
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

maximum price, overriding the global maximum price. Any other launch specifications that do not
include a specific maximum price still use the global maximum price.

Control spending
Spot Fleet stops launching instances when it has either reached the target capacity or the maximum
amount you’re willing to pay. To control the amount you pay per hour for your fleet, you can specify the
SpotMaxTotalPrice for Spot Instances and the OnDemandMaxTotalPrice for On-Demand Instances.
When the maximum total price is reached, Spot Fleet stops launching instances even if it hasn’t met the
target capacity.

The following examples show two different scenarios. In the first, Spot Fleet stops launching instances
when it has met the target capacity. In the second, Spot Fleet stops launching instances when it has
reached the maximum amount you’re willing to pay.

Example: Stop launching instances when target capacity is reached

Given a request for m4.large On-Demand Instances, where:

• On-Demand Price: $0.10 per hour

• OnDemandTargetCapacity: 10
• OnDemandMaxTotalPrice: $1.50

Spot Fleet launches 10 On-Demand Instances because the total of $1.00 (10 instances x $0.10) does not
exceed the OnDemandMaxTotalPrice of $1.50.

Example: Stop launching instances when maximum total price is reached

Given a request for m4.large On-Demand Instances, where:

• On-Demand Price: $0.10 per hour

• OnDemandTargetCapacity: 10
• OnDemandMaxTotalPrice: $0.80

If Spot Fleet launches the On-Demand target capacity (10 On-Demand Instances), the total cost per
hour would be $1.00. This is more than the amount ($0.80) specified for OnDemandMaxTotalPrice.
To prevent spending more than you're willing to pay, Spot Fleet launches only 8 On-Demand
Instances (below the On-Demand target capacity) because launching more would exceed the
OnDemandMaxTotalPrice.

Spot Fleet instance weighting

When you request a fleet of Spot Instances, you can define the capacity units that each instance type
would contribute to your application's performance, and adjust your maximum price for each Spot
Instance pool accordingly using instance weighting.

By default, the price that you specify is per instance hour. When you use the instance weighting feature,
the price that you specify is per unit hour. You can calculate your price per unit hour by dividing your
price for an instance type by the number of units that it represents. Spot Fleet calculates the number
of Spot Instances to launch by dividing the target capacity by the instance weight. If the result isn't an
integer, the Spot Fleet rounds it up to the next integer, so that the size of your fleet is not below its
target capacity. Spot Fleet can select any pool that you specify in your launch specification, even if the
capacity of the instances launched exceeds the requested target capacity.

The following tables provide examples of calculations to determine the price per unit for a Spot Fleet
request with a target capacity of 10.

376
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Instance type Instance Price per Price per unit Number of instances launched
weight instance hour hour

r3.xlarge 2 $0.05 .025 5

(.05 divided by (10 divided by 2)

2)

Instance type Instance Price per Price per unit Number of instances launched
weight instance hour hour

r3.8xlarge 8 $0.10 .0125 2

(.10 divided by (10 divided by 8, result rounded

8) up)

Use Spot Fleet instance weighting as follows to provision the target capacity that you want in the pools
with the lowest price per unit at the time of fulfillment:

1. Set the target capacity for your Spot Fleet either in instances (the default) or in the units of your
choice, such as virtual CPUs, memory, storage, or throughput.
2. Set the price per unit.
3. For each launch configuration, specify the weight, which is the number of units that the instance
type represents toward the target capacity.

Instance weighting example

Consider a Spot Fleet request with the following configuration:

• A target capacity of 24
• A launch specification with an instance type r3.2xlarge and a weight of 6
• A launch specification with an instance type c3.xlarge and a weight of 5

The weights represent the number of units that instance type represents toward the target capacity. If
the first launch specification provides the lowest price per unit (price for r3.2xlarge per instance hour
divided by 6), the Spot Fleet would launch four of these instances (24 divided by 6).

If the second launch specification provides the lowest price per unit (price for c3.xlarge per instance
hour divided by 5), the Spot Fleet would launch five of these instances (24 divided by 5, result rounded
up).

Instance weighting and allocation strategy

Consider a Spot Fleet request with the following configuration:

• A target capacity of 30
• A launch specification with an instance type c3.2xlarge and a weight of 8
• A launch specification with an instance type m3.xlarge and a weight of 8
• A launch specification with an instance type r3.xlarge and a weight of 8

The Spot Fleet would launch four instances (30 divided by 8, result rounded up). With the lowestPrice
strategy, all four instances come from the pool that provides the lowest price per unit. With the

377
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

diversified strategy, the Spot Fleet launches one instance in each of the three pools, and the fourth
instance in whichever pool provides the lowest price per unit.

Walkthrough: Using Spot Fleet with instance weighting

This walkthrough uses a fictitious company called Example Corp to illustrate the process of requesting a
Spot Fleet using instance weighting.

Objective
Example Corp, a pharmaceutical company, wants to leverage the computational power of Amazon EC2
for screening chemical compounds that might be used to fight cancer.

Planning
Example Corp first reviews Spot Best Practices. Next, Example Corp determines the following
requirements for their Spot Fleet.

Instance types

Example Corp has a compute- and memory-intensive application that performs best with at least 60 GB
of memory and eight virtual CPUs (vCPUs). They want to maximize these resources for the application at
the lowest possible price. Example Corp decides that any of the following EC2 instance types would meet
their needs:

Instance type Memory (GiB) vCPUs

r3.2xlarge 61 8

r3.4xlarge 122 16

r3.8xlarge 244 32

Target capacity in units

With instance weighting, target capacity can equal a number of instances (the default) or a combination
of factors such as cores (vCPUs), memory (GiBs), and storage (GBs). By considering the base for their
application (60 GB of RAM and eight vCPUs) as 1 unit, Example Corp decides that 20 times this amount
would meet their needs. So the company sets the target capacity of their Spot Fleet request to 20.

Instance weights

After determining the target capacity, Example Corp calculates instance weights. To calculate the
instance weight for each instance type, they determine the units of each instance type that are required
to reach the target capacity as follows:

• r3.2xlarge (61.0 GB, 8 vCPUs) = 1 unit of 20

• r3.4xlarge (122.0 GB, 16 vCPUs) = 2 units of 20
• r3.8xlarge (244.0 GB, 32 vCPUs) = 4 units of 20

Therefore, Example Corp assigns instance weights of 1, 2, and 4 to the respective launch configurations
in their Spot Fleet request.

Price per unit hour

Example Corp uses the On-Demand price per instance hour as a starting point for their price. They could
also use recent Spot prices, or a combination of the two. To calculate the price per unit hour, they divide
their starting price per instance hour by the weight. For example:

378
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Instance type On-Demand price Instance weight Price per unit hour

r3.2xLarge $0.7 1 $0.7

r3.4xLarge $1.4 2 $0.7

r3.8xLarge $2.8 4 $0.7

Example Corp could use a global price per unit hour of $0.7 and be competitive for all three instance
types. They could also use a global price per unit hour of $0.7 and a specific price per unit hour of $0.9 in
the r3.8xlarge launch specification.

Verify permissions
Before creating a Spot Fleet request, Example Corp verifies that it has an IAM role with the required
permissions. For more information, see Spot Fleet permissions (p. 403).

Create the request

Example Corp creates a file, config.json, with the following configuration for its Spot Fleet request:

{
"SpotPrice": "0.70",
"TargetCapacity": 20,
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "r3.2xlarge",
"SubnetId": "subnet-482e4972",
"WeightedCapacity": 1
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "r3.4xlarge",
"SubnetId": "subnet-482e4972",
"WeightedCapacity": 2
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "r3.8xlarge",
"SubnetId": "subnet-482e4972",
"SpotPrice": "0.90",
"WeightedCapacity": 4
}
]
}

Example Corp creates the Spot Fleet request using the request-spot-fleet command.

aws ec2 request-spot-fleet --spot-fleet-request-config file://config.json

For more information, see Spot Fleet requests (p. 400).

Fulfillment
The allocation strategy determines which Spot Instance pools your Spot Instances come from.

With the lowestPrice strategy (which is the default strategy), the Spot Instances come from the pool
with the lowest price per unit at the time of fulfillment. To provide 20 units of capacity, the Spot Fleet

379
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

launches either 20 r3.2xlarge instances (20 divided by 1), 10 r3.4xlarge instances (20 divided by 2),
or 5 r3.8xlarge instances (20 divided by 4).

If Example Corp used the diversified strategy, the Spot Instances would come from all three pools.
The Spot Fleet would launch 6 r3.2xlarge instances (which provide 6 units), 3 r3.4xlarge instances
(which provide 6 units), and 2 r3.8xlarge instances (which provide 8 units), for a total of 20 units.

Spot Instance pricing history

When you request Spot Instances, we recommend that you use the default maximum price (the On-
Demand price). If you want to specify a maximum price, we recommend that you review the Spot price
history before you do so. You can view the Spot price history for the last 90 days, filtering by instance
type, operating system, and Availability Zone.

Spot Instance prices are set by Amazon EC2 and adjust gradually based on long-term trends in supply
and demand for Spot Instance capacity. For the current Spot Instance prices see Amazon EC2 Spot
Instances Pricing.

To view the Spot price history (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Spot Requests.
3. If you are new to Spot Instances, you see a welcome page. Choose Get started, scroll to the bottom
of the screen, and then choose Cancel.
4. Choose Pricing history.
5. Choose the operating system (Product), Instance type, and Date range for which to view the price
history. Move your pointer over the graph to display the prices at specific times in the selected date
range.

6. (Optional) To review the Spot price history for a specific Availability Zone, you can filter the
Availability Zones by removing Availability Zones from the graph. To remove an Availability Zone
from the graph, select the zone to remove it. You can also select a different product, instance type,
or date range.

To view the Spot price history using the command line

You can use one of the following commands. For more information, see Access Amazon EC2 (p. 3).

• describe-spot-price-history (AWS CLI)

380
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

• Get-EC2SpotPriceHistory (AWS Tools for Windows PowerShell)

Savings from purchasing Spot Instances

You can view the usage and savings information for Spot Instances at the per-fleet level, or for all
running Spot Instances. At the per-fleet level, the usage and savings information includes all instances
launched and terminated by the fleet. You can view this information from the last hour or the last three
days.

The following screenshot from the Spot Requests page shows the Spot usage and savings information
for a Spot Fleet.

You can view the following usage and savings information:

• Spot Instances – The number of Spot Instances launched and terminated by the Spot Fleet. When
viewing the savings summary, the number represents all your running Spot Instances.
• vCPU-hours – The number of vCPU hours used across all the Spot Instances for the selected time
frame.
• Mem(GiB)-hours – The number of GiB hours used across all the Spot Instances for the selected time
frame.
• On-Demand total – The total amount you would've paid for the selected time frame had you launched
these instances as On-Demand Instances.
• Spot total – The total amount to pay for the selected time frame.
• Savings – The percentage that you are saving by not paying the On-Demand price.
• Average cost per vCPU-hour – The average hourly cost of using the vCPUs across all the Spot
Instances for the selected time frame, calculated as follows: Average cost per vCPU-hour = Spot total
/ vCPU-hours.
• Average cost per mem(GiB)-hour – The average hourly cost of using the GiBs across all the Spot
Instances for the selected time frame, calculated as follows: Average cost per mem(GiB)-hour = Spot
total / Mem(GiB)-hours.
• Details table – The different instance types (the number of instances per instance type is in
parentheses) that comprise the Spot Fleet. When viewing the savings summary, these comprise all
your running Spot Instances.

Savings information can only be viewed using the Amazon EC2 console.

To view the savings information for a Spot Fleet (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

381
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

2. On the navigation pane, choose Spot Requests.

3. Select a Spot Fleet request and choose Savings.
4. By default, the page displays usage and savings information for the last three days. You can choose
last hour or the last three days. For Spot Fleets that were launched less than an hour ago, the page
shows the estimated savings for the hour.

To view the savings information for all running Spot Instances (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. On the navigation pane, choose Spot Requests.
3. Choose Savings Summary.

Spot Instance requests

To use Spot Instances, you create a Spot Instance request that includes the desired number of instances,
the instance type, the Availability Zone, and the maximum price that you are willing to pay per
instance hour. If your maximum price exceeds the current Spot price, Amazon EC2 fulfills your request
immediately if capacity is available. Otherwise, Amazon EC2 waits until your request can be fulfilled or
until you cancel the request.

The following illustration shows how Spot requests work. Notice that the request type (one-time
or persistent) determines whether the request is opened again when Amazon EC2 interrupts a Spot
Instance or if you stop a Spot Instance. If the request is persistent, the request is opened again after your
Spot Instance is interrupted. If the request is persistent and you stop your Spot Instance, the request only
opens after you start your Spot Instance.

Contents
• Spot Instance request states (p. 383)
• Define a duration for your Spot Instances (p. 384)

382
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

• Specify a tenancy for your Spot Instances (p. 385)

• Service-linked role for Spot Instance requests (p. 385)
• Create a Spot Instance request (p. 387)
• Find running Spot Instances (p. 389)
• Tag Spot Instance requests (p. 390)
• Cancel a Spot Instance request (p. 395)
• Stop a Spot Instance (p. 396)
• Start a Spot Instance (p. 397)
• Terminate a Spot Instance (p. 398)
• Spot Instance request example launch specifications (p. 399)

Spot Instance request states

A Spot Instance request can be in one of the following states:

• open – The request is waiting to be fulfilled.

• active – The request is fulfilled and has an associated Spot Instance.
• failed – The request has one or more bad parameters.
• closed – The Spot Instance was interrupted or terminated.
• disabled – You stopped the Spot Instance.
• cancelled – You canceled the request, or the request expired.

The following illustration represents the transitions between the request states. Notice that the
transitions depend on the request type (one-time or persistent).

A one-time Spot Instance request remains active until Amazon EC2 launches the Spot Instance, the
request expires, or you cancel the request. If the Spot price exceeds your maximum price or capacity is
not available, your Spot Instance is terminated and the Spot Instance request is closed.

383
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

A persistent Spot Instance request remains active until it expires or you cancel it, even if the request is
fulfilled. If the Spot price exceeds your maximum price or capacity is not available, your Spot Instance
is interrupted. After your instance is interrupted, when your maximum price exceeds the Spot price or
capacity becomes available again, the Spot Instance is started if stopped or resumed if hibernated. You
can stop a Spot Instance and start it again if capacity is available and your maximum price exceeds the
current Spot price. If the Spot Instance is terminated (irrespective of whether the Spot Instance is in a
stopped or running state), the Spot Instance request is opened again and Amazon EC2 launches a new
Spot Instance. For more information, see Stop a Spot Instance (p. 396), Start a Spot Instance (p. 397),
and Terminate a Spot Instance (p. 398).

You can track the status of your Spot Instance requests, as well as the status of the Spot Instances
launched, through the status. For more information, see Spot request status (p. 443).

Define a duration for your Spot Instances

Spot Instances with a defined duration (also known as Spot blocks) are designed not to be interrupted
and will run continuously for the duration you select. This makes them ideal for jobs that take a finite
time to complete, such as batch processing, encoding and rendering, modeling and analysis, and
continuous integration.

You can use a duration of 1, 2, 3, 4, 5, or 6 hours. The price that you pay depends on the specified
duration. To view the current prices for a 1-hour duration or a 6-hour duration, see Spot Instance Prices.
You can use these prices to estimate the cost of the 2, 3, 4, and 5-hour durations. When a request with
a duration is fulfilled, the price for your Spot Instance is fixed, and this price remains in effect until the
instance terminates. You are billed at this price for each hour or partial hour that the instance is running.
A partial instance hour is billed to the nearest second.

When you define a duration in your Spot request, the duration period for each Spot Instance starts as
soon as the instance receives its instance ID. The Spot Instance runs until you terminate it or the duration
period ends. At the end of the duration period, Amazon EC2 marks the Spot Instance for termination
and provides a Spot Instance termination notice, which gives the instance a two-minute warning before
it terminates. In rare situations, Spot blocks may be interrupted due to Amazon EC2 capacity needs. In
these cases, we provide a two-minute warning before we terminate an instance, and you are not charged
for the terminated instances even if you used them.

New accounts or accounts with no previous billing history with AWS are not eligible for Spot Instances
with a defined duration (also known as Spot blocks).

To launch Spot Instances with a defined duration (console)

Follow the Create a Spot Fleet request (p. 407) procedure. To launch Spot Instances with a defined
duration, for Tell us your application or task need, choose Defined duration workloads.

To launch Spot Instances with a defined duration (AWS CLI)

To specify a duration for your Spot Instances, include the --block-duration-minutes option with
the request-spot-instances command. For example, the following command creates a Spot request that
launches Spot Instances that run for two hours.

aws ec2 request-spot-instances \

--instance-count 5 \
--block-duration-minutes 120 \
--type "one-time" \
--launch-specification file://specification.json

To retrieve the cost for Spot Instances with a defined duration (AWS CLI)

Use the describe-spot-instance-requests command to retrieve the fixed cost for your Spot Instances with
a specified duration. The information is in the actualBlockHourlyPrice field.

384
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Specify a tenancy for your Spot Instances

You can run a Spot Instance on single-tenant hardware. Dedicated Spot Instances are physically
isolated from instances that belong to other AWS accounts. For more information, see Dedicated
Instances (p. 496) and the Amazon EC2 Dedicated Instances product page.

To run a Dedicated Spot Instance, do one of the following:

• Specify a tenancy of dedicated when you create the Spot Instance request. For more information,
see Create a Spot Instance request (p. 387).
• Request a Spot Instance in a VPC with an instance tenancy of dedicated. For more information, see
Create a VPC with an instance tenancy of dedicated (p. 498). You cannot request a Spot Instance with
a tenancy of default if you request it in a VPC with an instance tenancy of dedicated.

The following instance types support Dedicated Spot Instances.

Current generation

• c4.8xlarge
• d2.8xlarge
• i3.16xlarge
• m4.10xlarge
• m4.16xlarge
• p2.16xlarge
• r4.16xlarge
• x1.32xlarge

Previous generation

• c3.8xlarge
• cc2.8xlarge
• cr1.8xlarge
• g2.8xlarge
• i2.8xlarge
• r3.8xlarge

Service-linked role for Spot Instance requests

Amazon EC2 uses service-linked roles for the permissions that it requires to call other AWS services on
your behalf. A service-linked role is a unique type of IAM role that is linked directly to an AWS service.
Service-linked roles provide a secure way to delegate permissions to AWS services because only the
linked service can assume a service-linked role. For more information, see Using Service-Linked Roles in
the IAM User Guide.

Amazon EC2 uses the service-linked role named AWSServiceRoleForEC2Spot to launch and manage
Spot Instances on your behalf.

Permissions granted by AWSServiceRoleForEC2Spot

Amazon EC2 uses AWSServiceRoleForEC2Spot to complete the following actions:

• ec2:DescribeInstances – Describe Spot Instances

385
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

• ec2:StopInstances – Stop Spot Instances

• ec2:StartInstances – Start Spot Instances

Create the service-linked role

Under most circumstances, you don't need to manually create a service-linked role. Amazon EC2 creates
the AWSServiceRoleForEC2Spot service-linked role the first time you request a Spot Instance using the
console.

If you had an active Spot Instance request before October 2017, when Amazon EC2 began supporting
this service-linked role, Amazon EC2 created the AWSServiceRoleForEC2Spot role in your AWS account.
For more information, see A New Role Appeared in My Account in the IAM User Guide.

Ensure that this role exists before you use the AWS CLI or an API to request a Spot Instance.

To create AWSServiceRoleForEC2Spot using the console

1. Open the IAM console at https://console.aws.amazon.com/iam/.

2. In the navigation pane, choose Roles.
3. Choose Create role.
4. On the Select type of trusted entity page, choose EC2, EC2 - Spot Instances, Next: Permissions.
5. On the next page, choose Next:Review.
6. On the Review page, choose Create role.

To create AWSServiceRoleForEC2Spot using the AWS CLI

Use the create-service-linked-role command as follows.

aws iam create-service-linked-role --aws-service-name spot.amazonaws.com

If you no longer need to use Spot Instances, we recommend that you delete the
AWSServiceRoleForEC2Spot role. After this role is deleted from your account, Amazon EC2 will create
the role again if you request Spot Instances.

Grant access to CMKs for use with encrypted AMIs and EBS snapshots

If you specify an encrypted AMI (p. 161) or an encrypted Amazon EBS snapshot (p. 1176) for your
Spot Instances and you use a customer managed customer master key (CMK) for encryption, you must
grant the AWSServiceRoleForEC2Spot role permission to use the CMK so that Amazon EC2 can launch
Spot Instances on your behalf. To do this, you must add a grant to the CMK, as shown in the following
procedure.

When providing permissions, grants are an alternative to key policies. For more information, see Using
Grants and Using Key Policies in AWS KMS in the AWS Key Management Service Developer Guide.

To grant the AWSServiceRoleForEC2Spot role permissions to use the CMK

• Use the create-grant command to add a grant to the CMK and to specify the principal (the
AWSServiceRoleForEC2Spot service-linked role) that is given permission to perform the operations
that the grant permits. The CMK is specified by the key-id parameter and the ARN of the
CMK. The principal is specified by the grantee-principal parameter and the ARN of the
AWSServiceRoleForEC2Spot service-linked role.

aws kms create-grant \

386
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

--region us-east-1 \
--key-id arn:aws:kms:us-
east-1:444455556666:key/1234abcd-12ab-34cd-56ef-1234567890ab \
--grantee-principal arn:aws:iam::111122223333:role/AWSServiceRoleForEC2Spot \
--operations "Decrypt" "Encrypt" "GenerateDataKey"
"GenerateDataKeyWithoutPlaintext" "CreateGrant" "DescribeKey" "ReEncryptFrom"
"ReEncryptTo"

Create a Spot Instance request

The procedure for requesting a Spot Instance is similar to the procedure for launching an On-Demand
Instance. You can request a Spot Instance in the following ways:

• To request a Spot Instance using the console, use the launch instance wizard. For more information,
see To create a Spot Instance request (console) (p. 387).
• To request a Spot Instance using the CLI, use the request-spot-instances command or the run-instances
command. For more information, see To create a Spot Instance request using request-spot-instances
(AWS CLI) and To create a Spot Instance request using run-instances (AWS CLI).
• To request a Spot Instance with a defined duration using the console, follow the Create a Spot Fleet
request (p. 407) procedure. For Tell us your application or task need, choose Defined duration
workloads. For more information, see Define a duration for your Spot Instances (p. 384).
• To request a Spot Instance with a defined duration using the CLI, use the request-spot-instances
command and specify the --block-duration-minutes parameter. For more information, see
Define a duration for your Spot Instances (p. 384).

After you've submitted your Spot Instance request, you can't change the parameters of the request. This
means that you can't make changes to the maximum price that you're willing to pay.

If you request multiple Spot Instances at one time, Amazon EC2 creates separate Spot Instance requests
so that you can track the status of each request separately. For more information about tracking Spot
Instance requests, see Spot request status (p. 443).

To launch a fleet that includes Spot Instances and On-Demand Instances, see Create a Spot Fleet
request (p. 407).
Note
You can't launch a Spot Instance and an On-Demand Instance in the same call using the launch
instance wizard or the run-instances command.

Prerequisites

Before you begin, decide on your maximum price, how many Spot Instances you'd like, and what instance
type to use. To review Spot price trends, see Spot Instance pricing history (p. 380).

To create a Spot Instance request (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation bar at the top of the screen, select a Region.
3. From the Amazon EC2 console dashboard, choose Launch Instance.
4. On the Choose an Amazon Machine Image (AMI) page, choose an AMI. For more information, see
Step 1: Choose an Amazon Machine Image (AMI) (p. 527).
5. On the Choose an Instance Type page, select the hardware configuration and size of the instance
to launch, and then choose Next: Configure Instance Details. For more information, see Step 2:
Choose an Instance Type (p. 528).

387
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

6. On the Configure Instance Details page, configure the Spot Instance request as follows:

• Number of instances: Enter the number of instances to launch.

Note
Amazon EC2 creates a separate request for each Spot Instance.
• (Optional) To help ensure that you maintain the correct number of instances to handle demand
on your application, you can choose Launch into Auto Scaling Group to create a launch
configuration and an Auto Scaling group. Auto Scaling scales the number of instances in the group
according to your specifications. For more information, see the Amazon EC2 Auto Scaling User
Guide.
• Purchasing option: Choose Request Spot instances to launch a Spot Instance. When you choose
this option, the following fields appear.
• Current price: The current Spot price in each Availability Zone is displayed for the instance type
that you selected.
• (Optional) Maximum price: You can leave the field empty, or you can specify the maximum
amount you're willing to pay.
• If you leave the field empty, then the maximum price defaults to the current On-Demand price.
Your Spot Instance launches at the current Spot price, not exceeding the On-Demand price.
• If you specify a maximum price that is more than the current Spot Price, your Spot Instance
launches and is charged at the current Spot price.
• If you specify a maximum price that is lower than the Spot price, your Spot Instance is not
launched.
• Persistent request: Choose Persistent request to resubmit the Spot Instance request if your Spot
Instance is interrupted.
• Interruption behavior: By default, the Spot service terminates a Spot Instance when it is
interrupted. If you choose Persistent request, you can then specify that the Spot service stops
or hibernates your Spot Instance when it's interrupted. For more information, see Interruption
behaviors (p. 453).
• (Optional) Request valid to: Choose Edit to specify when the Spot Instance request expires.

For more information about configuring your Spot Instance, see Step 3: Configure Instance
Details (p. 528).
7. The AMI you selected includes one or more volumes of storage, including the root device volume. On
the Add Storage page, you can specify additional volumes to attach to the instance by choosing Add
New Volume. For more information, see Step 4: Add Storage (p. 530).
8. On the Add Tags page, specify tags (p. 1311) by providing key and value combinations. For more
information, see Step 5: Add Tags (p. 531).
9. On the Configure Security Group page, use a security group to define firewall rules for your
instance. These rules specify which incoming network traffic is delivered to your instance. All other
traffic is ignored. (For more information about security groups, see Amazon EC2 security groups for
Linux instances (p. 1058).) Select or create a security group, and then choose Review and Launch.
For more information, see Step 6: Configure Security Group (p. 531).
10. On the Review Instance Launch page, check the details of your instance, and make any necessary
changes by choosing the appropriate Edit link. When you are ready, choose Launch. For more
information, see Step 7: Review Instance Launch and Select Key Pair (p. 531).
11. In the Select an existing key pair or create a new key pair dialog box, you can choose an existing
key pair, or create a new one. For example, choose Choose an existing key pair, then select the key
pair that you created when getting set up. For more information, see Amazon EC2 key pairs and
Linux instances (p. 1044).
Important
If you choose the Proceed without key pair option, you won't be able to connect to the
instance unless you choose an AMI that is configured to allow users another way to log in.

388
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

12. To launch your instance, select the acknowledgment check box, then choose Launch Instances.

If the instance fails to launch or the state immediately goes to terminated instead of running, see
Troubleshoot instance launch issues (p. 1326).

To create a Spot Instance request using request-spot-instances (AWS CLI)

Use the request-spot-instances command to create a one-time request.

aws ec2 request-spot-instances \

--instance-count 5 \
--type "one-time" \
--launch-specification file://specification.json

Use the request-spot-instances command to create a persistent request.

aws ec2 request-spot-instances \

--instance-count 5 \
--type "persistent" \
--launch-specification file://specification.json

For example launch specification files to use with these commands, see Spot Instance request example
launch specifications (p. 399). If you download a launch specification file from the console, you must
use the request-spot-fleet command instead (the console specifies a Spot request using a Spot Fleet).

To create a Spot Instance request using run-instances (AWS CLI)

Use the run-instances command and specify the Spot Instance options in the --instance-market-
options parameter.

aws ec2 run-instances \

--image-id ami-0abcdef1234567890 \
--instance-type t2.micro \
--count 5 \
--subnet-id subnet-08fc749671b2d077c \
--key-name MyKeyPair \
--security-group-ids sg-0b0384b66d7d692f9 \
--instance-market-options file://spot-options.json

The following is the data structure to specify in the JSON file for --instance-market-options. You
can also specify BlockDurationMinutes, ValidUntil, and InstanceInterruptionBehavior. If
you do not specify a field in the data structure, the default value is used. This example creates a one-
time request and specifies 0.02 as the maximum price you're willing to pay for the Spot Instance.

{
"MarketType": "spot",
"SpotOptions": {
"MaxPrice": "0.02",
"SpotInstanceType": "one-time"
}
}

Find running Spot Instances

Amazon EC2 launches a Spot Instance when the maximum price exceeds the Spot price and capacity is
available. A Spot Instance runs until it is interrupted or you terminate it yourself. If your maximum price
is exactly equal to the Spot price, there is a chance that your Spot Instance remains running, depending
on demand.

389
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

To find running Spot Instances (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Spot Requests. You can see both Spot Instance requests and Spot
Fleet requests. If a Spot Instance request has been fulfilled, Capacity is the ID of the Spot Instance.
For a Spot Fleet, Capacity indicates how much of the requested capacity has been fulfilled. To view
the IDs of the instances in a Spot Fleet, choose the expand arrow, or select the fleet and choose
Instances.
Note
For Spot Instance requests that are created by a Spot Fleet, the requests are not tagged
instantly with the system tag that indicates the Spot Fleet to which they belong, and for a
period of time may appear separate from Spot Fleet request.

Alternatively, in the navigation pane, choose Instances. In the top right corner, choose the settings

icon ( ), and then under Attribute columns, select Instance lifecycle. For each instance, Instance
lifecycle is either normal, spot, or scheduled.

To find running Spot Instances (AWS CLI)

To enumerate your Spot Instances, use the describe-spot-instance-requests command with the --query
option.

aws ec2 describe-spot-instance-requests \

--query "SpotInstanceRequests[*].{ID:InstanceId}"

The following is example output:

[
{
"ID": "i-1234567890abcdef0"
},
{
"ID": "i-0598c7d356eba48d7"
}
]

Alternatively, you can enumerate your Spot Instances using the describe-instances command with the --
filters option.

aws ec2 describe-instances \

--filters "Name=instance-lifecycle,Values=spot"

To describe a single Spot Instance instance, use the describe-spot-instance-requests command with the
--spot-instance-request-ids option.

aws ec2 describe-spot-instance-requests \

--spot-instance-request-ids sir-08b93456

Tag Spot Instance requests

To help categorize and manage your Spot Instance requests, you can tag them with custom metadata.
You can assign a tag to a Spot Instance request when you create it, or afterward. You can assign tags
using the Amazon EC2 console or a command line tool.

390
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

When you tag a Spot Instance request, the instances and volumes that are launched by the Spot Instance
request are not automatically tagged. You need to explicitly tag the instances and volumes launched
by the Spot Instance request. You can assign a tag to a Spot Instance and volumes during launch, or
afterward.

For more information about how tags work, see Tag your Amazon EC2 resources (p. 1311).

Contents
• Prerequisites (p. 391)
• Tag a new Spot Instance request (p. 393)
• Tag an existing Spot Instance request (p. 393)
• View Spot Instance request tags (p. 394)

Prerequisites

Grant the IAM user the permission to tag resources. For more information about IAM policies and
example policies, see Example: Tag resources (p. 1016).

The IAM policy you create is determined by which method you use for creating a Spot Instance request.

• If you use the launch instance wizard or run-instances to request Spot Instances, see To grant an
IAM user the permission to tag resources when using the launch instance wizard or run-instances.
• If you use the Spot console to request Spot Instances with a defined duration or use the request-
spot-instances command to request Spot Instances, see To grant an IAM user the permission to tag
resources when using request-spot-instances.

To grant an IAM user the permission to tag resources when using the launch instance wizard or run-
instances

Create a IAM policy that includes the following:

• The ec2:RunInstances action. This grants the IAM user permission to launch an instance.
• For Resource, specify spot-instances-request. This allows users to create Spot Instance
requests, which request Spot Instances.
• The ec2:CreateTags action. This grants the IAM user permission to create tags.
• For Resource, specify *. This allows users to tag all resources that are created during instance launch.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowLaunchInstances",
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:us-east-1::image/*",
"arn:aws:ec2:us-east-1:*:subnet/*",
"arn:aws:ec2:us-east-1:*:network-interface/*",
"arn:aws:ec2:us-east-1:*:security-group/*",
"arn:aws:ec2:us-east-1:*:key-pair/*",
"arn:aws:ec2:us-east-1:*:volume/*",
"arn:aws:ec2:us-east-1:*:instance/*",
"arn:aws:ec2:us-east-1:*:spot-instances-request/*"
]

391
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

},
{
"Sid": "TagSpotInstanceRequests",
"Effect": "Allow",
"Action": "ec2:CreateTags",
"Resource": "*"
}
]
}

Note
When you use the RunInstances action to create Spot Instance requests and tag the Spot
Instance requests on create, you need to be aware of how Amazon EC2 evaluates the spot-
instances-request resource in the RunInstances statement.
The spot-instances-request resource is evaluated in the IAM policy as follows:

• If you don't tag a Spot Instance request on create, Amazon EC2 does not evaluate the spot-
instances-request resource in the RunInstances statement.
• If you tag a Spot Instance request on create, Amazon EC2 evaluates the spot-instances-
request resource in the RunInstances statement.

Therefore, for the spot-instances-request resource, the following rules apply to the IAM
policy:

• If you use RunInstances to create a Spot Instance request and you don't intend to tag the Spot
Instance request on create, you don’t need to explicitly allow the spot-instances-request
resource; the call will succeed.
• If you use RunInstances to create a Spot Instance request and intend to tag the Spot Instance
request on create, you must include the spot-instances-request resource in the
RunInstances allow statement, otherwise the call will fail.
• If you use RunInstances to create a Spot Instance request and intend to tag the Spot Instance
request on create, you must specify the spot-instances-request resource or include a *
wildcard in the CreateTags allow statement, otherwise the call will fail.

For example IAM policies, including policies that are not supported for Spot Instance requests,
see Work with Spot Instances (p. 1011).

To grant an IAM user the permission to tag resources when using request-spot-instances

Create a IAM policy that includes the following:

• The ec2:RequestSpotInstances action. This grants the IAM user permission to create a Spot
Instance request.
• The ec2:CreateTags action. This grants the IAM user permission to create tags.
• For Resource, specify spot-instances-request. This allows users to tag only the Spot Instance
request.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "TagSpotInstanceRequest",
"Effect": "Allow",
"Action": [
"ec2:RequestSpotInstances",
"ec2:CreateTags"

392
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

],
"Resource": "arn:aws:ec2:us-east-1:111122223333:spot-instances-request/*"
}

Tag a new Spot Instance request

To tag a new Spot Instance request using the console

1. Follow the Create a Spot Instance request (p. 387) procedure.

2. To add a tag, on the Add Tags page, choose Add Tag, and enter the key and value for the tag.
Choose Add another tag for each additional tag.

For each tag, you can tag the Spot Instance request, the Spot Instances, and the volumes with
the same tag. To tag all three, ensure that Instances, Volumes, and Spot Instance Requests are
selected. To tag only one or two, ensure that the resources you want to tag are selected, and the
other resources are cleared.
3. Complete the required fields to create a Spot Instance request, and then choose Launch. For more
information, see Create a Spot Instance request (p. 387).

To tag a new Spot Instance request using the AWS CLI

To tag a Spot Instance request when you create it, configure the Spot Instance request configuration as
follows:

• Specify the tags for the Spot Instance request using the --tag-specification parameter.
• For ResourceType, specify spot-instances-request. If you specify another value, the Spot
Instance request will fail.
• For Tags, specify the key-value pair. You can specify more than one key-value pair.

In the following example, the Spot Instance request is tagged with two tags: Key=Environment and
Value=Production, and Key=Cost-Center and Value=123.

aws ec2 request-spot-instances \

--instance-count 5 \
--type "one-time" \
--launch-specification file://specification.json \
--tag-specification 'ResourceType=spot-instances-
request,Tags=[{Key=Environment,Value=Production},{Key=Cost-Center,Value=123}]'

Tag an existing Spot Instance request

To tag an existing Spot Instance request using the console

After you have created a Spot Instance request, you can add tags to the Spot Instance request using the
console.

1. Open the Spot console at https://console.aws.amazon.com/ec2spot.

2. Select your Spot Instance request.
3. Choose the Tags tab and choose Create Tag.

To tag an existing Spot Instance using the console

After your Spot Instance request has launched your Spot Instance, you can add tags to the instance using
the console. For more information, see Add and delete tags on an individual resource (p. 1317).

To tag an existing Spot Instance request or Spot Instance using the AWS CLI

393
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Use the create-tags command to tag existing resources. In the following example, the existing Spot
Instance request and the Spot Instance are tagged with Key=purpose and Value=test.

aws ec2 create-tags \

--resources sir-08b93456 i-1234567890abcdef0 \
--tags Key=purpose,Value=test

View Spot Instance request tags

To view Spot Instance request tags using the console

1. Open the Spot console at https://console.aws.amazon.com/ec2spot.

2. Select your Spot Instance request and choose the Tags tab.

To describe Spot Instance request tags

Use the describe-tags command to view the tags for the specified resource. In the following example,
you describe the tags for the specified request.

aws ec2 describe-tags \

--filters "Name=resource-id,Values=sir-11112222-3333-4444-5555-66666EXAMPLE"

{
"Tags": [
{
"Key": "Environment",
"ResourceId": "sir-11112222-3333-4444-5555-66666EXAMPLE",
"ResourceType": "spot-instances-request",
"Value": "Production"
},
{
"Key": "Another key",
"ResourceId": "sir-11112222-3333-4444-5555-66666EXAMPLE",
"ResourceType": "spot-instances-request",
"Value": "Another value"
}
]
}

You can also view the tags of a Spot Instance request by describing the Spot Instance request.

Use the describe-spot-instance-requests command to view the configuration of the specified Spot
Instance request, which includes any tags that were specified for the request.

aws ec2 describe-spot-instance-requests \

--spot-instance-request-ids sir-11112222-3333-4444-5555-66666EXAMPLE

{
"SpotInstanceRequests": [
{
"CreateTime": "2020-06-24T14:22:11+00:00",
"InstanceId": "i-1234567890EXAMPLE",
"LaunchSpecification": {
"SecurityGroups": [
{
"GroupName": "launch-wizard-6",
"GroupId": "sg-1234567890EXAMPLE"
}

394
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

],
"BlockDeviceMappings": [
{
"DeviceName": "/dev/xvda",
"Ebs": {
"DeleteOnTermination": true,
"VolumeSize": 8,
"VolumeType": "gp2"
}
}
],
"ImageId": "ami-1234567890EXAMPLE",
"InstanceType": "t2.micro",
"KeyName": "my-key-pair",
"NetworkInterfaces": [
{
"DeleteOnTermination": true,
"DeviceIndex": 0,
"SubnetId": "subnet-11122233"
}
],
"Placement": {
"AvailabilityZone": "eu-west-1c",
"Tenancy": "default"
},
"Monitoring": {
"Enabled": false
}
},
"LaunchedAvailabilityZone": "eu-west-1c",
"ProductDescription": "Linux/UNIX",
"SpotInstanceRequestId": "sir-1234567890EXAMPLE",
"SpotPrice": "0.012600",
"State": "active",
"Status": {
"Code": "fulfilled",
"Message": "Your spot request is fulfilled.",
"UpdateTime": "2020-06-25T18:30:21+00:00"
},
"Tags": [
{
"Key": "Environment",
"Value": "Production"
},
{
"Key": "Another key",
"Value": "Another value"
}
],
"Type": "one-time",
"InstanceInterruptionBehavior": "terminate"
}
]
}

Cancel a Spot Instance request

If you no longer want your Spot Instance request, you can cancel it. You can only cancel Spot Instance
requests that are open, active, or disabled.

• Your Spot Instance request is open when your request has not yet been fulfilled and no instances have
been launched.
• Your Spot Instance request is active when your request has been fulfilled and Spot Instances have
launched as a result.

395
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

• Your Spot Instance request is disabled when you stop your Spot Instance.

If your Spot Instance request is active and has an associated running Spot Instance, canceling the
request does not terminate the instance. For more information about terminating a Spot Instance, see
Terminate a Spot Instance (p. 398).

To cancel a Spot Instance request (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Spot Requests and select the Spot request.
3. Choose Actions, Cancel request.
4. (Optional) If you are finished with the associated Spot Instances, you can terminate them. In the
Cancel Spot request dialog box, select Terminate instances, and then choose Confirm.

To cancel a Spot Instance request (AWS CLI)

• Use the cancel-spot-instance-requests command to cancel the specified Spot request.

aws ec2 cancel-spot-instance-requests --spot-instance-request-ids sir-08b93456

Stop a Spot Instance

If you don’t need your Spot Instances now, but you want to restart them later without losing the data
persisted in the Amazon EBS volume, you can stop them. The steps for stopping a Spot Instance are
similar to the steps for stopping an On-Demand Instance. You can only stop a Spot Instance if the Spot
Instance was launched from a persistent Spot Instance request.
Note
While a Spot Instance is stopped, you can modify some of its instance attributes, but not the
instance type.
We don't charge usage for a stopped Spot Instance, or data transfer fees, but we do charge for
the storage for any Amazon EBS volumes.

Limitations

• You can't stop a Spot Instance if it is part of a fleet or launch group, Availability Zone group, or Spot
block.

New console

To stop a Spot Instance (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and select the Spot Instance.
3. Choose Instance state, Stop instance.
4. When prompted for confirmation, choose Stop.

Old console

To stop a Spot Instance (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

396
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

2. In the navigation pane, choose Instances and select the Spot Instance.
3. Choose Actions, Instance State, Stop.

AWS CLI

To stop a Spot Instance (AWS CLI)

• Use the stop-instances command to manually stop one or more Spot Instances.

aws ec2 stop-instances --instance-ids i-1234567890abcdef0

Start a Spot Instance

You can start a Spot Instance that you previously stopped. The steps for starting a Spot Instance are
similar to the steps for starting an On-Demand Instance.

Prerequisites

You can only start a Spot Instance if:

• You manually stopped the Spot Instance.

• The Spot Instance is an EBS-backed instance.
• Spot Instance capacity is available.
• The Spot price is lower than your maximum price.

Limitations

• You can't start a Spot Instance if it is part of fleet or launch group, Availability Zone group, or Spot
block.

New console

To start a Spot Instance (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and select the Spot Instance.
3. Choose Instance state, Start instance.

Old console

To start a Spot Instance (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and select the Spot Instance.
3. Choose Actions, Instance State, Start.

AWS CLI

To start a Spot Instance (AWS CLI)

• Use the start-instances command to manually start one or more Spot Instances.

397
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

aws ec2 start-instances --instance-ids i-1234567890abcdef0

Terminate a Spot Instance

If your Spot Instance request is active and has an associated running Spot Instance, or your Spot
Instance request is disabled and has an associated stopped Spot Instance, canceling the request does
not terminate the instance; you must terminate the running Spot Instance manually.

If you terminate a running or stopped Spot Instance that was launched by a persistent Spot request,
the Spot request returns to the open state so that a new Spot Instance can be launched. To cancel a
persistent Spot request and terminate its Spot Instances, you must cancel the Spot request first and then
terminate the Spot Instances. Otherwise, the persistent Spot request can launch a new instance. For
more information about canceling a Spot Instance request, see Cancel a Spot Instance request (p. 395).

New console

To manually terminate a Spot Instance using the console

1. Before you terminate an instance, verify that you won't lose any data by checking that your
Amazon EBS volumes won't be deleted on termination and that you've copied any data that you
need from your instance store volumes to persistent storage, such as Amazon EBS or Amazon
S3.
2. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.
3. In the navigation pane, choose Instances.
4. To confirm that the instance is a Spot Instance, check that spot appears in the Instance lifecycle
column.
5. Select the instance, and choose Actions, Instance state, Terminate instance.
6. Choose Terminate when prompted for confirmation.

Old console

To manually terminate a Spot Instance using the console

1. Before you terminate an instance, verify that you won't lose any data by checking that your
Amazon EBS volumes won't be deleted on termination and that you've copied any data that you
need from your instance store volumes to persistent storage, such as Amazon EBS or Amazon
S3.
2. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.
3. In the navigation pane, choose Instances.
4. To confirm that the instance is a Spot Instance, check that spot appears in the Lifecycle column.
5. Select the instance, and choose Actions, Instance State, Terminate.
6. Choose Yes, Terminate when prompted for confirmation.

AWS CLI

To manually terminate a Spot Instance using the AWS CLI

• Use the terminate-instances command to manually terminate Spot Instances.

aws ec2 terminate-instances --instance-ids i-1234567890abcdef0 i-0598c7d356eba48d7

398
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Spot Instance request example launch specifications

The following examples show launch configurations that you can use with the request-spot-instances
command to create a Spot Instance request. For more information, see Create a Spot Instance
request (p. 387).

1. Launch Spot Instances (p. 399)

2. Launch Spot Instances in the specified Availability Zone (p. 399)
3. Launch Spot Instances in the specified subnet (p. 399)
4. Launch a Dedicated Spot Instance (p. 400)

Example 1: Launch Spot Instances

The following example does not include an Availability Zone or subnet. Amazon EC2 selects an
Availability Zone for you. Amazon EC2 launches the instances in the default subnet of the selected
Availability Zone.

{
"ImageId": "ami-1a2b3c4d",
"KeyName": "my-key-pair",
"SecurityGroupIds": [ "sg-1a2b3c4d" ],
"InstanceType": "m3.medium",
"IamInstanceProfile": {
"Arn": "arn:aws:iam::123456789012:instance-profile/my-iam-role"
}
}

Example 2: Launch Spot Instances in the specified Availability Zone

The following example includes an Availability Zone. Amazon EC2 launches the instances in the default
subnet of the specified Availability Zone.

{
"ImageId": "ami-1a2b3c4d",
"KeyName": "my-key-pair",
"SecurityGroupIds": [ "sg-1a2b3c4d" ],
"InstanceType": "m3.medium",
"Placement": {
"AvailabilityZone": "us-west-2a"
},
"IamInstanceProfile": {
"Arn": "arn:aws:iam::123456789012:instance-profile/my-iam-role"
}
}

Example 3: Launch Spot Instances in the specified subnet

The following example includes a subnet. Amazon EC2 launches the instances in the specified subnet. If
the VPC is a nondefault VPC, the instance does not receive a public IPv4 address by default.

{
"ImageId": "ami-1a2b3c4d",
"SecurityGroupIds": [ "sg-1a2b3c4d" ],
"InstanceType": "m3.medium",
"SubnetId": "subnet-1a2b3c4d",
"IamInstanceProfile": {
"Arn": "arn:aws:iam::123456789012:instance-profile/my-iam-role"

399
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

}
}

To assign a public IPv4 address to an instance in a nondefault VPC, specify the

AssociatePublicIpAddress field as shown in the following example. When you specify a network
interface, you must include the subnet ID and security group ID using the network interface, rather than
using the SubnetId and SecurityGroupIds fields shown in example 3.

{
"ImageId": "ami-1a2b3c4d",
"KeyName": "my-key-pair",
"InstanceType": "m3.medium",
"NetworkInterfaces": [
{
"DeviceIndex": 0,
"SubnetId": "subnet-1a2b3c4d",
"Groups": [ "sg-1a2b3c4d" ],
"AssociatePublicIpAddress": true
}
],
"IamInstanceProfile": {
"Arn": "arn:aws:iam::123456789012:instance-profile/my-iam-role"
}
}

Example 4: Launch a Dedicated Spot Instance

The following example requests Spot Instance with a tenancy of dedicated. A Dedicated Spot Instance
must be launched in a VPC.

{
"ImageId": "ami-1a2b3c4d",
"KeyName": "my-key-pair",
"SecurityGroupIds": [ "sg-1a2b3c4d" ],
"InstanceType": "c3.8xlarge",
"SubnetId": "subnet-1a2b3c4d",
"Placement": {
"Tenancy": "dedicated"
}
}

Spot Fleet requests

To use a Spot Fleet, you create a Spot Fleet request that includes the target capacity, an optional On-
Demand portion, one or more launch specifications for the instances, and the maximum price that you
are willing to pay. Amazon EC2 attempts to maintain your Spot Fleet's target capacity as Spot prices
change. For more information, see How Spot Fleet works (p. 371).

There are two types of Spot Fleet requests: request and maintain. You can create a Spot Fleet to
submit a one-time request for your desired capacity, or require it to maintain a target capacity over time.
Both types of requests benefit from Spot Fleet's allocation strategy.

When you make a one-time request, Spot Fleet places the required requests but does not attempt to
replenish Spot Instances if capacity is diminished. If capacity is not available, Spot Fleet does not submit
requests in alternative Spot pools.

To maintain a target capacity, Spot Fleet places requests to meet the target capacity and automatically
replenish any interrupted instances.

400
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

It is not possible to modify the target capacity of a one-time request after it's been submitted. To change
the target capacity, cancel the request and submit a new one.

A Spot Fleet request remains active until it expires or you cancel it. When you cancel a Spot Fleet request,
you may specify whether canceling your Spot Fleet request terminates the Spot Instances in your Spot
Fleet.

Each launch specification includes the information that Amazon EC2 needs to launch an instance, such as
an AMI, instance type, subnet or Availability Zone, and one or more security groups.

Contents
• Spot Fleet request states (p. 401)
• Spot Fleet health checks (p. 402)
• Plan a Spot Fleet request (p. 402)
• Spot Fleet permissions (p. 403)
• Create a Spot Fleet request (p. 407)
• Tag a Spot Fleet (p. 410)
• Monitor your Spot Fleet (p. 416)
• Modify a Spot Fleet request (p. 417)
• Cancel a Spot Fleet request (p. 418)
• Spot Fleet example configurations (p. 419)

Spot Fleet request states

A Spot Fleet request can be in one of the following states:

• submitted – The Spot Fleet request is being evaluated and Amazon EC2 is preparing to launch the
target number of instances.
• active – The Spot Fleet has been validated and Amazon EC2 is attempting to maintain the target
number of running Spot Instances. The request remains in this state until it is modified or canceled.
• modifying – The Spot Fleet request is being modified. The request remains in this state until the
modification is fully processed or the Spot Fleet is canceled. A one-time request cannot be modified,
and this state does not apply to such Spot requests.
• cancelled_running – The Spot Fleet is canceled and does not launch additional Spot Instances. Its
existing Spot Instances continue to run until they are interrupted or terminated. The request remains
in this state until all instances are interrupted or terminated.
• cancelled_terminating – The Spot Fleet is canceled and its Spot Instances are terminating. The
request remains in this state until all instances are terminated.
• cancelled – The Spot Fleet is canceled and has no running Spot Instances. The Spot Fleet request is
deleted two days after its instances were terminated.

The following illustration represents the transitions between the request states. If you exceed your Spot
Fleet limits, the request is canceled immediately.

401
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Spot Fleet health checks

Spot Fleet checks the health status of the Spot Instances in the fleet every two minutes. The health
status of an instance is either healthy or unhealthy. Spot Fleet determines the health status of an
instance using the status checks provided by Amazon EC2. If the status of either the instance status
check or the system status check is impaired for three consecutive health checks, the health status of
the instance is unhealthy. Otherwise, the health status is healthy. For more information, see Status
checks for your instances (p. 743).

You can configure your Spot Fleet to replace unhealthy instances. After enabling health check
replacement, an instance is replaced after its health status is reported as unhealthy. The Spot Fleet
could go below its target capacity for up to a few minutes while an unhealthy instance is being replaced.

Requirements

• Health check replacement is supported only with Spot Fleets that maintain a target capacity, not with
one-time Spot Fleets.
• You can configure your Spot Fleet to replace unhealthy instances only when you create it.
• IAM users can use health check replacement only if they have permission to call the
ec2:DescribeInstanceStatus action.

Plan a Spot Fleet request

Before you create a Spot Fleet request, review Spot Best Practices. Use these best practices when you
plan your Spot Fleet request so that you can provision the type of instances you want at the lowest
possible price. We also recommend that you do the following:

• Determine whether you want to create a Spot Fleet that submits a one-time request for the desired
target capacity, or one that maintains a target capacity over time.
• Determine the instance types that meet your application requirements.
• Determine the target capacity for your Spot Fleet request. You can set the target capacity in instances
or in custom units. For more information, see Spot Fleet instance weighting (p. 376).

402
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

• Determine what portion of the Spot Fleet target capacity must be On-Demand capacity. You can
specify 0 for On-Demand capacity.
• Determine your price per unit, if you are using instance weighting. To calculate the price per unit,
divide the price per instance hour by the number of units (or weight) that this instance represents. If
you are not using instance weighting, the default price per unit is the price per instance hour.
• Review the possible options for your Spot Fleet request. For more information, see the request-spot-
fleet command in the AWS CLI Command Reference. For additional examples, see Spot Fleet example
configurations (p. 419).

Spot Fleet permissions

If your IAM users will create or manage a Spot Fleet, you need to grant them the required permissions.

If you use the Amazon EC2 console to create a Spot Fleet, it creates a service-linked role named
AWSServiceRoleForEC2SpotFleet and a role named aws-ec2-spot-fleet-tagging-role that
grant the Spot Fleet the permissions to request, launch, terminate, and tag resources on your behalf. If
you use the AWS CLI or an API, you must ensure that these roles exist.

Use the following instructions to grant the required permissions and create the roles.

Permissions and roles

• Grant permission to IAM users for Spot Fleet (p. 403)
• Service-linked role for Spot Fleet (p. 405)
• IAM role for Spot Fleet (p. 406)

Grant permission to IAM users for Spot Fleet

If your IAM users will create or manage a Spot Fleet, be sure to grant them the required permissions as
follows.

To grant an IAM user permissions for Spot Fleet

1. Open the IAM console at https://console.aws.amazon.com/iam/.

2. In the navigation pane, choose Policies, Create policy.
3. On the Create policy page, choose JSON, and replace the text with the following.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:RunInstances",
"ec2:CreateTags",
"ec2:RequestSpotFleet",
"ec2:ModifySpotFleetRequest",
"ec2:CancelSpotFleetRequests",
"ec2:DescribeSpotFleetRequests",
"ec2:DescribeSpotFleetInstances",
"ec2:DescribeSpotFleetRequestHistory"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": "iam:PassRole",
"Resource": "arn:aws:iam::*:role/aws-ec2-spot-fleet-tagging-role"
},

403
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

{
"Effect": "Allow",
"Action": [
"iam:CreateServiceLinkedRole",
"iam:ListRoles",
"iam:ListInstanceProfiles"
],
"Resource": "*"
}
]
}

The preceding example policy grants an IAM user the permissions required for most Spot Fleet use
cases. To limit the user to specific API actions, specify only those API actions instead.

Required EC2 and IAM APIs

The following APIs must be included in the policy:

• ec2:RunInstances – Required to launch instances in a Spot Fleet

• ec2:CreateTags – Required to tag the Spot Fleet request, instances, or volumes
• iam:PassRole – Required to specify the Spot Fleet role
• iam:CreateServiceLinkedRole – Required to create the service-linked role
• iam:ListRoles – Required to enumerate existing IAM roles
• iam:ListInstanceProfiles – Required to enumerate existing instance profiles

Important
If you specify a role for the IAM instance profile in the launch specification or launch
template, you must grant the IAM user the permission to pass the role to the service. To do
this, in the IAM policy include "arn:aws:iam::*:role/IamInstanceProfile-role"
as a resource for the iam:PassRole action. For more information, see Granting a User
Permissions to Pass a Role to an AWS Service in the IAM User Guide.

Spot Fleet APIs

Add the following Spot Fleet API actions to your policy, as needed:

• ec2:RequestSpotFleet
• ec2:ModifySpotFleetRequest
• ec2:CancelSpotFleetRequests
• ec2:DescribeSpotFleetRequests
• ec2:DescribeSpotFleetInstances
• ec2:DescribeSpotFleetRequestHistory

Optional IAM APIs

(Optional) To enable an IAM user to create roles or instance profiles using the IAM console, you must
add the following actions to the policy:

• iam:AddRoleToInstanceProfile
• iam:AttachRolePolicy
• iam:CreateInstanceProfile
• iam:CreateRole
• iam:GetRole
404
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

• iam:ListPolicies
4. Choose Review policy.
5. On the Review policy page, enter a policy name and description, and choose Create policy.
6. In the navigation pane, choose Users and select the user.
7. Choose Permissions, Add permissions.
8. Choose Attach existing policies directly. Select the policy that you created earlier and choose Next:
Review.
9. Choose Add permissions.

Service-linked role for Spot Fleet

Amazon EC2 uses service-linked roles for the permissions that it requires to call other AWS services on
your behalf. A service-linked role is a unique type of IAM role that is linked directly to an AWS service.
Service-linked roles provide a secure way to delegate permissions to AWS services because only the
linked service can assume a service-linked role. For more information, see Using Service-Linked Roles in
the IAM User Guide.

Amazon EC2 uses the service-linked role named AWSServiceRoleForEC2SpotFleet to launch and
manage instances on your behalf.
Important
If you specify an encrypted AMI (p. 161) or an encrypted Amazon EBS snapshot (p. 1176) in your
Spot Fleet, you must grant the AWSServiceRoleForEC2SpotFleet role permission to use the
CMK so that Amazon EC2 can launch instances on your behalf. For more information, see Grant
access to CMKs for use with encrypted AMIs and EBS snapshots (p. 406).

Permissions granted by AWSServiceRoleForEC2SpotFleet

Amazon EC2 uses AWSServiceRoleForEC2SpotFleet to complete the following actions:

• ec2:RequestSpotInstances - Request Spot Instances

• ec2:RunInstances - Launch instances
• ec2:TerminateInstances - Terminate instances
• ec2:DescribeImages - Describe Amazon Machine Images (AMIs) for the instances
• ec2:DescribeInstanceStatus - Describe the status of the instances
• ec2:DescribeSubnets - Describe the subnets for the instances
• ec2:CreateTags - Add tags to the Spot Fleet request, instances, and volumes
• elasticloadbalancing:RegisterInstancesWithLoadBalancer - Add the specified instances
to the specified load balancer
• elasticloadbalancing:RegisterTargets - Register the specified targets with the specified
target group

Create the service-linked role

Under most circumstances, you don't need to manually create a service-linked role. Amazon EC2 creates
the AWSServiceRoleForEC2SpotFleet service-linked role the first time you create a Spot Fleet using the
console.

If you use the AWS CLI or an API, you must ensure that this role exists.

If you had an active Spot Fleet request before October 2017, when Amazon EC2 began supporting this
service-linked role, Amazon EC2 created the AWSServiceRoleForEC2SpotFleet role in your AWS account.
For more information, see A New Role Appeared in My AWS Account in the IAM User Guide.

Ensure that this role exists before you use the AWS CLI or an API to create a Spot Fleet.

405
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

To create AWSServiceRoleForEC2SpotFleet using the console

1. Open the IAM console at https://console.aws.amazon.com/iam/.

2. In the navigation pane, choose Roles.
3. Choose Create role.
4. For Select type of trusted entity, choose AWS service.
5. In the list of services, choose EC2.
6. In the Select your use case section, choose EC2 - Spot Fleet
7. Choose Next: Permissions.
8. On the next page, choose Next:Review.
9. On the Review page, choose Create role.

To create AWSServiceRoleForEC2SpotFleet using the AWS CLI

Use the create-service-linked-role command as follows.

aws iam create-service-linked-role --aws-service-name spotfleet.amazonaws.com

If you no longer need to use Spot Fleet, we recommend that you delete the
AWSServiceRoleForEC2SpotFleet role. After this role is deleted from your account, Amazon EC2 will
create the role again if you request a Spot Fleet using the console. For more information, see Deleting a
Service-Linked Role in the IAM User Guide.

Grant access to CMKs for use with encrypted AMIs and EBS snapshots
If you specify an encrypted AMI (p. 161) or an encrypted Amazon EBS snapshot (p. 1176) in your Spot
Fleet request and you use a customer managed customer master key (CMK) for encryption, you must
grant the AWSServiceRoleForEC2SpotFleet role permission to use the CMK so that Amazon EC2 can
launch instances on your behalf. To do this, you must add a grant to the CMK, as shown in the following
procedure.

When providing permissions, grants are an alternative to key policies. For more information, see Using
Grants and Using Key Policies in AWS KMS in the AWS Key Management Service Developer Guide.

To grant the AWSServiceRoleForEC2SpotFleet role permissions to use the CMK

• Use the create-grant command to add a grant to the CMK and to specify the principal (the
AWSServiceRoleForEC2SpotFleet service-linked role) that is given permission to perform the
operations that the grant permits. The CMK is specified by the key-id parameter and the ARN of
the CMK. The principal is specified by the grantee-principal parameter and the ARN of the
AWSServiceRoleForEC2SpotFleet service-linked role.

aws kms create-grant \

--region us-east-1 \
--key-id arn:aws:kms:us-
east-1:444455556666:key/1234abcd-12ab-34cd-56ef-1234567890ab \
--grantee-principal arn:aws:iam::111122223333:role/AWSServiceRoleForEC2SpotFleet \
--operations "Decrypt" "Encrypt" "GenerateDataKey"
"GenerateDataKeyWithoutPlaintext" "CreateGrant" "DescribeKey" "ReEncryptFrom"
"ReEncryptTo"

IAM role for Spot Fleet

The aws-ec2-spot-fleet-tagging-role IAM role grants the Spot Fleet permission to tag the Spot
Fleet request, instances, and volumes. For more information, see Tag a Spot Fleet (p. 410).

406
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Important
If you choose to tag instances in the fleet and you choose to maintain target capacity (the
Spot Fleet request is of type maintain), the differences in permissions of the IAM user and
the IamFleetRole might lead to inconsistent tagging behavior of instances in the fleet. If the
IamFleetRole does not include the CreateTags permission, some of the instances launched
by the fleet might not be tagged. While we are working to fix this inconsistency, to ensure that
all instances launched by the fleet are tagged, we recommend that you use the aws-ec2-spot-
fleet-tagging-role role for the IamFleetRole. Alternatively, to use an existing role, attach
the AmazonEC2SpotFleetTaggingRole AWS Managed Policy to the existing role. Otherwise,
you need to manually add the CreateTags permission to your existing policy.

To create the IAM role for tagging a Spot Fleet

1. Open the IAM console at https://console.aws.amazon.com/iam/.

2. In the navigation pane, choose Roles.
3. On the Select type of trusted entity page, choose AWS service, EC2, EC2 - Spot Fleet Tagging,
Next: Permissions.
4. On the Attached permissions policy page, choose Next:Review.
5. On the Review page, enter a name for the role (for example, aws-ec2-spot-fleet-tagging-
role) and choose Create role.

Create a Spot Fleet request

Using the AWS Management Console, quickly create a Spot Fleet request by choosing only your
application or task need and minimum compute specs. Amazon EC2 configures a fleet that best meets
your needs and follows Spot best practice. For more information, see Quickly create a Spot Fleet request
(console) (p. 407). Otherwise, you can modify any of the default settings. For more information, see
Create a Spot Fleet request using defined parameters (console) (p. 407).

Quickly create a Spot Fleet request (console)

Follow these steps to quickly create a Spot Fleet request.

To create a Spot Fleet request using the recommended settings (console)

1. Open the Spot console at https://console.aws.amazon.com/ec2spot.

2. If you are new to Spot, you see a welcome page; choose Get started. Otherwise, choose Request
Spot Instances.
3. For Tell us your application or task need, choose Load balancing workloads, Flexible workloads,
Big data workloads, or Defined duration workloads.
4. Under Configure your instances, for Minimum compute unit, choose the minimum hardware
specifications (vCPUs, memory, and storage) that you need for your application or task, either as
specs or as an instance type.

• For as specs, specify the required number of vCPUs and amount of memory.
• For as an instance type, accept the default instance type, or choose Change instance type to
choose a different instance type.
5. Under Tell us how much capacity you need, for Total target capacity, specify the number of units
to request for target capacity. You can choose instances or vCPUs.
6. Review the recommended Fleet request settings based on your application or task selection, and
choose Launch.

Create a Spot Fleet request using defined parameters (console)

You can create a Spot Fleet using the parameters that you define.

407
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

To create a Spot Fleet request using defined parameters (console)

1. Open the Spot console at https://console.aws.amazon.com/ec2spot.

2. If you are new to Spot, you see a welcome page; choose Get started. Otherwise, choose Request
Spot Instances.
3. For Tell us your application or task need, choose Load balancing workloads, Flexible workloads,
Big data workloads, or Defined duration workloads.
4. For Configure your instances, do the following:

a. (Optional) For Launch template, choose a launch template. The launch template must specify
an Amazon Machine Image (AMI), as you cannot override the AMI using Spot Fleet if you specify
a launch template.
Important
If you intend to specify Optional On-Demand portion, you must choose a launch
template.
b. For AMI, choose one of the basic AMIs provided by AWS, or choose Search for AMI to use an
AMI from our user community, the AWS Marketplace, or one of your own.
c. For Minimum compute unit, choose the minimum hardware specifications (vCPUs, memory,
and storage) that you need for your application or task, either as specs or as an instance type.

• For as specs, specify the required number of vCPUs and amount of memory.
• For as an instance type, accept the default instance type, or choose Change instance type to
choose a different instance type.
d. For Network, choose an existing VPC or create a new one.

[Existing VPC] Choose the VPC.

[New VPC] Choose Create new VPC to go the Amazon VPC console. When you are done, return
to the wizard and refresh the list.
e. (Optional) For Availability Zone, let AWS choose the Availability Zones for your Spot Instances,
or specify one or more Availability Zones.

If you have more than one subnet in an Availability Zone, choose the appropriate subnet from
Subnet. To add subnets, choose Create new subnet to go to the Amazon VPC console. When
you are done, return to the wizard and refresh the list.
f. (Optional) For Key pair name, choose an existing key pair or create a new one.

[Existing key pair] Choose the key pair.

[New key pair] Choose Create new key pair to go the Amazon VPC console. When you are done,
return to the wizard and refresh the list.
5. (Optional) For Additional configurations, do the following:

a. (Optional) To add storage, specify additional instance store volumes or Amazon EBS volumes,
depending on the instance type.
b. (Optional) To enable Amazon EBS optimization, for EBS-optimized, choose Launch EBS-
optimized instances.
c. (Optional) To add temporary block-level storage for your instances, for Instance store, choose
Attach at launch.
d. (Optional) By default, basic monitoring is enabled for your instances. To enable detailed
monitoring, for Monitoring, choose Enable CloudWatch detailed monitoring.
e. (Optional) To replace unhealthy instances, for Health check, choose Replace unhealthy
instances. To enable this option, you must first choose Maintain target capacity.

408
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

f. (Optional) To run a Dedicated Spot Instance, for Tenancy, choose Dedicated - run a dedicated
instance.
g. (Optional) For Security groups, choose one or more security groups or create a new one.

[Existing security group] Choose one or more security groups.

[New security group] Choose Create new security group to go the Amazon VPC console. When
you are done, return to the wizard and refresh the list.
h. (Optional) To make your instances reachable from the internet, for Auto-assign IPv4 Public IP,
choose Enable.
i. (Optional) To launch your Spot Instances with an IAM role, for IAM instance profile, choose the
role.
j. (Optional) To run a start-up script, copy it to User data.
k. (Optional) To add a tag, choose Add new tag and enter the key and value for the tag. Repeat for
each tag.

For each tag, to tag the instances and the Spot Fleet request with the same tag, ensure that
both Instance tags and Fleet tags are selected. To tag only the instances launched by the fleet,
clear Fleet tags. To tag only the Spot Fleet request, clear Instance tags.
6. For Tell us how much capacity you need, do the following:

a. For Total target capacity, specify the number of units to request for target capacity. You can
choose instances or vCPUs. To specify a target capacity of 0 so that you can add capacity later,
choose Maintain target capacity.
b. (Optional) For Optional On-Demand portion, specify the number of On-Demand units to
request. The number must be less than the Total target capacity. Amazon EC2 calculates the
difference, and allocates the difference to Spot units to request.
Important
To specify an optional On-Demand portion, you must first choose a launch template.
c. (Optional) By default, the Spot service terminates Spot Instances when they are interrupted.
To maintain the target capacity, select Maintain target capacity. You can then specify that the
Spot service terminates, stops, or hibernates Spot Instances when they are interrupted. To do
so, choose the corresponding option from Interruption behavior.
d. (Optional) To allow Spot Fleet to launch a replacement Spot Instance when an instance
rebalance notification is emitted for an existing Spot Instance in the fleet, select Capacity
rebalance. For more information, see Capacity Rebalancing (p. 374).
Note
When a replacement instance is launched, the instance marked for rebalance is not
automatically terminated. You can terminate it, or you can leave it running. You are
charged for both instances while they are running.
The instance marked for rebalance is at an elevated risk of interruption, and you will
receive a two-minute Spot Instance interruption notice before Amazon EC2 interrupts
it.
e. (Optional) To control the amount you pay per hour for all the Spot Instances in your fleet,
select Maintain target cost for Spot (advanced - optional) and then enter the maximum total
amount you're willing to pay per hour. When the maximum total amount is reached, Spot Fleet
stops launching Spot Instances even if it hasn’t met the target capacity. For more information,
see Control spending (p. 376).
7. For Fleet request settings, do the following:

a. Review the fleet request and fleet allocation strategy based on your application or task
selection. To change the instance types or allocation strategy, clear Apply recommendations.

409
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

b. (Optional) To remove instance types, for Fleet request, choose Remove. To add instance types,
choose Select instance types.
c. (Optional) For Fleet allocation strategy, choose the strategy that meets your needs. For more
information, see Allocation strategy for Spot Instances (p. 372).
8. For Additional request details, do the following:

a. Review the additional request details. To make changes, clear Apply defaults.
b. (Optional) For IAM fleet role, you can use the default role or choose a different role. To use the
default role after changing the role, choose Use default role.
c. (Optional) For Maximum price, you can use the default maximum price (the On-Demand price)
or specify the maximum price you are willing to pay. If your maximum price is lower than the
Spot price for the instance types that you selected, your Spot Instances are not launched.
d. (Optional) To create a request that is valid only during a specific time period, edit Request valid
from and Request valid until.
e. (Optional) By default, we terminate your Spot Instances when the request expires. To keep them
running after your request expires, clear Terminate the instances when the request expires.
f. (Optional) To register your Spot Instances with a load balancer, choose Receive traffic from one
or more load balancers and choose one or more Classic Load Balancers or target groups.
9. (Optional) To download a copy of the launch configuration for use with the AWS CLI, choose JSON
config.
10. Choose Launch.

The Spot Fleet request type is fleet. When the request is fulfilled, requests of type instance are
added, where the state is active and the status is fulfilled.

To create a Spot Fleet request using the AWS CLI

• Use the request-spot-fleet command to create a Spot Fleet request.

aws ec2 request-spot-fleet --spot-fleet-request-config file://config.json

For example configuration files, see Spot Fleet example configurations (p. 419).

The following is example output:

{
"SpotFleetRequestId": "sfr-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE"
}

Tag a Spot Fleet

To help categorize and manage your Spot Fleet requests, you can tag them with custom metadata. You
can assign a tag to a Spot Fleet request when you create it, or afterward. You can assign tags using the
Amazon EC2 console or a command line tool.

When you tag a Spot Fleet request, the instances and volumes that are launched by the Spot Fleet are
not automatically tagged. You need to explicitly tag the instances and volumes launched by the Spot
Fleet. You can choose to assign tags to only the Spot Fleet request, or to only the instances launched by
the fleet, or to only the volumes attached to the instances launched by the fleet, or to all three.
Note
Volume tags are only supported for volumes that are attached to On-Demand Instances. You
can't tag volumes that are attached to Spot Instances.

410
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

For more information about how tags work, see Tag your Amazon EC2 resources (p. 1311).

Contents
• Prerequisite (p. 411)
• Tag a new Spot Fleet (p. 411)
• Tag a new Spot Fleet and the instances and volumes that it launches (p. 412)
• Tag an existing Spot Fleet (p. 415)
• View Spot Fleet request tags (p. 415)

Prerequisite

Grant the IAM user the permission to tag resources. For more information, see Example: Tag
resources (p. 1016).

To grant an IAM user the permission to tag resources

Create a IAM policy that includes the following:

• The ec2:CreateTags action. This grants the IAM user permission to create tags.
• The ec2:RequestSpotFleet action. This grants the IAM user permission to create a Spot Fleet
request.
• For Resource, you must specify "*". This allows users to tag all resource types.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "TagSpotFleetRequest",
"Effect": "Allow",
"Action": [
"ec2:CreateTags",
"ec2:RequestSpotFleet"
],
"Resource": "*"
}

Important
We currently do not support resource-level permissions for the spot-fleet-request
resource. If you specify spot-fleet-request as a resource, you will get an unauthorized
exception when you try to tag the fleet. The following example illustrates how not to set the
policy.

{
"Effect": "Allow",
"Action": [
"ec2:CreateTags",
"ec2:RequestSpotFleet"
],
"Resource": "arn:aws:ec2:us-east-1:111122223333:spot-fleet-request/*"
}

Tag a new Spot Fleet

To tag a new Spot Fleet request using the console

1. Follow the Create a Spot Fleet request using defined parameters (console) (p. 407) procedure.

411
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

2. To add a tag, expand Additional configurations, choose Add new tag, and enter the key and value
for the tag. Repeat for each tag.

For each tag, you can tag the Spot Fleet request and the instances with the same tag. To tag both,
ensure that both Instance tags and Fleet tags are selected. To tag only the Spot Fleet request, clear
Instance tags. To tag only the instances launched by the fleet, clear Fleet tags.
3. Complete the required fields to create a Spot Fleet request, and then choose Launch. For more
information, see Create a Spot Fleet request using defined parameters (console) (p. 407).

To tag a new Spot Fleet request using the AWS CLI

To tag a Spot Fleet request when you create it, configure the Spot Fleet request configuration as follows:

• Specify the tags for the Spot Fleet request in SpotFleetRequestConfig.

• For ResourceType, specify spot-fleet-request. If you specify another value, the fleet request will
fail.
• For Tags, specify the key-value pair. You can specify more than one key-value pair.

In the following example, the Spot Fleet request is tagged with two tags: Key=Environment and
Value=Production, and Key=Cost-Center and Value=123.

{
"SpotFleetRequestConfig": {
"AllocationStrategy": "lowestPrice",
"ExcessCapacityTerminationPolicy": "default",
"IamFleetRole": "arn:aws:iam::111122223333:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-0123456789EXAMPLE",
"InstanceType": "c4.large"
}
],
"SpotPrice": "5",
"TargetCapacity": 2,
"TerminateInstancesWithExpiration": true,
"Type": "maintain",
"ReplaceUnhealthyInstances": true,
"InstanceInterruptionBehavior": "terminate",
"InstancePoolsToUseCount": 1,
"TagSpecifications": [
{
"ResourceType": "spot-fleet-request",
"Tags": [
{
"Key": "Environment",
"Value":"Production"
},
{
"Key": "Cost-Center",
"Value":"123"
}
]
}
]
}
}

Tag a new Spot Fleet and the instances and volumes that it launches
To tag a new Spot Fleet request and the instances and volumes that it launches using the AWS CLI

412
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

To tag a Spot Fleet request when you create it, and to tag the instances and volumes when they are
launched by the fleet, configure the Spot Fleet request configuration as follows:

Spot Fleet request tags:

• Specify the tags for the Spot Fleet request in SpotFleetRequestConfig.

• For ResourceType, specify spot-fleet-request. If you specify another value, the fleet request will
fail.
• For Tags, specify the key-value pair. You can specify more than one key-value pair.

Instance tags:

• Specify the tags for the instances in LaunchSpecifications.

• For ResourceType, specify instance. If you specify another value, the fleet request will fail.
• For Tags, specify the key-value pair. You can specify more than one key-value pair.

Alternatively, you can specify the tags for the instance in the launch template (p. 534) that is
referenced in the Spot Fleet request.

Volume tags:

• Specify the tags for the volumes in the launch template (p. 534) that is referenced in the Spot Fleet
request. Volume tagging in LaunchSpecifications is not supported.

In the following example, the Spot Fleet request is tagged with two tags: Key=Environment and
Value=Production, and Key=Cost-Center and Value=123. The instances that are launched by the fleet are
tagged with one tag (which is the same as one of the tags for the Spot Fleet request): Key=Cost-Center
and Value=123.

{
"SpotFleetRequestConfig": {
"AllocationStrategy": "lowestPrice",
"ExcessCapacityTerminationPolicy": "default",
"IamFleetRole": "arn:aws:iam::111122223333:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-0123456789EXAMPLE",
"InstanceType": "c4.large",
"TagSpecifications": [
{
"ResourceType": "instance",
"Tags": [
{
"Key": "Cost-Center",
"Value": "123"
}
]
}
]
}
],
"SpotPrice": "5",
"TargetCapacity": 2,
"TerminateInstancesWithExpiration": true,
"Type": "maintain",
"ReplaceUnhealthyInstances": true,
"InstanceInterruptionBehavior": "terminate",
"InstancePoolsToUseCount": 1,

413
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

"TagSpecifications": [
{
"ResourceType": "spot-fleet-request",
"Tags": [
{
"Key": "Environment",
"Value":"Production"
},
{
"Key": "Cost-Center",
"Value":"123"
}
]
}
]
}
}

To tag instances launched by a Spot Fleet using the AWS CLI

To tag instances when they are launched by the fleet, you can either specify the tags in the launch
template (p. 534) that is referenced in the Spot Fleet request, or you can specify the tags in the Spot
Fleet request configuration as follows:

• Specify the tags for the instances in LaunchSpecifications.

• For ResourceType, specify instance. If you specify another value, the fleet request will fail.
• For Tags, specify the key-value pair. You can specify more than one key-value pair.

In the following example, the instances that are launched by the fleet are tagged with one tag:
Key=Cost-Center and Value=123.

{
"SpotFleetRequestConfig": {
"AllocationStrategy": "lowestPrice",
"ExcessCapacityTerminationPolicy": "default",
"IamFleetRole": "arn:aws:iam::111122223333:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-0123456789EXAMPLE",
"InstanceType": "c4.large",
"TagSpecifications": [
{
"ResourceType": "instance",
"Tags": [
{
"Key": "Cost-Center",
"Value": "123"
}
]
}
]
}
],
"SpotPrice": "5",
"TargetCapacity": 2,
"TerminateInstancesWithExpiration": true,
"Type": "maintain",
"ReplaceUnhealthyInstances": true,
"InstanceInterruptionBehavior": "terminate",
"InstancePoolsToUseCount": 1
}

414
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

To tag volumes attached to On-Demand Instances launched by a Spot Fleet using the AWS CLI

To tag volumes when they are created by the fleet, you must specify the tags in the launch
template (p. 534) that is referenced in the Spot Fleet request.
Note
Volume tags are only supported for volumes that are attached to On-Demand Instances. You
can't tag volumes that are attached to Spot Instances.
Volume tagging in LaunchSpecifications is not supported.

Tag an existing Spot Fleet

To tag an existing Spot Fleet request using the console

After you have created a Spot Fleet request, you can add tags to the fleet request using the console.

1. Open the Spot console at https://console.aws.amazon.com/ec2spot.

2. Select your Spot Fleet request.
3. Choose the Tags tab and choose Create Tag.

To tag an existing Spot Fleet request using the AWS CLI

You can use the create-tags command to tag existing resources. In the following example, the existing
Spot Fleet request is tagged with Key=purpose and Value=test.

aws ec2 create-tags \

--resources sfr-11112222-3333-4444-5555-66666EXAMPLE \
--tags Key=purpose,Value=test

View Spot Fleet request tags

To view Spot Fleet request tags using the console

1. Open the Spot console at https://console.aws.amazon.com/ec2spot.

2. Select your Spot Fleet request and choose the Tags tab.

To describe Spot Fleet request tags

Use the describe-tags command to view the tags for the specified resource. In the following example,
you describe the tags for the specified Spot Fleet request.

aws ec2 describe-tags \

--filters "Name=resource-id,Values=sfr-11112222-3333-4444-5555-66666EXAMPLE"

{
"Tags": [
{
"Key": "Environment",
"ResourceId": "sfr-11112222-3333-4444-5555-66666EXAMPLE",
"ResourceType": "spot-fleet-request",
"Value": "Production"
},
{
"Key": "Another key",
"ResourceId": "sfr-11112222-3333-4444-5555-66666EXAMPLE",
"ResourceType": "spot-fleet-request",

415
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

"Value": "Another value"

}
]
}

You can also view the tags of a Spot Fleet request by describing the Spot Fleet request.

Use the describe-spot-fleet-requests command to view the configuration of the specified Spot Fleet
request, which includes any tags that were specified for the fleet request.

aws ec2 describe-spot-fleet-requests \

--spot-fleet-request-ids sfr-11112222-3333-4444-5555-66666EXAMPLE

{
"SpotFleetRequestConfigs": [
{
"ActivityStatus": "fulfilled",
"CreateTime": "2020-02-13T02:49:19.709Z",
"SpotFleetRequestConfig": {
"AllocationStrategy": "capacityOptimized",
"OnDemandAllocationStrategy": "lowestPrice",
"ExcessCapacityTerminationPolicy": "Default",
"FulfilledCapacity": 2.0,
"OnDemandFulfilledCapacity": 0.0,
"IamFleetRole": "arn:aws:iam::111122223333:role/aws-ec2-spot-fleet-tagging-
role",
"LaunchSpecifications": [
{
"ImageId": "ami-0123456789EXAMPLE",
"InstanceType": "c4.large"
}
],
"TargetCapacity": 2,
"OnDemandTargetCapacity": 0,
"Type": "maintain",
"ReplaceUnhealthyInstances": false,
"InstanceInterruptionBehavior": "terminate"
},
"SpotFleetRequestId": "sfr-11112222-3333-4444-5555-66666EXAMPLE",
"SpotFleetRequestState": "active",
"Tags": [
{
"Key": "Environment",
"Value": "Production"
},
{
"Key": "Another key",
"Value": "Another value"
}
]
}
]
}

Monitor your Spot Fleet

The Spot Fleet launches Spot Instances when your maximum price exceeds the Spot price and capacity is
available. The Spot Instances run until they are interrupted or you terminate them.

To monitor your Spot Fleet (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

416
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

2. In the navigation pane, choose Spot Requests.

3. Select your Spot Fleet request. To see the configuration details, choose Description.
4. To list the Spot Instances for the Spot Fleet, choose Instances.
5. To view the history for the Spot Fleet, choose History.

To monitor your Spot Fleet (AWS CLI)

Use the describe-spot-fleet-requests command to describe your Spot Fleet requests.

aws ec2 describe-spot-fleet-requests

Use the describe-spot-fleet-instances command to describe the Spot Instances for the specified Spot
Fleet.

aws ec2 describe-spot-fleet-instances \

--spot-fleet-request-id sfr-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE

Use the describe-spot-fleet-request-history command to describe the history for the specified Spot Fleet
request.

aws ec2 describe-spot-fleet-request-history \

--spot-fleet-request-id sfr-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE \
--start-time 2015-05-18T00:00:00Z

Modify a Spot Fleet request

You can modify an active Spot Fleet request to complete the following tasks:

• Increase the target capacity and On-Demand portion

• Decrease the target capacity and On-Demand portion

Note
You can't modify a one-time Spot Fleet request. You can only modify a Spot Fleet request if you
selected Maintain target capacity when you created the Spot Fleet request.

When you increase the target capacity, the Spot Fleet launches additional Spot Instances. When you
increase the On-Demand portion, the Spot Fleet launches additional On-Demand Instances.

When you increase the target capacity, the Spot Fleet launches the additional Spot Instances according
to the allocation strategy for its Spot Fleet request. If the allocation strategy is lowestPrice, the Spot
Fleet launches the instances from the lowest-priced Spot Instance pool in the Spot Fleet request. If the
allocation strategy is diversified, the Spot Fleet distributes the instances across the pools in the Spot
Fleet request.

When you decrease the target capacity, the Spot Fleet cancels any open requests that exceed the new
target capacity. You can request that the Spot Fleet terminate Spot Instances until the size of the fleet
reaches the new target capacity. If the allocation strategy is lowestPrice, the Spot Fleet terminates
the instances with the highest price per unit. If the allocation strategy is diversified, the Spot Fleet
terminates instances across the pools. Alternatively, you can request that the Spot Fleet keep the fleet at
its current size, but not replace any Spot Instances that are interrupted or that you terminate manually.

When a Spot Fleet terminates an instance because the target capacity was decreased, the instance
receives a Spot Instance interruption notice.

417
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

To modify a Spot Fleet request (console)

1. Open the Spot console at https://console.aws.amazon.com/ec2spot/home/fleet.

2. Select your Spot Fleet request.
3. Choose Actions, Modify target capacity.
4. In Modify target capacity, do the following:

a. Enter the new target capacity and On-Demand portion.

b. (Optional) If you are decreasing the target capacity but want to keep the fleet at its current size,
clear Terminate instances.
c. Choose Submit.

To modify a Spot Fleet request using the AWS CLI

Use the modify-spot-fleet-request command to update the target capacity of the specified Spot Fleet
request.

aws ec2 modify-spot-fleet-request \

--spot-fleet-request-id sfr-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE \
--target-capacity 20

You can modify the previous command as follows to decrease the target capacity of the specified Spot
Fleet without terminating any Spot Instances as a result.

aws ec2 modify-spot-fleet-request \

--spot-fleet-request-id sfr-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE \
--target-capacity 10 \
--excess-capacity-termination-policy NoTermination

Cancel a Spot Fleet request

When you are finished using your Spot Fleet, you can cancel the Spot Fleet request. This cancels all
Spot requests associated with the Spot Fleet, so that no new Spot Instances are launched for your Spot
Fleet. You must specify whether the Spot Fleet should terminate its Spot Instances. If you terminate
the instances, the Spot Fleet request enters the cancelled_terminating state. Otherwise, the Spot
Fleet request enters the cancelled_running state and the instances continue to run until they are
interrupted or you terminate them manually.

To cancel a Spot Fleet request (console)

1. Open the Spot console at https://console.aws.amazon.com/ec2spot/home/fleet.

2. Select your Spot Fleet request.
3. Choose Actions, Cancel spot request.
4. In Cancel spot request, verify that you want to cancel the Spot Fleet. To keep the fleet at its current
size, clear Terminate instances. When you are ready, choose Confirm.

To cancel a Spot Fleet request using the AWS CLI

Use the cancel-spot-fleet-requests command to cancel the specified Spot Fleet request and terminate
the instances.

aws ec2 cancel-spot-fleet-requests \

--spot-fleet-request-ids sfr-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE \

418
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

--terminate-instances

The following is example output:

{
"SuccessfulFleetRequests": [
{
"SpotFleetRequestId": "sfr-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE",
"CurrentSpotFleetRequestState": "cancelled_terminating",
"PreviousSpotFleetRequestState": "active"
}
],
"UnsuccessfulFleetRequests": []
}

You can modify the previous command as follows to cancel the specified Spot Fleet request without
terminating the instances.

aws ec2 cancel-spot-fleet-requests \

--spot-fleet-request-ids sfr-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE \
--no-terminate-instances

The following is example output:

{
"SuccessfulFleetRequests": [
{
"SpotFleetRequestId": "sfr-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE",
"CurrentSpotFleetRequestState": "cancelled_running",
"PreviousSpotFleetRequestState": "active"
}
],
"UnsuccessfulFleetRequests": []
}

Spot Fleet example configurations

The following examples show launch configurations that you can use with the request-spot-
fleet command to create a Spot Fleet request. For more information, see Create a Spot Fleet
request (p. 407).
Note
For Spot Fleet, you can't specify an network interface ID in a launch specification. Make sure you
omit the NetworkInterfaceID parameter in your launch specification.

Examples
• Example 1: Launch Spot Instances using the lowest-priced Availability Zone or subnet in the
Region (p. 420)
• Example 2: Launch Spot Instances using the lowest-priced Availability Zone or subnet in a specified
list (p. 420)
• Example 3: Launch Spot Instances using the lowest-priced instance type in a specified list (p. 421)
• Example 4. Override the price for the request (p. 422)
• Example 5: Launch a Spot Fleet using the diversified allocation strategy (p. 424)
• Example 6: Launch a Spot Fleet using instance weighting (p. 426)
• Example 7: Launch a Spot Fleet with On-Demand capacity (p. 427)
• Example 8: Configure Capacity Rebalancing to launch replacement Spot Instances (p. 427)

419
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Example 1: Launch Spot Instances using the lowest-priced Availability Zone or subnet in the
Region
The following example specifies a single launch specification without an Availability Zone or subnet. The
Spot Fleet launches the instances in the lowest-priced Availability Zone that has a default subnet. The
price you pay does not exceed the On-Demand price.

{
"TargetCapacity": 20,
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"KeyName": "my-key-pair",
"SecurityGroups": [
{
"GroupId": "sg-1a2b3c4d"
}
],
"InstanceType": "m3.medium",
"IamInstanceProfile": {
"Arn": "arn:aws:iam::123456789012:instance-profile/my-iam-role"
}
}
]
}

Example 2: Launch Spot Instances using the lowest-priced Availability Zone or subnet in a
specified list
The following examples specify two launch specifications with different Availability Zones or subnets,
but the same instance type and AMI.

Availability Zones

The Spot Fleet launches the instances in the default subnet of the lowest-priced Availability Zone that
you specified.

{
"TargetCapacity": 20,
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"KeyName": "my-key-pair",
"SecurityGroups": [
{
"GroupId": "sg-1a2b3c4d"
}
],
"InstanceType": "m3.medium",
"Placement": {
"AvailabilityZone": "us-west-2a, us-west-2b"
},
"IamInstanceProfile": {
"Arn": "arn:aws:iam::123456789012:instance-profile/my-iam-role"
}
}
]
}

Subnets

420
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

You can specify default subnets or nondefault subnets, and the nondefault subnets can be from a
default VPC or a nondefault VPC. The Spot service launches the instances in whichever subnet is in the
lowest-priced Availability Zone.

You can't specify different subnets from the same Availability Zone in a Spot Fleet request.

{
"TargetCapacity": 20,
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"KeyName": "my-key-pair",
"SecurityGroups": [
{
"GroupId": "sg-1a2b3c4d"
}
],
"InstanceType": "m3.medium",
"SubnetId": "subnet-a61dafcf, subnet-65ea5f08",
"IamInstanceProfile": {
"Arn": "arn:aws:iam::123456789012:instance-profile/my-iam-role"
}
}
]
}

If the instances are launched in a default VPC, they receive a public IPv4 address by default. If the
instances are launched in a nondefault VPC, they do not receive a public IPv4 address by default. Use
a network interface in the launch specification to assign a public IPv4 address to instances launched in
a nondefault VPC. When you specify a network interface, you must include the subnet ID and security
group ID using the network interface.

...
{
"ImageId": "ami-1a2b3c4d",
"KeyName": "my-key-pair",
"InstanceType": "m3.medium",
"NetworkInterfaces": [
{
"DeviceIndex": 0,
"SubnetId": "subnet-1a2b3c4d",
"Groups": [ "sg-1a2b3c4d" ],
"AssociatePublicIpAddress": true
}
],
"IamInstanceProfile": {
"Arn": "arn:aws:iam::880185128111:instance-profile/my-iam-role"
}
}
...

Example 3: Launch Spot Instances using the lowest-priced instance type in a specified list
The following examples specify two launch configurations with different instance types, but the same
AMI and Availability Zone or subnet. The Spot Fleet launches the instances using the specified instance
type with the lowest price.

Availability Zone

{
"TargetCapacity": 20,

421
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"SecurityGroups": [
{
"GroupId": "sg-1a2b3c4d"
}
],
"InstanceType": "cc2.8xlarge",
"Placement": {
"AvailabilityZone": "us-west-2b"
}
},
{
"ImageId": "ami-1a2b3c4d",
"SecurityGroups": [
{
"GroupId": "sg-1a2b3c4d"
}
],
"InstanceType": "r3.8xlarge",
"Placement": {
"AvailabilityZone": "us-west-2b"
}
}
]
}

Subnet

{
"TargetCapacity": 20,
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"SecurityGroups": [
{
"GroupId": "sg-1a2b3c4d"
}
],
"InstanceType": "cc2.8xlarge",
"SubnetId": "subnet-1a2b3c4d"
},
{
"ImageId": "ami-1a2b3c4d",
"SecurityGroups": [
{
"GroupId": "sg-1a2b3c4d"
}
],
"InstanceType": "r3.8xlarge",
"SubnetId": "subnet-1a2b3c4d"
}
]
}

Example 4. Override the price for the request

We recommended that you use the default maximum price, which is the On-Demand price. If you
prefer, you can specify a maximum price for the fleet request and maximum prices for individual launch
specifications.

422
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

The following examples specify a maximum price for the fleet request and maximum prices for two
of the three launch specifications. The maximum price for the fleet request is used for any launch
specification that does not specify a maximum price. The Spot Fleet launches the instances using the
instance type with the lowest price.

Availability Zone

{
"SpotPrice": "1.00",
"TargetCapacity": 30,
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "c3.2xlarge",
"Placement": {
"AvailabilityZone": "us-west-2b"
},
"SpotPrice": "0.10"
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "c3.4xlarge",
"Placement": {
"AvailabilityZone": "us-west-2b"
},
"SpotPrice": "0.20"
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "c3.8xlarge",
"Placement": {
"AvailabilityZone": "us-west-2b"
}
}
]
}

Subnet

{
"SpotPrice": "1.00",
"TargetCapacity": 30,
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "c3.2xlarge",
"SubnetId": "subnet-1a2b3c4d",
"SpotPrice": "0.10"
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "c3.4xlarge",
"SubnetId": "subnet-1a2b3c4d",
"SpotPrice": "0.20"
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "c3.8xlarge",
"SubnetId": "subnet-1a2b3c4d"
}
]

423
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Example 5: Launch a Spot Fleet using the diversified allocation strategy

The following example uses the diversified allocation strategy. The launch specifications have
different instance types but the same AMI and Availability Zone or subnet. The Spot Fleet distributes the
30 instances across the three launch specifications, such that there are 10 instances of each type. For
more information, see Allocation strategy for Spot Instances (p. 372).

Availability Zone

{
"SpotPrice": "0.70",
"TargetCapacity": 30,
"AllocationStrategy": "diversified",
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "c4.2xlarge",
"Placement": {
"AvailabilityZone": "us-west-2b"
}
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "m3.2xlarge",
"Placement": {
"AvailabilityZone": "us-west-2b"
}
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "r3.2xlarge",
"Placement": {
"AvailabilityZone": "us-west-2b"
}
}
]
}

Subnet

{
"SpotPrice": "0.70",
"TargetCapacity": 30,
"AllocationStrategy": "diversified",
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "c4.2xlarge",
"SubnetId": "subnet-1a2b3c4d"
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "m3.2xlarge",
"SubnetId": "subnet-1a2b3c4d"
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "r3.2xlarge",
"SubnetId": "subnet-1a2b3c4d"

424
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

}
]
}

A best practice to increase the chance that a spot request can be fulfilled by EC2 capacity in the event
of an outage in one of the Availability Zones is to diversify across zones. For this scenario, include each
Availability Zone available to you in the launch specification. And, instead of using the same subnet each
time, use three unique subnets (each mapping to a different zone).

Availability Zone

{
"SpotPrice": "0.70",
"TargetCapacity": 30,
"AllocationStrategy": "diversified",
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "c4.2xlarge",
"Placement": {
"AvailabilityZone": "us-west-2a"
}
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "m3.2xlarge",
"Placement": {
"AvailabilityZone": "us-west-2b"
}
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "r3.2xlarge",
"Placement": {
"AvailabilityZone": "us-west-2c"
}
}
]
}

Subnet

{
"SpotPrice": "0.70",
"TargetCapacity": 30,
"AllocationStrategy": "diversified",
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "c4.2xlarge",
"SubnetId": "subnet-1a2b3c4d"
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "m3.2xlarge",
"SubnetId": "subnet-2a2b3c4d"
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "r3.2xlarge",
"SubnetId": "subnet-3a2b3c4d"

425
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

}
]
}

Example 6: Launch a Spot Fleet using instance weighting

The following examples use instance weighting, which means that the price is per unit hour instead of
per instance hour. Each launch configuration lists a different instance type and a different weight. The
Spot Fleet selects the instance type with the lowest price per unit hour. The Spot Fleet calculates the
number of Spot Instances to launch by dividing the target capacity by the instance weight. If the result
isn't an integer, the Spot Fleet rounds it up to the next integer, so that the size of your fleet is not below
its target capacity.

If the r3.2xlarge request is successful, Spot provisions 4 of these instances. Divide 20 by 6 for a total
of 3.33 instances, then round up to 4 instances.

If the c3.xlarge request is successful, Spot provisions 7 of these instances. Divide 20 by 3 for a total of
6.66 instances, then round up to 7 instances.

For more information, see Spot Fleet instance weighting (p. 376).

Availability Zone

{
"SpotPrice": "0.70",
"TargetCapacity": 20,
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "r3.2xlarge",
"Placement": {
"AvailabilityZone": "us-west-2b"
},
"WeightedCapacity": 6
},
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "c3.xlarge",
"Placement": {
"AvailabilityZone": "us-west-2b"
},
"WeightedCapacity": 3
}
]
}

Subnet

{
"SpotPrice": "0.70",
"TargetCapacity": 20,
"IamFleetRole": "arn:aws:iam::123456789012:role/aws-ec2-spot-fleet-tagging-role",
"LaunchSpecifications": [
{
"ImageId": "ami-1a2b3c4d",
"InstanceType": "r3.2xlarge",
"SubnetId": "subnet-1a2b3c4d",
"WeightedCapacity": 6
},
{
"ImageId": "ami-1a2b3c4d",

426
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

"InstanceType": "c3.xlarge",
"SubnetId": "subnet-1a2b3c4d",
"WeightedCapacity": 3
}
]
}

Example 7: Launch a Spot Fleet with On-Demand capacity

To ensure that you always have instance capacity, you can include a request for On-Demand capacity in
your Spot Fleet request. If there is capacity, the On-Demand request is always fulfilled. The balance of
the target capacity is fulfilled as Spot if there is capacity and availability.

The following example specifies the desired target capacity as 10, of which 5 must be On-Demand
capacity. Spot capacity is not specified; it is implied in the balance of the target capacity minus the On-
Demand capacity. Amazon EC2 launches 5 capacity units as On-Demand, and 5 capacity units (10-5=5) as
Spot if there is available Amazon EC2 capacity and availability.

For more information, see On-Demand in Spot Fleet (p. 372).

{
"IamFleetRole": "arn:aws:iam::781603563322:role/aws-ec2-spot-fleet-tagging-role",
"AllocationStrategy": "lowestPrice",
"TargetCapacity": 10,
"SpotPrice": null,
"ValidFrom": "2018-04-04T15:58:13Z",
"ValidUntil": "2019-04-04T15:58:13Z",
"TerminateInstancesWithExpiration": true,
"LaunchSpecifications": [],
"Type": "maintain",
"OnDemandTargetCapacity": 5,
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-0dbb04d4a6cca5ad1",
"Version": "2"
},
"Overrides": [
{
"InstanceType": "t2.medium",
"WeightedCapacity": 1,
"SubnetId": "subnet-d0dc51fb"
}
]
}
]
}

Example 8: Configure Capacity Rebalancing to launch replacement Spot Instances

The following example configures the Spot Fleet to launch a replacement Spot Instance when Amazon
EC2 emits a rebalance recommendation for a Spot Instance in the fleet. To configure the automatic
replacement of Spot Instances, for ReplacementStrategy, specify launch.
Note
When a replacement instance is launched, the instance marked for rebalance is not
automatically terminated. You can terminate it, or you can leave it running. You are charged for
both instances while they are running.

The effectiveness of the Capacity Rebalancing strategy depends on the number of Spot Instance pools
specified in the Spot Fleet request. We recommend that you configure the fleet with a diversified set of
instance types and Availability Zones, and for AllocationStrategy, specify capacityOptimized.

427
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

For more information about what you should consider when configuring a Spot Fleet for Capacity
Rebalancing, see Capacity Rebalancing (p. 374).

{
"SpotFleetRequestConfig": {
"AllocationStrategy": "capacityOptimized",
"IamFleetRole": "arn:aws:iam::000000000000:role/aws-ec2-spot-fleet-tagging-role",
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateName": "LaunchTemplate",
"Version": "1"
},
"Overrides": [
{
"InstanceType": "c3.large",
"WeightedCapacity": 1,
"Placement": {
"AvailabilityZone": "us-east-1a"
}
},
{
"InstanceType": "c4.large",
"WeightedCapacity": 1,
"Placement": {
"AvailabilityZone": "us-east-1a"
}
},
{
"InstanceType": "c5.large",
"WeightedCapacity": 1,
"Placement": {
"AvailabilityZone": "us-east-1a"
}
}
]
}
],
"TargetCapacity": 5,
"SpotMaintenanceStrategies": {
"CapacityRebalance": {
"ReplacementStrategy": "launch"
}
}
}
}

CloudWatch metrics for Spot Fleet

Amazon EC2 provides Amazon CloudWatch metrics that you can use to monitor your Spot Fleet.
Important
To ensure accuracy, we recommend that you enable detailed monitoring when using
these metrics. For more information, see Enable or turn off detailed monitoring for your
instances (p. 761).

For more information about CloudWatch metrics provided by Amazon EC2, see Monitor your instances
using CloudWatch (p. 761).

Spot Fleet metrics

The AWS/EC2Spot namespace includes the following metrics, plus the CloudWatch metrics for the Spot
Instances in your fleet. For more information, see Instance metrics (p. 764).

428
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Metric Description

AvailableInstancePoolsCount The Spot Instance pools specified in the Spot Fleet request.

Units: Count

BidsSubmittedForCapacity The capacity for which Amazon EC2 has submitted Spot Fleet
requests.

Units: Count

EligibleInstancePoolCount The Spot Instance pools specified in the Spot Fleet request
where Amazon EC2 can fulfill requests. Amazon EC2 does not
fulfill requests in pools where the maximum price you're willing
to pay for Spot Instances is less than the Spot price or the Spot
price is greater than the price for On-Demand Instances.

Units: Count

FulfilledCapacity The capacity that Amazon EC2 has fulfilled.

Units: Count

MaxPercentCapacityAllocation The maximum value of PercentCapacityAllocation across

all Spot Fleet pools specified in the Spot Fleet request.

Units: Percent

PendingCapacity The difference between TargetCapacity and

FulfilledCapacity.

Units: Count

PercentCapacityAllocation The capacity allocated for the Spot Instance pool

for the specified dimensions. To get the maximum
value recorded across all Spot Instance pools, use
MaxPercentCapacityAllocation.

Units: Percent

TargetCapacity The target capacity of the Spot Fleet request.

Units: Count

TerminatingCapacity The capacity that is being terminated because the provisioned

capacity is greater than the target capacity.

Units: Count

If the unit of measure for a metric is Count, the most useful statistic is Average.

Spot Fleet dimensions

To filter the data for your Spot Fleet, use the following dimensions.

Dimensions Description

AvailabilityZone Filter the data by Availability Zone.

429
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Dimensions Description

FleetRequestId Filter the data by Spot Fleet request.

InstanceType Filter the data by instance type.

View the CloudWatch metrics for your Spot Fleet

You can view the CloudWatch metrics for your Spot Fleet using the Amazon CloudWatch console. These
metrics are displayed as monitoring graphs. These graphs show data points if the Spot Fleet is active.

Metrics are grouped first by namespace, and then by the various combinations of dimensions within each
namespace. For example, you can view all Spot Fleet metrics or Spot Fleet metrics groups by Spot Fleet
request ID, instance type, or Availability Zone.

To view Spot Fleet metrics

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. In the navigation pane, choose Metrics.
3. Choose the EC2 Spot namespace.
Note
If the EC2 Spot namespace is not displayed, there are two reasons for this. Either you've
not yet used Spot Fleet—only the AWS services that you're using send metrics to Amazon
CloudWatch. Or, if you’ve not used Spot Fleet for the past two weeks, the namespace does
not appear.
4. (Optional) To filter the metrics by dimension, select one of the following:

• Fleet Request Metrics – Group by Spot Fleet request

• By Availability Zone – Group by Spot Fleet request and Availability Zone
• By Instance Type – Group by Spot Fleet request and instance type
• By Availability Zone/Instance Type – Group by Spot Fleet request, Availability Zone, and instance
type
5. To view the data for a metric, select the check box next to the metric.

430
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Use Amazon EventBridge to monitor Spot Fleet events

When the state of a Spot Fleet changes, the Spot Fleet emits a notification. The notification is made
available as an event that is sent to Amazon EventBridge (formerly known as Amazon CloudWatch
Events). Events are emitted on a best effort basis.

With Amazon EventBridge, you can create rules that trigger programmatic actions in response to
an event. For example, you can create two EventBridge rules, one that's triggered when a fleet state
changes, and one that's triggered when an instance in the fleet is terminated. If the fleet state changes,
the first EventBridge rule invokes an SNS topic to send an email notification to you. If an instance is
terminated, the second EventBridge rule invokes a Lambda function to launch a new instance.

Topics
• Spot Fleet event types (p. 431)
• Use Amazon EventBridge (p. 435)

Spot Fleet event types

There are five Spot Fleet event types. For each event type, there are several sub-types.

The events are sent to EventBridge in JSON format. The following fields in the event form the event
pattern that is defined in the rule, and which trigger an action:

"source": "aws.ec2spotfleet"

Identifies that the event is from Spot Fleet.

"detail-type": "EC2 Spot Fleet State Change"

Identifies the event type.

"sub-type": "submitted"

Identifies the event sub-type.

Event types
• EC2 Spot Fleet State Change (p. 431)
• EC2 Spot Fleet Spot Instance Request Change (p. 432)
• EC2 Spot Fleet Instance Change (p. 433)
• EC2 Spot Fleet Information (p. 434)
• EC2 Spot Fleet Error (p. 435)

EC2 Spot Fleet State Change

Spot Fleet sends an EC2 Spot Fleet State Change event to Amazon EventBridge when a Spot Fleet
changes state.

The following is example data for this event.

{
"version": "0",
"id": "d1af1091-6cc3-2e24-203a-3b870e455d5b",
"detail-type": "EC2 Spot Fleet State Change",
"source": "aws.ec2spotfleet",
"account": "123456789012",

431
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

"time": "2020-11-09T08:57:06Z",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:123456789012:spot-fleet-request/sfr-4b6d274d-0cea-4b2c-
b3be-9dc627ad1f55"
],
"detail": {
"sub-type": "submitted"
}
}

The possible values for sub-type are:

submitted

The Spot Fleet request is being evaluated and Amazon EC2 is preparing to launch the target number
of instances.
active

The Spot Fleet request has been validated and Amazon EC2 is attempting to maintain the target
number of running Spot Instances.
progress

The Spot Fleet request is in the process of being fulfilled.

cancelled_terminating

The Spot Fleet request is deleted and its instances are terminating. The request remains in this state
until all instances are terminated.
cancelled_running

The Spot Fleet request is deleted and does not launch additional instances. Its existing instances
continue to run until they are interrupted or terminated. The request remains in this state until all
instances are interrupted or terminated.
cancelled

The Spot Fleet request is deleted and has no running instances. The Spot Fleet will be deleted two
days after its instances are terminated.
modify_in_progress

The Spot Fleet request is being modified. The request remains in this state until the modification is
fully processed or the Spot Fleet request is deleted.
modify_succeeded

The Spot Fleet request was modified.

expired

The Spot Fleet request has expired. If the request was created with
TerminateInstancesWithExpiration set, a subsequent event indicates that the instances are
terminated.

EC2 Spot Fleet Spot Instance Request Change

Spot Fleet sends an EC2 Spot Fleet Spot Instance Request Change event to Amazon
EventBridge when a Spot Instance request in the fleet changes state.

The following is example data for this event.

432
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

{
"version": "0",
"id": "cd141ef0-14af-d670-a71d-fe46e9971bd2",
"detail-type": "EC2 Spot Fleet Spot Instance Request Change",
"source": "aws.ec2spotfleet",
"account": "123456789012",
"time": "2020-11-09T08:53:21Z",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:123456789012:spot-fleet-request/sfr-
a98d2133-941a-47dc-8b03-0f94c6852ad1"
],
"detail": {
"spot-instance-request-id": "sir-a2w9gc5h",
"description": "SpotInstanceRequestId sir-a2w9gc5h, PreviousState:
cancelled_running",
"sub-type": "cancelled"
}
}

The possible values for sub-type are:

submitted

The request is submitted.

disabled

You stopped the Spot Instance.

active

The request is fulfilled and has an associated Spot Instance.

cancelled

You cancelled the request, or the request expired.

EC2 Spot Fleet Instance Change

Spot Fleet sends an EC2 Spot Fleet Instance Change event to Amazon EventBridge when an
instance in the fleet changes state.

The following is example data for this event.

{
"version": "0",
"id": "11591686-5bd7-bbaa-eb40-d46529c2710f",
"detail-type": "EC2 Spot Fleet Instance Change",
"source": "aws.ec2spotfleet",
"account": "123456789012",
"time": "2020-11-09T07:25:02Z",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:123456789012:spot-fleet-request/sfr-c8a764a4-bedc-4b62-
af9c-0095e6e3ba61"
],
"detail": {
"instance-id": "i-08b90df1e09c30c9b",
"description": "{\"instanceType\":\"r4.2xlarge\",\"image\":\"ami-032930428bf1abbff
\",\"productDescription\":\"Linux/UNIX\",\"availabilityZone\":\"us-east-1a\"}",
"sub-type": "launched"
}

433
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

The possible values for sub-type are:

launched

A new instance was launched.

terminated

The instance was terminated.

termination_notified

An instance termination notification was sent.

EC2 Spot Fleet Information

Spot Fleet sends an EC2 Spot Fleet Information event to Amazon EventBridge when there is an
error during fulfillment. The information event does not block the fleet from attempting to fulfil its
target capacity.

The following is example data for this event.

{
"version": "0",
"id": "73a60f70-3409-a66c-635c-7f66c5f5b669",
"detail-type": "EC2 Spot Fleet Information",
"source": "aws.ec2spotfleet",
"account": "123456789012",
"time": "2020-11-08T20:56:12Z",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:123456789012:spot-fleet-request/sfr-2531ea06-
af18-4647-8757-7d69c94971b1"
],
"detail": {
"description": "r3.8xlarge, ami-032930428bf1abbff, Linux/UNIX, us-east-1a, Spot bid
price is less than Spot market price $0.5291",
"sub-type": "launchSpecUnusable"
}
}

The possible values for sub-type are:

launchSpecUnusable

The price in a launch specification is not valid because it is below the Spot price or the Spot price is
above the On-Demand price.
fleetProgressHalted

The price in every launch specification is not valid. A launch specification might become valid if the
Spot price changes.
registerWithLoadBalancersFailed

An attempt to register instances with load balancers failed. For more information, see the
description of the event.
launchSpecTemporarilyBlacklisted

The configuration is not valid and several attempts to launch instances have failed. For more
information, see the description of the event.

434
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

EC2 Spot Fleet Error

Spot Fleet sends an EC2 Spot Fleet Error event to Amazon EventBridge when there is an error
during fulfillment. The error event blocks the fleet from attempting to fulfil its target capacity.

The following is example data for this event.

{
"version": "0",
"id": "10adc4e7-675c-643e-125c-5bfa1b1ba5d2",
"detail-type": "EC2 Spot Fleet Error",
"source": "aws.ec2spotfleet",
"account": "123456789012",
"time": "2020-11-09T06:56:07Z",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:123456789012:spot-fleet-request/
sfr-38725d30-25f1-4f30-83ce-2907c56dba17"
],
"detail": {
"description": "r4.2xlarge, ami-032930428bf1abbff, Linux/UNIX: The
associatePublicIPAddress parameter can only be specified for the network interface with
DeviceIndex 0. ",
"sub-type": "spotFleetRequestConfigurationInvalid"
}
}

The possible values for sub-type are:

allLaunchSpecsTemporarilyBlacklisted

None of the configurations are valid, and several attempts to launch instances have failed. For more
information, see the description of the event.
spotFleetRequestConfigurationInvalid

The configuration is not valid. For more information, see the description of the event.
spotInstanceCountLimitExceeded

You’ve reached the limit on the number of Spot Instances that you can launch.

Use Amazon EventBridge

When a notification of a state change is emitted for a Spot Fleet, the event for the notification is sent to
Amazon EventBridge. If EventBridge detects an event pattern that matches a pattern defined in a rule,
EventBridge invokes a target (or targets) specified in the rule.

The preceding sections provide example events for the Spot Fleet notifications. The event JSON files
contain the fields that form the event pattern.

The following fields form the event pattern that is defined in the rule:

"source": "aws.ec2spotfleet"

Identifies that the event is from Spot Fleet.

"detail-type": "EC2 Spot Fleet State Change"

Identifies the event type.

435
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

"sub-type": "submitted"

Identifies the event sub-type.

You can write an EventBridge rule and automate what actions to take when the event pattern matches
the rule.

Examples
• Create an EventBridge rule to send a notification (p. 436)
• Create an EventBridge rule to trigger a Lambda function (p. 436)

Create an EventBridge rule to send a notification

The following example creates an EventBridge rule to send an email, text message, or mobile push
notification every time that Amazon EC2 emits a Spot Fleet notification. The signal in this example is
emitted as an EC2 Spot Fleet State Change event, which triggers the action defined by the rule.

To create an EventBridge rule to send a notification when a Spot Fleet state changes

1. Open the Amazon EventBridge console at https://console.aws.amazon.com/events/.

2. Choose Create rule.
3. Enter a Name for the rule, and, optionally, a description.

A rule can't have the same name as another rule in the same Region and on the same event bus.
4. For Define pattern, choose Event pattern.
5. Under Event matching pattern, choose Custom pattern.
6. In the Event pattern box, add the following pattern to match the EC2 Spot Fleet State
Change event, and then choose Save.

{
"source": [ "aws.ec2spotfleet" ],
"detail-type": [ "EC2 Spot Fleet State Change" ],
"sub-type": [ "submitted" ]
}

7. For Select event bus, choose AWS default event bus. When an AWS service in your account emits an
event, it always goes to your account's default event bus.
8. Confirm that Enable the rule on the selected event bus is toggled on.
9. For Target, choose SNS topic to send an email, text message, or mobile push notification when the
event occurs.
10. For Topic, choose an existing topic. You first need to create an Amazon SNS topic using the Amazon
SNS console. For more information, see Using Amazon SNS for application-to-person (A2P)
messaging in the Amazon Simple Notification Service Developer Guide.
11. For Configure input, choose the input for the email, text message, or mobile push notification.
12. Choose Create.

For more information, see Creating a rule for an AWS service and Event Patterns in the Amazon
EventBridge User Guide

Create an EventBridge rule to trigger a Lambda function

The following example creates an EventBridge rule to trigger a Lambda function every time that Amazon
EC2 emits a Spot Fleet notification. The signal in this example is emitted as an EC2 Spot Fleet

436
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Instance Change event, which triggers the action defined by the rule. Before creating the EventBridge
rule, you must create the Lambda function.

To create an EventBridge rule to trigger a Lambda function when an instance in a Spot Fleet
changes state

1. Open the AWS Lambda console at https://console.aws.amazon.com/lambda/.

2. Choose Create function.
3. Enter a name for your function, configure the code, and then choose Create function.

For more information about using Lambda, see Create a Lambda function with the console in the
AWS Lambda Developer Guide.
4. Open the Amazon EventBridge console at https://console.aws.amazon.com/events/.
5. Choose Create rule.
6. Enter a Name for the rule, and, optionally, a description.

A rule can't have the same name as another rule in the same Region and on the same event bus.
7. For Define pattern, choose Event pattern.
8. Under Event matching pattern, you can choose Pre-defined pattern by service or Custom pattern.
The Custom pattern allows you to create a more detailed rule.

a. If you choose Pre-defined pattern by service, do the following:

i. For Service provider, choose AWS.

ii. For Service name, choose EC2 Spot Fleet.
iii. For Event type, select the required event type. For this example, choose EC2 Spot Fleet
Instance Change.
b. If you choose Custom pattern, do the following:

• In the Event pattern box, add the following pattern to match the EC2 Spot Fleet
Instance Change event and launched sub-type for this example, and then choose Save.

{
"source": [ "aws.ec2spotfleet" ],
"detail-type": [ "EC2 Spot Fleet Instance Change" ],
"sub-type": [ "launched" ]
}

9. For Target, choose Lambda function, and for Function, choose the function that you created to
respond when the event occurs.
10. Choose Create.

In this example, the Lambda function will be triggered when the EC2 Fleet Instance Change
event with the sub-type launched occurs.

For a tutorial on how to create a Lambda function and an EventBridge rule that runs the Lambda
function, see Tutorial: Log the State of an Amazon EC2 Instance Using EventBridge in the AWS Lambda
Developer Guide.

Automatic scaling for Spot Fleet

Automatic scaling is the ability to increase or decrease the target capacity of your Spot Fleet
automatically based on demand. A Spot Fleet can either launch instances (scale out) or terminate
instances (scale in), within the range that you choose, in response to one or more scaling policies.

Spot Fleet supports the following types of automatic scaling:

437
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

• Target tracking scaling (p. 439) – Increase or decrease

the current capacity of the fleet based on a target value for a specific metric. This is similar to the
way that your thermostat maintains the temperature of your home—you select temperature and the
thermostat does the rest.
• Step scaling (p. 440) – Increase or decrease the current capacity of the fleet based on a set of scaling
adjustments, known as step adjustments, that vary based on the size of the alarm breach.
• Scheduled scaling (p. 442) – Increase or decrease the current capacity of the fleet based on the date
and time.

If you are using instance weighting (p. 376), keep in mind that Spot Fleet can exceed the target capacity
as needed. Fulfilled capacity can be a floating-point number but target capacity must be an integer,
so Spot Fleet rounds up to the next integer. You must take these behaviors into account when you
look at the outcome of a scaling policy when an alarm is triggered. For example, suppose that the
target capacity is 30, the fulfilled capacity is 30.1, and the scaling policy subtracts 1. When the alarm is
triggered, the automatic scaling process subtracts 1 from 30.1 to get 29.1 and then rounds it up to 30, so
no scaling action is taken. As another example, suppose that you selected instance weights of 2, 4, and 8,
and a target capacity of 10, but no weight 2 instances were available so Spot Fleet provisioned instances
of weights 4 and 8 for a fulfilled capacity of 12. If the scaling policy decreases target capacity by 20%
and an alarm is triggered, the automatic scaling process subtracts 12*0.2 from 12 to get 9.6 and then
rounds it up to 10, so no scaling action is taken.

The scaling policies that you create for Spot Fleet support a cooldown period. This is the number of
seconds after a scaling activity completes where previous trigger-related scaling activities can influence
future scaling events. For scale-out policies, while the cooldown period is in effect, the capacity that
has been added by the previous scale-out event that initiated the cooldown is calculated as part of
the desired capacity for the next scale out. The intention is to continuously (but not excessively) scale
out. For scale in policies, the cooldown period is used to block subsequent scale in requests until it has
expired. The intention is to scale in conservatively to protect your application's availability. However, if
another alarm triggers a scale-out policy during the cooldown period after a scale-in, automatic scaling
scales out your scalable target immediately.

We recommend that you scale based on instance metrics with a 1-minute frequency because that
ensures a faster response to utilization changes. Scaling on metrics with a 5-minute frequency can
result in slower response time and scaling on stale metric data. To send metric data for your instances
to CloudWatch in 1-minute periods, you must specifically enable detailed monitoring. For more
information, see Enable or turn off detailed monitoring for your instances (p. 761) and Create a Spot
Fleet request using defined parameters (console) (p. 407).

For more information about configuring scaling for Spot Fleet, see the following resources:

• application-autoscaling section of the AWS CLI Command Reference

• Application Auto Scaling API Reference
• Application Auto Scaling User Guide

IAM permissions required for Spot Fleet automatic scaling

Automatic scaling for Spot Fleet is made possible by a combination of the Amazon EC2, Amazon
CloudWatch, and Application Auto Scaling APIs. Spot Fleet requests are created with Amazon EC2,
alarms are created with CloudWatch, and scaling policies are created with Application Auto Scaling.

In addition to the IAM permissions for Spot Fleet (p. 403) and Amazon EC2, the IAM user that accesses
fleet scaling settings must have the appropriate permissions for the services that support dynamic
scaling. IAM users must have permissions to use the actions shown in the following example policy.

438
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"application-autoscaling:*",
"ec2:DescribeSpotFleetRequests",
"ec2:ModifySpotFleetRequest",
"cloudwatch:DeleteAlarms",
"cloudwatch:DescribeAlarmHistory",
"cloudwatch:DescribeAlarms",
"cloudwatch:DescribeAlarmsForMetric",
"cloudwatch:GetMetricStatistics",
"cloudwatch:ListMetrics",
"cloudwatch:PutMetricAlarm",
"cloudwatch:DisableAlarmActions",
"cloudwatch:EnableAlarmActions",
"iam:CreateServiceLinkedRole",
"sns:CreateTopic",
"sns:Subscribe",
"sns:Get*",
"sns:List*"
],
"Resource": "*"
}
]
}

You can also create your own IAM policies that allow more fine-grained permissions for calls to the
Application Auto Scaling API. For more information, see Authentication and Access Control in the
Application Auto Scaling User Guide.

The Application Auto Scaling service also needs permission to describe your Spot Fleet and
CloudWatch alarms, and permissions to modify your Spot Fleet target capacity on your behalf.
If you enable automatic scaling for your Spot Fleet, it creates a service-linked role named
AWSServiceRoleForApplicationAutoScaling_EC2SpotFleetRequest. This service-linked role
grants Application Auto Scaling permission to describe the alarms for your policies, to monitor the
current capacity of the fleet, and to modify the capacity of the fleet. The original managed Spot Fleet
role for Application Auto Scaling was aws-ec2-spot-fleet-autoscale-role, but it is no longer
required. The service-linked role is the default role for Application Auto Scaling. For more information,
see Service-Linked Roles in the Application Auto Scaling User Guide.

Scale Spot Fleet using a target tracking policy

With target tracking scaling policies, you select a metric and set a target value. Spot Fleet creates and
manages the CloudWatch alarms that trigger the scaling policy and calculates the scaling adjustment
based on the metric and the target value. The scaling policy adds or removes capacity as required to keep
the metric at, or close to, the specified target value. In addition to keeping the metric close to the target
value, a target tracking scaling policy also adjusts to the fluctuations in the metric due to a fluctuating
load pattern and minimizes rapid fluctuations in the capacity of the fleet.

You can create multiple target tracking scaling policies for a Spot Fleet, provided that each of them
uses a different metric. The fleet scales based on the policy that provides the largest fleet capacity. This
enables you to cover multiple scenarios and ensure that there is always enough capacity to process your
application workloads.

To ensure application availability, the fleet scales out proportionally to the metric as fast as it can, but
scales in more gradually.

When a Spot Fleet terminates an instance because the target capacity was decreased, the instance
receives a Spot Instance interruption notice.

439
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Do not edit or delete the CloudWatch alarms that Spot Fleet manages for a target tracking scaling policy.
Spot Fleet deletes the alarms automatically when you delete the target tracking scaling policy.

Limitation

The Spot Fleet request must have a request type of maintain. Automatic scaling is not supported for
requests of type request, or Spot blocks.

To configure a target tracking policy (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Spot Requests.
3. Select your Spot Fleet request and choose Auto Scaling.
4. If automatic scaling is not configured, choose Configure.
5. Use Scale capacity between to set the minimum and maximum capacity for your fleet. Automatic
scaling does not scale your fleet below the minimum capacity or above the maximum capacity.
6. For Policy name, enter a name for the policy.
7. Choose a Target metric.
8. Enter a Target value for the metric.
9. (Optional) Set Cooldown period to modify the default cooldown period.
10. (Optional) Select Disable scale-in to omit creating a scale-in policy based on the current
configuration. You can create a scale-in policy using a different configuration.
11. Choose Save.

To configure a target tracking policy using the AWS CLI

1. Register the Spot Fleet request as a scalable target using the register-scalable-target command.
2. Create a scaling policy using the put-scaling-policy command.

Scale Spot Fleet using step scaling policies

With step scaling policies, you specify CloudWatch alarms to trigger the scaling process. For example,
if you want to scale out when CPU utilization reaches a certain level, create an alarm using the
CPUUtilization metric provided by Amazon EC2.

When you create a step scaling policy, you must specify one of the following scaling adjustment types:

• Add – Increase the target capacity of the fleet by a specified number of capacity units or a specified
percentage of the current capacity.
• Remove – Decrease the target capacity of the fleet by a specified number of capacity units or a
specified percentage of the current capacity.
• Set to – Set the target capacity of the fleet to the specified number of capacity units.

When an alarm is triggered, the automatic scaling process calculates the new target capacity using the
fulfilled capacity and the scaling policy, and then updates the target capacity accordingly. For example,
suppose that the target capacity and fulfilled capacity are 10 and the scaling policy adds 1. When
the alarm is triggered, the automatic scaling process adds 1 to 10 to get 11, so Spot Fleet launches 1
instance.

When a Spot Fleet terminates an instance because the target capacity was decreased, the instance
receives a Spot Instance interruption notice.

440
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Limitation

The Spot Fleet request must have a request type of maintain. Automatic scaling is not supported for
requests of type request, or Spot blocks.

Prerequisites

• Consider which CloudWatch metrics are important to your application. You can create CloudWatch
alarms based on metrics provided by AWS or your own custom metrics.
• For the AWS metrics that you will use in your scaling policies, enable CloudWatch metrics collection if
the service that provides the metrics does not enable it by default.

To create a CloudWatch alarm

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. In the navigation pane, choose Alarms.
3. Choose Create alarm.
4. On the Specify metric and conditions page, choose Select metric.
5. Choose EC2 Spot, Fleet Request Metrics, select a metric (for example, CPUUtilization), and then
choose Select metric.

The Specify metric and conditions page appears, showing a graph and other information about the
metric you selected.
6. For Period, choose the evaluation period for the alarm, for example, 1 minute. When evaluating the
alarm, each period is aggregated into one data point.
Note
A shorter period creates a more sensitive alarm.
7. For Conditions, define the alarm by defining the threshold condition. For example, you can define
a threshold to trigger the alarm whenever the value of the metric is greater than or equal to 80
percent.
8. Under Additional configuration, for Datapoints to alarm, specify how many datapoints (evaluation
periods) must be in the ALARM state to trigger the alarm, for example, 1 evaluation period or 2 out
of 3 evaluation periods. This creates an alarm that goes to ALARM state if that many consecutive
periods are breaching. For more information, see Evaluating an Alarm in the Amazon CloudWatch
User Guide.
9. For Missing data treatment, choose one of the options (or leave the default of Treat missing data
as missing). For more information, see Configuring How CloudWatch Alarms Treat Missing Data in
the Amazon CloudWatch User Guide.
10. Choose Next.
11. (Optional) To receive notification of a scaling event, for Notification, you can choose or create
the Amazon SNS topic you want to use to receive notifications. Otherwise, you can delete the
notification now and add one later as needed.
12. Choose Next.
13. Under Add a description, enter a name and description for the alarm and choose Next.
14. Choose Create alarm.

To configure a step scaling policy for your Spot Fleet (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Spot Requests.
3. Select your Spot Fleet request and choose Auto Scaling.

441
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

4. If automatic scaling is not configured, choose Configure.

5. Use Scale capacity between to set the minimum and maximum capacity for your fleet. Automatic
scaling does not scale your fleet below the minimum capacity or above the maximum capacity.
6. Initially, Scaling policies contains policies named ScaleUp and ScaleDown. You can complete these
policies, or choose Remove policy to delete them. You can also choose Add policy.
7. To define a policy, do the following:

a. For Policy name, enter a name for the policy.

b. For Policy trigger, select an existing alarm or choose Create new alarm to open the Amazon
CloudWatch console and create an alarm.
c. For Modify capacity, select a scaling adjustment type, select a number, and select a unit.
d. (Optional) To perform step scaling, choose Define steps. By default, an add policy has a lower
bound of -infinity and an upper bound of the alarm threshold. By default, a remove policy has
a lower bound of the alarm threshold and an upper bound of +infinity. To add another step,
choose Add step.
e. (Optional) To modify the default value for the cooldown period, select a number from
Cooldown period.
8. Choose Save.

To configure step scaling policies for your Spot Fleet using the AWS CLI

1. Register the Spot Fleet request as a scalable target using the register-scalable-target command.
2. Create a scaling policy using the put-scaling-policy command.
3. Create an alarm that triggers the scaling policy using the put-metric-alarm command.

Scale Spot Fleet using scheduled scaling

Scaling based on a schedule enables you to scale your application in response to predictable changes in
demand. To use scheduled scaling, you create scheduled actions, which tell Spot Fleet to perform scaling
activities at specific times. When you create a scheduled action, you specify an existing Spot Fleet, when
the scaling activity should occur, minimum capacity, and maximum capacity. You can create scheduled
actions that scale one time only or that scale on a recurring schedule.

You can only create a scheduled action for Spot Fleets that already exist. You can't create a scheduled
action at the same time that you create a Spot Fleet.

Limitation

The Spot Fleet request must have a request type of maintain. Automatic scaling is not supported for
requests of type request, or Spot blocks.

To create a one-time scheduled action

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Spot Requests.
3. Select your Spot Fleet request and choose the Scheduled Scaling tab near the bottom of the screen.
4. Choose Create Scheduled Action.
5. For Name, specify a name for the scheduled action.
6. Enter a value for Minimum capacity, Maximum capacity, or both.
7. For Recurrence, choose Once.
8. (Optional) Choose a date and time for Start time, End time, or both.

442
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

9. Choose Submit.

To scale on a recurring schedule

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Spot Requests.
3. Select your Spot Fleet request and choose the Scheduled Scaling tab near the bottom of the screen.
4. For Recurrence, choose one of the predefined schedules (for example, Every day), or choose
Custom and enter a cron expression. For more information about the cron expressions supported by
scheduled scaling, see Cron Expressions in the Amazon CloudWatch Events User Guide.
5. (Optional) Choose a date and time for Start time, End time, or both.
6. Choose Submit.

To edit a scheduled action

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Spot Requests.
3. Select your Spot Fleet request and choose the Scheduled Scaling tab near the bottom of the screen.
4. Select the scheduled action and choose Actions, Edit.
5. Make the needed changes and choose Submit.

To delete a scheduled action

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Spot Requests.
3. Select your Spot Fleet request and choose the Scheduled Scaling tab near the bottom of the screen.
4. Select the scheduled action and choose Actions, Delete.
5. When prompted for confirmation, choose Delete.

To manage scheduled scaling using the AWS CLI

Use the following commands:

• put-scheduled-action
• describe-scheduled-actions
• delete-scheduled-action

Spot request status

To help you track your Spot Instance requests and plan your use of Spot Instances, use the request status
provided by Amazon EC2. For example, the request status can provide the reason why your Spot request
isn't fulfilled yet, or list the constraints that are preventing the fulfillment of your Spot request.

At each step of the process—also called the Spot request lifecycle—specific events determine successive
request states.

Contents
• Lifecycle of a Spot request (p. 444)
• Get request status information (p. 447)

443
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

• Spot request status codes (p. 447)

Lifecycle of a Spot request

The following diagram shows you the paths that your Spot request can follow throughout its lifecycle,
from submission to termination. Each step is depicted as a node, and the status code for each node
describes the status of the Spot request and Spot Instance.

Pending evaluation

As soon as you create a Spot Instance request, it goes into the pending-evaluation state unless one
or more request parameters are not valid (bad-parameters).

Status code Request state Instance state

pending-evaluation open n/a

bad-parameters closed n/a

Holding

If one or more request constraints are valid but can't be met yet, or if there is not enough capacity, the
request goes into a holding state waiting for the constraints to be met. The request options affect the
likelihood of the request being fulfilled. For example, if you specify a maximum price below the current
Spot price, your request stays in a holding state until the Spot price goes below your maximum price.
If you specify an Availability Zone group, the request stays in a holding state until the Availability Zone
constraint is met.

444
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

In the event of an outage of one of the Availability Zones, there is a chance that the spare EC2 capacity
available for Spot Instance requests in other Availability Zones can be affected.

Status code Request state Instance state

capacity-not-available open n/a

price-too-low open n/a

not-scheduled-yet open n/a

launch-group-constraint open n/a

az-group-constraint open n/a

placement-group- open n/a

constraint

constraint-not- open n/a

fulfillable

Pending evaluation/fulfillment-terminal

Your Spot Instance request can go to a terminal state if you create a request that is valid only during
a specific time period and this time period expires before your request reaches the pending fulfillment
phase. It might also happen if you cancel the request, or if a system error occurs.

Status code Request state Instance state

schedule-expired cancelled n/a

canceled-before- cancelled n/a

fulfillment*

bad-parameters failed n/a

system-error closed n/a

* If you cancel the request.

Pending fulfillment

When the constraints you specified (if any) are met and your maximum price is equal to or higher than
the current Spot price, your Spot request goes into the pending-fulfillment state.

At this point, Amazon EC2 is getting ready to provision the instances that you requested. If the process
stops at this point, it is likely to be because it was canceled by the user before a Spot Instance was
launched. It might also be because an unexpected system error occurred.

Status code Request state Instance state

pending-fulfillment open n/a

Fulfilled

445
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

When all the specifications for your Spot Instances are met, your Spot request is fulfilled. Amazon
EC2 launches the Spot Instances, which can take a few minutes. If a Spot Instance is hibernated or
stopped when interrupted, it remains in this state until the request can be fulfilled again or the request is
canceled.

Status code Request state Instance state

fulfilled active pending → running

fulfilled active stopped → running

If you stop a Spot Instance, your Spot request goes into the marked-for-stop or instance-
stopped-by-user state until the Spot Instance can be started again or the request is cancelled.

Status code Request state Instance state

marked-for-stop active stopping

instance-stopped-by-user* disabled or cancelled** stopped

* A Spot Instance goes into the instance-stopped-by-user state if you stop the instance or run the
shutdown command from the instance. After you've stopped the instance, you can start it again. On
restart, the Spot Instance request returns to the pending-evaluation state and then Amazon EC2
launches a new Spot Instance when the constraints are met.

** The Spot request state is disabled if you stop the Spot Instance but do not cancel the request. The
request state is cancelled if your Spot Instance is stopped and the request expires.

Fulfilled-terminal

Your Spot Instances continue to run as long as your maximum price is at or above the Spot price, there
is available capacity for your instance type, and you don't terminate the instance. If a change in the
Spot price or available capacity requires Amazon EC2 to terminate your Spot Instances, the Spot request
goes into a terminal state. A request also goes into the terminal state if you cancel the Spot request or
terminate the Spot Instances.

Status code Request state Instance state

request-canceled-and- cancelled running

instance-running

marked-for-stop active running

marked-for-termination active running

instance-stopped-by- disabled stopped

price

instance-stopped-by-user disabled stopped

instance-stopped-no- disabled stopped

capacity

instance-terminated-by- closed (one-time), terminated

price open(persistent)

446
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Status code Request state Instance state

instance-terminated-by- closed terminated

schedule

instance-terminated-by- cancelled terminated

service

instance-terminated-by- closed or cancelled * terminated

user

instance-terminated-no- closed (one-time), open terminated

capacity (persistent)

instance-terminated- closed (one-time), open terminated

launch-group-constraint (persistent)

* The request state is closed if you terminate the instance but do not cancel the request. The request
state is cancelled if you terminate the instance and cancel the request. Even if you terminate a Spot
Instance before you cancel its request, there might be a delay before Amazon EC2 detects that your Spot
Instance was terminated. In this case, the request state can either be closed or cancelled.

Persistent requests

When your Spot Instances are terminated (either by you or Amazon EC2), if the Spot request is a
persistent request, it returns to the pending-evaluation state and then Amazon EC2 can launch a
new Spot Instance when the constraints are met.

Get request status information

You can get request status information using the AWS Management Console or a command line tool.

To get request status information (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Spot Requests and select the Spot request.
3. To check the status, on the Description tab, check the Status field.

To get request status information using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• describe-spot-instance-requests (AWS CLI)

• Get-EC2SpotInstanceRequest (AWS Tools for Windows PowerShell)

Spot request status codes

Spot request status information is composed of a status code, the update time, and a status message.
Together, these help you determine the disposition of your Spot request.

The following are the Spot request status codes:

az-group-constraint

Amazon EC2 cannot launch all the instances you requested in the same Availability Zone.

447
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

bad-parameters

One or more parameters for your Spot request are not valid (for example, the AMI you specified does
not exist). The status message indicates which parameter is not valid.
canceled-before-fulfillment

The user canceled the Spot request before it was fulfilled.

capacity-not-available

There is not enough capacity available for the instances that you requested.
constraint-not-fulfillable

The Spot request can't be fulfilled because one or more constraints are not valid (for example, the
Availability Zone does not exist). The status message indicates which constraint is not valid.
fulfilled

The Spot request is active, and Amazon EC2 is launching your Spot Instances.
instance-stopped-by-price

Your instance was stopped because the Spot price exceeded your maximum price.
instance-stopped-by-user

Your instance was stopped because a user stopped the instance or ran the shutdown command from
the instance.
instance-stopped-no-capacity

Your instance was stopped because there was no longer enough Spot capacity available for the
instance.
instance-terminated-by-price

Your instance was terminated because the Spot price exceeded your maximum price. If your request
is persistent, the process restarts, so your request is pending evaluation.
instance-terminated-by-schedule

Your Spot Instance was terminated at the end of its scheduled duration.
instance-terminated-by-service

Your instance was terminated from a stopped state.

instance-terminated-by-user or spot-instance-terminated-by-user

You terminated a Spot Instance that had been fulfilled, so the request state is closed (unless it's a
persistent request) and the instance state is terminated.
instance-terminated-launch-group-constraint

One or more of the instances in your launch group was terminated, so the launch group constraint is
no longer fulfilled.
instance-terminated-no-capacity

Your instance was terminated because there is no longer enough Spot capacity available for the
instance.
launch-group-constraint

Amazon EC2 cannot launch all the instances that you requested at the same time. All instances in a
launch group are started and terminated together.

448
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

limit-exceeded

The limit on the number of EBS volumes or total volume storage was exceeded. For more
information about these limits and how to request an increase, see Amazon EBS Limits in the
Amazon Web Services General Reference.
marked-for-stop

The Spot Instance is marked for stopping.

marked-for-termination

The Spot Instance is marked for termination.

not-scheduled-yet

The Spot request is not evaluated until the scheduled date.

pending-evaluation

After you make a Spot Instance request, it goes into the pending-evaluation state while the
system evaluates the parameters of your request.
pending-fulfillment

Amazon EC2 is trying to provision your Spot Instances.

placement-group-constraint

The Spot request can't be fulfilled yet because a Spot Instance can't be added to the placement
group at this time.
price-too-low

The request can't be fulfilled yet because your maximum price is below the Spot price. In this case,
no instance is launched and your request remains open.
request-canceled-and-instance-running

You canceled the Spot request while the Spot Instances are still running. The request is cancelled,
but the instances remain running.
schedule-expired

The Spot request expired because it was not fulfilled before the specified date.
system-error

There was an unexpected system error. If this is a recurring issue, please contact AWS Support for
assistance.

EC2 instance rebalance recommendations

An EC2 Instance rebalance recommendation is a new signal that notifies you when a Spot Instance
is at elevated risk of interruption. The signal can arrive sooner than the two-minute Spot Instance
interruption notice (p. 457), giving you the opportunity to proactively manage the Spot Instance. You
can decide to rebalance your workload to new or existing Spot Instances that are not at an elevated risk
of interruption.

It is not always possible for Amazon EC2 to send the rebalance recommendation signal before the two-
minute Spot Instance interruption notice. Therefore, the rebalance recommendation signal can arrive
along with the two-minute interruption notice.

Spot Instances with a defined duration (also known as Spot blocks) do not receive rebalance
recommendations.

449
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Note
Rebalance recommendations are only supported for Spot Instances that are launched after
November 5, 2020 00:00 UTC.

Topics
• Rebalance actions you can take (p. 450)
• Monitor rebalance recommendation signals (p. 450)
• Services that use the rebalance recommendation signal (p. 452)

Rebalance actions you can take

These are some of the possible rebalancing actions that you can take:

Graceful shutdown

When you receive the rebalance recommendation signal for a Spot Instance, you can start your
instance shutdown procedures, which might include ensuring that processes are completed before
stopping them. For example, you can upload system or application logs to Amazon Simple Storage
Service (Amazon S3), you can shut down Amazon SQS workers, or you can complete deregistration
from the Domain Name System (DNS). You can also save your work in external storage and resume it
at a later time.
Prevent new work from being scheduled

When you receive the rebalance recommendation signal for a Spot Instance, you can prevent new
work from being scheduled on the instance, while continuing to use the instance until the scheduled
work is completed.
Proactively launch new replacement instances

You can configure Auto Scaling groups, EC2 Fleet, or Spot Fleet to automatically launch replacement
Spot Instances when a rebalance recommendation signal is emitted. For more information, see
Amazon EC2 Auto Scaling Capacity Rebalancing in the Amazon EC2 Auto Scaling User Guide, and
Capacity Rebalancing (p. 554) for EC2 Fleet and Capacity Rebalancing (p. 374) for Spot Fleet in
this user guide.

Monitor rebalance recommendation signals

You can monitor the rebalance recommendation signal so that, when it is emitted, you can take the
actions that are specified in the preceding section. The rebalance recommendation signal is made
available as an event that is sent to Amazon EventBridge (formerly known as Amazon CloudWatch
Events) and as instance metadata on the Spot Instance.

Monitor rebalance recommendation signals:

• Use Amazon EventBridge (p. 450)
• Use instance metadata (p. 452)

Use Amazon EventBridge

When the rebalance recommendation signal is emitted for a Spot Instance, the event for the signal is
sent to Amazon EventBridge. If EventBridge detects an event pattern that matches a pattern defined in a
rule, EventBridge invokes a target (or targets) specified in the rule.

The following is an example event for the rebalance recommendation signal.

{
"version": "0",

450
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

"id": "12345678-1234-1234-1234-123456789012",
"detail-type": "EC2 Instance Rebalance Recommendation",
"source": "aws.ec2",
"account": "123456789012",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-2",
"resources": ["arn:aws:ec2:us-east-2:123456789012:instance/i-1234567890abcdef0"],
"detail": {
"instance-id": "i-1234567890abcdef0"
}
}

The following fields form the event pattern that is defined in the rule:

"detail-type": "EC2 Instance Rebalance Recommendation"

Identifies that the event is a rebalance recommendation event

source": "aws.ec2

Identifies that the event is from Amazon EC2

Create an EventBridge rule

You can write an EventBridge rule and automate what actions to take when the event pattern matches
the rule.

The following example creates an EventBridge rule to send an email, text message, or mobile push
notification every time Amazon EC2 emits a rebalance recommendation signal. The signal is emitted as
an EC2 Instance Rebalance Recommendation event, which triggers the action defined by the rule.

To create an EventBridge rule for a rebalance recommendation event

1. Open the Amazon EventBridge console at https://console.aws.amazon.com/events/.

2. Choose Create rule.
3. Enter a Name for the rule, and, optionally, a description.

A rule can't have the same name as another rule in the same Region and on the same event bus.
4. For Define pattern, choose Event pattern.
5. Under Event matching pattern, choose Custom pattern.
6. In the Event pattern box, add the following pattern to match the EC2 Instance Rebalance
Recommendation event, and then choose Save.

{
"source": [ "aws.ec2" ],
"detail-type": [ "EC2 Instance Rebalance Recommendation" ]
}

7. For Select event bus, choose AWS default event bus. When an AWS service in your account emits an
event, it always goes to your account's default event bus.
8. Confirm that Enable the rule on the selected event bus is toggled on.
9. For Target, choose SNS topic to send an email, text message, or mobile push notification when the
event occurs.
10. For Topic, choose an existing topic. You first need to create an Amazon SNS topic using the Amazon
SNS console. For more information, see Using Amazon SNS for application-to-person (A2P)
messaging in the Amazon Simple Notification Service Developer Guide.
11. For Configure input, choose the input for the email, text message, or mobile push notification.
12. Choose Create.

451
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

For more information, see Creating a rule for an AWS service and Event Patterns in the Amazon
EventBridge User Guide

Use instance metadata

The instance metadata category events/recommendations/rebalance provides the approximate

time, in UTC, when the rebalance recommendation signal was emitted for a Spot Instance.

We recommend that you check for rebalance recommendation signals every 5 seconds so that you don't
miss an opportunity to act on the rebalance recommendation.

If a Spot Instance receives a rebalance recommendation, the time that the signal was emitted is present
in the instance metadata. You can retrieve the time that the signal was emitted as follows.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" –v http://169.254.169.254/latest/meta-
data/events/recommendations/rebalance

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/events/recommendations/

rebalance

The following is example output, which indicates the time, in UTC, that the rebalance recommendation
signal was emitted for the Spot Instance.

{"noticeTime": "2020-10-27T08:22:00Z"}

If the signal has not been emitted for the instance, events/recommendations/rebalance is not
present and you receive an HTTP 404 error when you try to retrieve it.

Services that use the rebalance recommendation signal

Amazon EC2 Auto Scaling, EC2 Fleet, and Spot Fleet use the rebalance recommendation signal to make
it easy for you to maintain workload availability by proactively augmenting your fleet with a new Spot
Instance before a running instance receives the two-minute Spot Instance interruption notice. You can
have these services monitor and respond proactively to changes affecting the availability of your Spot
Instances. For more information, see the following:

• Amazon EC2 Auto Scaling Capacity Rebalancing in the Amazon EC2 Auto Scaling User Guide
• Capacity Rebalancing (p. 554) in the EC2 Fleet topic in this user guide
• Capacity Rebalancing (p. 374) in the Spot Fleet topic in this user guide

Spot Instance interruptions

You can launch Spot Instances on spare EC2 capacity for steep discounts in exchange for returning them
when Amazon EC2 needs the capacity back. When Amazon EC2 reclaims a Spot Instance, we call this
event a Spot Instance interruption.

Demand for Spot Instances can vary significantly from moment to moment, and the availability of Spot
Instances can also vary significantly depending on how many unused EC2 instances are available. It is

452
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

always possible that your Spot Instance might be interrupted. Therefore, you must ensure that your
application is prepared for a Spot Instance interruption.

An On-Demand Instance specified in an EC2 Fleet or Spot Fleet cannot be interrupted.

Contents
• Reasons for interruption (p. 453)
• Interruption behaviors (p. 453)
• Specify the interruption behavior (p. 455)
• Prepare for interruptions (p. 456)
• Prepare for instance hibernation (p. 456)
• Spot Instance interruption notices (p. 457)
• Find interrupted Spot Instances (p. 459)
• Determine whether Amazon EC2 interrupted a Spot Instance (p. 459)
• Billing for interrupted Spot Instances (p. 459)

Reasons for interruption

The following are the possible reasons that Amazon EC2 might interrupt your Spot Instances:

• Price – The Spot price is greater than your maximum price.

• Capacity – If there are not enough unused EC2 instances to meet the demand for On-Demand
Instances, Amazon EC2 interrupts Spot Instances. The order in which the instances are interrupted is
determined by Amazon EC2.
• Constraints – If your request includes a constraint such as a launch group or an Availability Zone group,
these Spot Instances are terminated as a group when the constraint can no longer be met.

Interruption behaviors
You can specify that Amazon EC2 should do one of the following when it interrupts a Spot Instance:

• Stop the Spot Instance

• Hibernate the Spot Instance
• Terminate the Spot Instance

The default is to terminate Spot Instances when they are interrupted. To change the interruption
behavior, see Specify the interruption behavior (p. 455).

Stop interrupted Spot Instances

You can specify the interruption behavior so that Amazon EC2 stops Spot Instances when they are
interrupted if the following requirements are met.

Requirements

• For a Spot Instance request, the type must be persistent. You cannot specify a launch group in the
Spot Instance request.
• For an EC2 Fleet or Spot Fleet request, the type must be maintain.
• The root volume must be an EBS volume, not an instance store volume.

After a Spot Instance is stopped by the Spot service, only the Spot service can restart the Spot Instance,
and the same launch specification must be used.

453
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

For a Spot Instance launched by a persistent Spot Instance request, the Spot service restarts the
stopped instance when capacity is available in the same Availability Zone and for the same instance type
as the stopped instance.

If instances in an EC2 Fleet or Spot Fleet are stopped and the fleet is of type maintain, the
Spot service launches replacement instances to maintain the target capacity. The Spot service
finds the best pools based on the specified allocation strategy (lowestPrice, diversified, or
InstancePoolsToUseCount); it does not prioritize the pool with the earlier stopped instances. Later, if
the allocation strategy leads to a pool containing the earlier stopped instances, the Spot service restarts
the stopped instances to meet the target capacity.

For example, consider a Spot Fleet with the lowestPrice allocation strategy. At initial launch, a
c3.large pool meets the lowestPrice criteria for the launch specification. Later, when the c3.large
instances are interrupted, the Spot service stops the instances and replenishes capacity from another
pool that fits the lowestPrice strategy. This time, the pool happens to be a c4.large pool and the
Spot service launches c4.large instances to meet the target capacity. Similarly, Spot Fleet could move
to a c5.large pool the next time. In each of these transitions, the Spot service does not prioritize pools
with earlier stopped instances, but rather prioritizes purely on the specified allocation strategy. The
lowestPrice strategy can lead back to pools with earlier stopped instances. For example, if instances
are interrupted in the c5.large pool and the lowestPrice strategy leads it back to the c3.large or
c4.large pools, the earlier stopped instances are restarted to fulfill target capacity.

While a Spot Instance is stopped, you can modify some of its instance attributes, but not the instance
type. If you detach or delete an EBS volume, it is not attached when the Spot Instance is started. If you
detach the root volume and the Spot service attempts to start the Spot Instance, instance start fails and
the Spot service terminates the stopped instance.

You can terminate a Spot Instance while it is stopped. If you cancel a Spot request, an EC2 Fleet, or a
Spot Fleet, the Spot service terminates any associated Spot Instances that are stopped.

While a Spot Instance is stopped, you are charged only for the EBS volumes, which are preserved. With
EC2 Fleet and Spot Fleet, if you have many stopped instances, you can exceed the limit on the number of
EBS volumes for your account.

Hibernate interrupted Spot Instances

You can specify the interruption behavior so that Amazon EC2 hibernates Spot Instances when they are
interrupted if the following requirements are met.

Requirements

• For a Spot Instance request, the type must be persistent. You cannot specify a launch group in the
Spot Instance request.
• For an EC2 Fleet or Spot Fleet request, the type must be maintain.
• The root volume must be an EBS volume, not an instance store volume, and it must be large enough to
store the instance memory (RAM) during hibernation.
• The following instances are supported: C3, C4, C5, M4, M5, R3, and R4, with less than 100 GB of
memory.
• The following operating systems are supported: Amazon Linux 2, Amazon Linux AMI, Ubuntu with an
AWS-tuned Ubuntu kernel (linux-aws) greater than 4.4.0-1041, and Windows Server 2008 R2 and later.
• Install the hibernation agent on a supported operating system, or use one of the following AMIs, which
already include the agent:
• Amazon Linux 2
• Amazon Linux AMI 2017.09.1 or later
• Ubuntu Xenial 16.04 20171121 or later
• Windows Server 2008 R2 AMI 2017.11.19 or later

454
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

• Windows Server 2012 or Windows Server 2012 R2 AMI 2017.11.19 or later

• Windows Server 2016 AMI 2017.11.19 or later
• Windows Server 2019
• Start the agent. We recommend that you use user data to start the agent on instance startup.
Alternatively, you could start the agent manually.

Recommendation

• We strongly recommend that you use an encrypted Amazon EBS volume as the root volume, because
instance memory is stored on the root volume during hibernation. This ensures that the contents
of memory (RAM) are encrypted when the data is at rest on the volume and when data is moving
between the instance and volume. Use one of the following three options to ensure that the root
volume is an encrypted Amazon EBS volume:
• EBS “single-step” encryption: In a single run-instances API call, you can launch encrypted EBS-
backed EC2 instances from an unencrypted AMI. For more information, see Use encryption with EBS-
backed AMIs (p. 161).
• EBS encryption by default: You can enable EBS encryption by default to ensure all new EBS
volumes created in your AWS account are encrypted. For more information, see Encryption by
default (p. 1178).
• Encrypted AMI: You can enable EBS encryption by using an encrypted AMI to launch your instance.
If your AMI does not have an encrypted root snapshot, you can copy it to a new AMI and request
encryption. For more information, see Encrypt an unencrypted image during copy (p. 165) and Copy
an AMI (p. 170).

When a Spot Instance is hibernated by the Spot service, the EBS volumes are preserved and instance
memory (RAM) is preserved on the root volume. The private IP addresses of the instance are also
preserved. Instance storage volumes and public IP addresses, other than Elastic IP addresses, are not
preserved. While the instance is hibernating, you are charged only for the EBS volumes. With EC2 Fleet
and Spot Fleet, if you have many hibernated instances, you can exceed the limit on the number of EBS
volumes for your account.

The agent prompts the operating system to hibernate when the instance receives a signal from the Spot
service. If the agent is not installed, the underlying operating system doesn't support hibernation, or
there isn't enough volume space to save the instance memory, hibernation fails and the Spot service
stops the instance instead.

When the Spot service hibernates a Spot Instance, you receive an interruption notice, but you do not
have two minutes before the Spot Instance is interrupted. Hibernation begins immediately. While the
instance is in the process of hibernating, instance health checks might fail. When the hibernation process
completes, the state of the instance is stopped.

Resuming a hibernated Spot Instance

After a Spot Instance is hibernated by the Spot service, it can only be resumed by the Spot service. The
Spot service resumes the instance when capacity becomes available with a Spot price that is less than
your specified maximum price.

For more information, see Prepare for instance hibernation (p. 456).

For information about hibernating On-Demand Instances, see Hibernate your Linux instance (p. 627).

Specify the interruption behavior

If you do not specify an interruption behavior, the default is to terminate Spot Instances when they are
interrupted. You can specify the interruption behavior when you create a Spot request. The way in which
you specify the interruption behavior is different depending on how you request Spot Instances.

455
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

If you request Spot Instances using the launch instance wizard (p. 526), you can specify the interruption
behavior as follows: Select the Persistent request check box and then, from Interruption behavior,
choose an interruption behavior.

If you request Spot Instances using the Spot console (p. 407), you can specify the interruption behavior
as follows: Select the Maintain target capacity check box and then, from Interruption behavior, choose
an interruption behavior.

If you configure Spot Instances in a launch template (p. 534), you can specify the interruption behavior
as follows: In the launch template, expand Advanced details and select the Request Spot Instances
check box. Choose Customize and then, from Interruption behavior, choose an interruption behavior.

If you configure Spot Instances in a launch configuration when using the request-spot-fleet CLI, you
can specify the interruption behavior as follows: For InstanceInterruptionBehavior, specify an
interruption behavior.

If you configure Spot Instances using the request-spot-instances CLI, you can specify the interruption
behavior as follows: For --instance-interruption-behavior, specify an interruption behavior.

Prepare for interruptions

Here are some best practices to follow when you use Spot Instances:

• Use the default maximum price, which is the On-Demand price.

• Ensure that your instance is ready to go as soon as the request is fulfilled by using an Amazon Machine
Image (AMI) that contains the required software configuration. You can also use user data to run
commands at start-up.
• Store important data regularly in a place that isn't affected when the Spot Instance terminates. For
example, you can use Amazon S3, Amazon EBS, or DynamoDB.
• Divide the work into small tasks (using a Grid, Hadoop, or queue-based architecture) or use
checkpoints so that you can save your work frequently.
• Amazon EC2 emits a rebalance recommendation signal to the Spot Instance when the instance is at an
elevated risk of interruption. You can rely on the rebalance recommendation to proactively manage
Spot Instance interruptions without having to wait for the two-minute Spot Instance interruption
notice. For more information, see EC2 instance rebalance recommendations (p. 449).
• Use the two-minute Spot Instance interruption notices to monitor the status of your Spot Instances.
For more information, see Spot Instance interruption notices (p. 457).
• While we make every effort to provide these warnings as soon as possible, it is possible that your Spot
Instance is interrupted before the warnings can be made available. Test your application to ensure that
it handles an unexpected instance interruption gracefully, even if you are monitoring for rebalance
recommendation signals and interruption notices. You can do so by running the application using an
On-Demand Instance and then terminating the On-Demand Instance yourself.

Prepare for instance hibernation

You must install a hibernation agent on your instance, unless you used an AMI that already includes the
agent. You must run the agent on instance startup, whether the agent was included in your AMI or you
installed it yourself.

The following procedures help you prepare a Linux instance. For directions to prepare a Windows
instance, see Preparing for instance hibernation in the Amazon EC2 User Guide for Windows Instances.

To prepare an Amazon Linux instance

1. Verify that your kernel supports hibernation and update the kernel if necessary.
2. If your AMI doesn't include the agent, install the agent using the following command.

456
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

sudo yum update; sudo yum install hibagent

3. Add the following to the user data:

#!/bin/bash
/usr/bin/enable-ec2-spot-hibernation

To prepare an Ubuntu instance

1. If your AMI doesn't include the agent, install the agent using the following command. The
hibernation agent is only available on Ubuntu 16.04 or later.

sudo apt-get install hibagent

2. Add the following to the user data.

#!/bin/bash
/usr/bin/enable-ec2-spot-hibernation

Spot Instance interruption notices

The best way for you to gracefully handle Spot Instance interruptions is to architect your application
to be fault-tolerant. To accomplish this, you can take advantage of Spot Instance interruption notices. A
Spot Instance interruption notice is a warning that is issued two minutes before Amazon EC2 stops or
terminates your Spot Instance. If you specify hibernation as the interruption behavior, you receive an
interruption notice, but you do not receive a two-minute warning because the hibernation process begins
immediately.

We recommend that you check for these interruption notices every 5 seconds.

The interruption notices are made available as a CloudWatch event and as items in the instance
metadata (p. 704) on the Spot Instance.

EC2 Spot Instance interruption notice

When Amazon EC2 is going to interrupt your Spot Instance, it emits an event two minutes prior to the
actual interruption (except for hibernation, which gets the interruption notice, but not two minutes in
advance, because hibernation begins immediately). This event can be detected by Amazon CloudWatch
Events. For more information about CloudWatch events, see the Amazon CloudWatch Events User Guide.
For a detailed example that walks you through how to create and use event rules, see Taking Advantage
of Amazon EC2 Spot Instance Interruption Notices.

The following is an example of the event for Spot Instance interruption. The possible values for
instance-action are hibernate, stop, and terminate.

{
"version": "0",
"id": "12345678-1234-1234-1234-123456789012",
"detail-type": "EC2 Spot Instance Interruption Warning",
"source": "aws.ec2",
"account": "123456789012",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-2",
"resources": ["arn:aws:ec2:us-east-2:123456789012:instance/i-1234567890abcdef0"],
"detail": {
"instance-id": "i-1234567890abcdef0",

457
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

"instance-action": "action"
}
}

instance-action

If your Spot Instance is marked to be stopped or terminated by the Spot service, the instance-action
item is present in your instance metadata (p. 704). Otherwise, it is not present. You can retrieve
instance-action as follows.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" –v http://169.254.169.254/latest/meta-
data/spot/instance-action

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/spot/instance-action

The instance-action item specifies the action and the approximate time, in UTC, when the action will
occur.

The following example indicates the time at which this instance will be stopped.

{"action": "stop", "time": "2017-09-18T08:22:00Z"}

The following example indicates the time at which this instance will be terminated.

{"action": "terminate", "time": "2017-09-18T08:22:00Z"}

If Amazon EC2 is not preparing to stop or terminate the instance, or if you terminated the instance
yourself, instance-action is not present and you receive an HTTP 404 error when you try to retrieve
it.

termination-time

This item is maintained for backward compatibility; you should use instance-action instead.

If your Spot Instance is marked for termination by the Spot service, the termination-time item is
present in your instance metadata. Otherwise, it is not present. You can retrieve termination-time as
follows.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"`
[ec2-user ~]$ if curl -H "X-aws-ec2-metadata-token: $TOKEN" -s http://169.254.169.254/
latest/meta-data/spot/termination-time | grep -q .*T.*Z; then echo terminated; fi

IMDSv1

[ec2-user ~]$ if curl -s http://169.254.169.254/latest/meta-data/spot/termination-time

| grep -q .*T.*Z; then echo terminated; fi

458
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

The termination-time item specifies the approximate time in UTC when the instance receives the
shutdown signal. For example:

2015-01-05T18:02:00Z

If Amazon EC2 is not preparing to terminate the instance, or if you terminated the Spot Instance yourself,
the termination-time item is either not present (so you receive an HTTP 404 error) or contains a
value that is not a time value.

If Amazon EC2 fails to terminate the instance, the request status is set to fulfilled. The
termination-time value remains in the instance metadata with the original approximate time, which
is now in the past.

Find interrupted Spot Instances

In the console, the Instances pane displays all instances, including Spot Instances. You can identify a
Spot Instance from the spot value in the Instance lifecycle column. The Instance state column indicates
whether the instance is pending, running, stopping, stopped, shutting-down, or terminated.
For a hibernated Spot Instance, the instance state is stopped.

To find an interrupted Spot Instance (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2.
In the navigation pane, choose Instances. In the top right corner, choose the settings icon ( ),
and under Attribute columns, select Instance lifecycle. For Spot Instances, Instance lifecycle is
spot.

Alternatively, in the navigation pane, choose Spot Requests. You can see both Spot Instance
requests and Spot Fleet requests. To view the IDs of the instances, select a Spot Instance request or a
Spot Fleet request and choose the Instances tab. Choose an instance ID to display the instance in the
Instances pane.
3. For each Spot Instance, you can view its state in the Instance State column.

To find interrupted Spot Instances (AWS CLI)

You can list your interrupted Spot Instances using the describe-instances command with the --filters
parameter. To list only the instance IDs in the output, add the --query parameter.

aws ec2 describe-instances \

--filters Name=instance-lifecycle,Values=spot Name=instance-state-
name,Values=terminated,stopped \
--query "Reservations[*].Instances[*].InstanceId"

Determine whether Amazon EC2 interrupted a Spot Instance

If a Spot Instance is stopped, hibernated, or terminated, you can use CloudTrail to see whether Amazon
EC2 interrupted the Spot Instance. In CloudTrail, the event name BidEvictedEvent indicates that
Amazon EC2 interrupted the Spot Instance. For more information about using CloudTrail, see Log
Amazon EC2 and Amazon EBS API calls with AWS CloudTrail (p. 805).

Billing for interrupted Spot Instances

When a Spot Instance (not in a Spot block) is interrupted, you’re charged as follows.

459
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

Who interrupts the Operating system Interrupted in the first Interrupted in any hour
Spot Instance hour after the first hour

If you stop or terminate Linux (excluding RHEL Charged for the seconds Charged for the seconds
the Spot Instance and SUSE) used used

Windows, RHEL, SUSE Charged for the full Charged for the full
hour even if you used a hours used, and
partial hour charged a full hour for
the interrupted partial
hour

If the Amazon EC2 Spot Linux (excluding RHEL No charge Charged for the seconds
service interrupts the and SUSE) used
Spot Instance
Windows, RHEL, SUSE No charge Charged for the
full hours used, but
no charge for the
interrupted partial hour

When a Spot Instance in a Spot block is interrupted, you’re charged as follows.

Who interrupts the Operating system Interrupted in the first Interrupted in any hour
Spot Instance hour after the first hour

If you stop or terminate Linux (excluding RHEL Charged for the seconds Charged for the seconds
the Spot Instance and SUSE) used used

Windows, RHEL, SUSE Charged for the full Charged for the full
hour even if you used a hours used, and
partial hour charged a full hour for
the interrupted partial
hour

If the Amazon EC2 Spot Linux (excluding RHEL No charge No charge

service interrupts the and SUSE)
Spot Instance
Windows, RHEL, SUSE No charge No charge

Spot Instance data feed

To help you understand the charges for your Spot Instances, Amazon EC2 provides a data feed that
describes your Spot Instance usage and pricing. This data feed is sent to an Amazon S3 bucket that you
specify when you subscribe to the data feed.

Data feed files arrive in your bucket typically once an hour, and each hour of usage is typically covered in
a single data file. These files are compressed (gzip) before they are delivered to your bucket. Amazon EC2
can write multiple files for a given hour of usage where files are large (for example, when file contents
for the hour exceed 50 MB before compression).
Note
If you don't have a Spot Instance running during a certain hour, you don't receive a data feed file
for that hour.

Contents
• Data feed file name and format (p. 461)

460
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

• Amazon S3 bucket requirements (p. 461)

• Subscribe to your Spot Instance data feed (p. 462)
• Describe your Spot Instance data feed (p. 462)
• Delete your Spot Instance data feed (p. 462)

Data feed file name and format

The Spot Instance data feed file name uses the following format (with the date and hour in UTC):

bucket-name.s3.amazonaws.com/optional-prefix/aws-account-id.YYYY-MM-DD-HH.n.unique-id.gz

For example, if your bucket name is my-bucket-name and your prefix is my-prefix, your file names are
similar to the following:

my-bucket-name.s3.amazonaws.com/my-prefix/111122223333.2019-03-17-20.001.pwBdGTJG.gz

For more information about bucket names, see Rules for bucket naming in the Amazon Simple Storage
Service Developer Guide.

The Spot Instance data feed files are tab-delimited. Each line in the data file corresponds to one instance
hour and contains the fields listed in the following table.

Field Description

Timestamp The timestamp used to determine the price charged for this instance usage.

UsageType The type of usage and instance type being charged for. For m1.small Spot
Instances, this field is set to SpotUsage. For all other instance types, this field is
set to SpotUsage:{instance-type}. For example, SpotUsage:c1.medium.

Operation The product being charged for. For Linux Spot Instances, this field is
set to RunInstances. For Windows Spot Instances, this field is set to
RunInstances:0002. Spot usage is grouped according to Availability Zone.

InstanceID The ID of the Spot Instance that generated this instance usage.

MyBidID The ID for the Spot Instance request that generated this instance usage.

MyMaxPrice The maximum price specified for this Spot Instance request.

MarketPrice The Spot price at the time specified in the Timestamp field.

Charge The price charged for this instance usage.

Version The version included in the data feed file name for this record.

Amazon S3 bucket requirements

When you subscribe to the data feed, you must specify an Amazon S3 bucket to store the data feed files.
Before you choose an Amazon S3 bucket for the data feed, consider the following:

• You must have FULL_CONTROL permission to the bucket, which includes permission for the
s3:GetBucketAcl and s3:PutBucketAcl actions.

If you're the bucket owner, you have this permission by default. Otherwise, the bucket owner must
grant your AWS account this permission.

461
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

• When you subscribe to a data feed, these permissions are used to update the bucket ACL to give the
AWS data feed account FULL_CONTROL permission. The AWS data feed account writes data feed files
to the bucket. If your account doesn't have the required permissions, the data feed files cannot be
written to the bucket.
Note
If you update the ACL and remove the permissions for the AWS data feed account, the data
feed files cannot be written to the bucket. You must resubscribe to the data feed to receive
the data feed files.
• Each data feed file has its own ACL (separate from the ACL for the bucket). The bucket owner
has FULL_CONTROL permission to the data files. The AWS data feed account has read and write
permissions.
• If you delete your data feed subscription, Amazon EC2 doesn't remove the read and write permissions
for the AWS data feed account on either the bucket or the data files. You must remove these
permissions yourself.

Subscribe to your Spot Instance data feed

To subscribe to your data feed, use the create-spot-datafeed-subscription command.

aws ec2 create-spot-datafeed-subscription \

--bucket my-bucket-name \
[--prefix my-prefix]

The following is example output:

{
"SpotDatafeedSubscription": {
"OwnerId": "111122223333",
"Bucket": "my-bucket-name",
"Prefix": "my-prefix",
"State": "Active"
}
}

Describe your Spot Instance data feed

To describe your data feed subscription, use the describe-spot-datafeed-subscription command.

aws ec2 describe-spot-datafeed-subscription

The following is example output:

{
"SpotDatafeedSubscription": {
"OwnerId": "123456789012",
"Prefix": "spotdata",
"Bucket": "my-s3-bucket",
"State": "Active"
}
}

Delete your Spot Instance data feed

To delete your data feed, use the delete-spot-datafeed-subscription command.

462
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

aws ec2 delete-spot-datafeed-subscription

Spot Instance limits

There is a limit on the number of running and requested Spot Instances per AWS account per Region.
Spot Instance limits are managed in terms of the number of virtual central processing units (vCPUs) that
your running Spot Instances are either using or will use pending the fulfillment of open Spot Instance
requests. If you terminate your Spot Instances but do not cancel the Spot Instance requests, the requests
count against your Spot Instance vCPU limit until Amazon EC2 detects the Spot Instance terminations
and closes the requests.

Topics
• Spot Instance limits (p. 463)
• Request a Spot Instance limit increase (p. 463)
• Monitor Spot Instance limits and usage (p. 464)
• Spot Fleet limits (p. 464)

Spot Instance limits

There are six Spot Instance limits, listed in the following table. Each limit specifies the vCPU limit for one
or more instance families. For information about the different instance families, generations, and sizes,
see Amazon EC2 Instance Types.

Spot Instance limit name Default vCPU limit

All Standard (A, C, D, H, I, M, R, T, Z) Spot Instance 1440 vCPUs

Requests

All F Spot Instance Requests 11 vCPUs

All G Spot Instance Requests 11 vCPUs

All Inf Spot Instance Requests 64 vCPUs

All P Spot Instance Requests 16 vCPUs

All X Spot Instance Requests 21 vCPUs

Note
New AWS accounts might start with limits that are lower than the limits described here. These
limits can increase over time.

With vCPU limits, you can use your limit in terms of the number of vCPUs that are required to launch
any combination of instance types that meet your changing application needs. For example, with an All
Standard Spot Instance Requests limit of 256 vCPUs, you could request 32 m5.2xlarge Spot Instances
(32 x 8 vCPUs) or 16 c5.4xlarge Spot Instances (16 x 16 vCPUs), or a combination of any Standard
Spot Instance types and sizes that total 256 vCPUs.

Request a Spot Instance limit increase

Even though Amazon EC2 automatically increases your Spot Instance limits based on your usage, you can
request a limit increase if necessary. For example, if you intend to launch more Spot Instances than your

463
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spot Instances

current limit allows, you can request a limit increase. You can also request a limit increase if you submit a
Spot Instance request and you receive the error Max spot instance count exceeded.

To request a Spot Instance limit increase

1. Open the Create case, Service limit increase form in the Support Center console at https://
console.aws.amazon.com/support/home#/case/create.
2. For Limit type, choose EC2 Spot Instances.
3. For Region, select the required Region.
4. For Primary instance type, select the Spot Instance limit for which you want to request a limit
increase.
5. For New limit value, enter the total number of vCPUs that you want to run concurrently. To
determine the total number of vCPUs that you need, see Amazon EC2 Instance Types to find the
number of vCPUs of each instance type.
6. (Conditional) You must create a separate limit request for each Spot Instance limit. To request an
increase for another Spot Instance limit, choose Add another request and repeat steps 4 and 5 in
this procedure.
7. For Use case description, enter your use case, and then choose Submit.

For more information about viewing limits and requesting a limit increase, see Amazon EC2 service
quotas (p. 1323).

Monitor Spot Instance limits and usage

You can view and manage your Spot Instance limits using the following:

• The Limits page in the Amazon EC2 console

• The Amazon EC2 Services quotas page in the Service Quotas console
• The get-service-quota AWS CLI

For more information, see Amazon EC2 service quotas (p. 1323) in the Amazon EC2 User Guide for Linux
Instances and Viewing a Service Quota in the Service Quotas User Guide.

With Amazon CloudWatch metrics integration, you can monitor EC2 usage against limits. You can
also configure alarms to warn about approaching limits. For more information, see Using Amazon
CloudWatch Alarms in the Service Quotas User Guide.

Spot Fleet limits

The usual Amazon EC2 limits apply to instances launched by a Spot Fleet or an EC2 Fleet, such as Spot
Instance limits and volume limits. In addition, the following limits apply:

• The number of active Spot Fleets and EC2 Fleets per Region: 1,000*
• The number of Spot Instance pools (unique combination of instance type and subnet): 300*
• The size of the user data in a launch specification: 16 KB*
• The target capacity per Spot Fleet or EC2 Fleet: 10,000
• The target capacity across all Spot Fleets and EC2 Fleets in a Region: 100,000
• A Spot Fleet request or an EC2 Fleet request can't span Regions.
• A Spot Fleet request or an EC2 Fleet request can't span different subnets from the same Availability
Zone.

464
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

If you need more than the default limits for target capacity, complete the AWS Support Center Create
case form to request a limit increase. For Limit type, choose EC2 Fleet, choose a Region, and then choose
Target Fleet Capacity per Fleet (in units) or Target Fleet Capacity per Region (in units), or both.

* These are hard limits. You cannot request a limit increase for these limits.

Burstable performance instances

If you launch your Spot Instances using a burstable performance instance type (p. 224), and if you plan
to use your burstable performance Spot Instances immediately and for a short duration, with no idle
time for accruing CPU credits, we recommend that you launch them in Standard mode (p. 236) to avoid
paying higher costs. If you launch burstable performance Spot Instances in Unlimited mode (p. 228)
and burst CPU immediately, you'll spend surplus credits for bursting. If you use the instance for a short
duration, the instance doesn't have time to accrue CPU credits to pay down the surplus credits, and you
are charged for the surplus credits when you terminate the instance.

Unlimited mode is suitable for burstable performance Spot Instances only if the instance runs long
enough to accrue CPU credits for bursting. Otherwise, paying for surplus credits makes burstable
performance Spot Instances more expensive than using other instances. For more information, see When
to use unlimited mode versus fixed CPU (p. 229).

Launch credits are meant to provide a productive initial launch experience for T2 instances by providing
sufficient compute resources to configure the instance. Repeated launches of T2 instances to access new
launch credits is not permitted. If you require sustained CPU, you can earn credits (by idling over some
period), use Unlimited mode (p. 228) for T2 Spot Instances, or use an instance type with dedicated CPU.

Dedicated Hosts
An Amazon EC2 Dedicated Host is a physical server with EC2 instance capacity fully dedicated to your
use. Dedicated Hosts allow you to use your existing per-socket, per-core, or per-VM software licenses,
including Windows Server, Microsoft SQL Server, SUSE, and Linux Enterprise Server.

For information about the configurations supported on Dedicated Hosts, see the Dedicated Hosts
Configuration Table.

Contents
• Differences between Dedicated Hosts and Dedicated Instances (p. 465)
• Bring your own license (p. 466)
• Dedicated Host instance capacity (p. 466)
• Dedicated Hosts restrictions (p. 467)
• Pricing and billing (p. 467)
• Work with Dedicated Hosts (p. 469)
• Work with shared Dedicated Hosts (p. 486)
• Host recovery (p. 490)
• Track configuration changes (p. 494)

Differences between Dedicated Hosts and Dedicated Instances

Dedicated Hosts and Dedicated Instances can both be used to launch Amazon EC2 instances onto
physical servers that are dedicated for your use.

There are no performance, security, or physical differences between Dedicated Instances and instances
on Dedicated Hosts. However, there are some differences between the two. The following table
highlights some of the key differences between Dedicated Hosts and Dedicated Instances:

465
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

  Dedicated Host Dedicated Instance

Billing Per-host billing Per-instance billing

Visibility of Provides visibility of the number of No visibility

sockets, cores, sockets and physical cores
and host ID

Host and Allows you to consistently deploy your Not supported

instance instances to the same physical server
affinity over time

Targeted Provides additional visibility and control Not supported

instance over how instances are placed on a
placement physical server

Automatic Supported. For more information, see Supported

instance Host recovery (p. 490).
recovery

Bring Your Supported Not supported

Own License
(BYOL)

Bring your own license

Dedicated Hosts allow you to use your existing per-socket, per-core, or per-VM software licenses. When
you bring your own license, you are responsible for managing your own licenses. However, Amazon EC2
has features that help you maintain license compliance, such as instance affinity and targeted placement.

These are the general steps to follow in order to bring your own volume licensed machine image into
Amazon EC2.

1. Verify that the license terms controlling the use of your machine images allow usage in a virtualized
cloud environment.
2. After you have verified that your machine image can be used within Amazon EC2, import it using VM
Import/Export. For information about how to import your machine image, see the VM Import/Export
User Guide.
3. After you import your machine image, you can launch instances from it onto active Dedicated Hosts in
your account.
4. When you run these instances, depending on the operating system, you might be required to activate
these instances against your own KMS server.

Note
To track how your images are used in AWS, enable host recording in AWS Config. You can use
AWS Config to record configuration changes to a Dedicated Host and use the output as a data
source for license reporting. For more information, see Track configuration changes (p. 494).

Dedicated Host instance capacity

Support for multiple instance types on the same Dedicated Host is available for the following instance
families: c5, m5, r5, c5n, r5n, and m5n. For example, when you allocate an r5 Dedicated Host, you can
use a host with 2 sockets and 48 physical cores on which you can run different instance types, such as
r5.2xlarge and r5.4xlarge. You can run any number of instances up to the core capacity associated

466
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

with the host. For example, the table below shows the different instance type combinations you can run
on a Dedicated Host.

Instance family Example instance type combinations

R5 • Example 1: 4 x r5.4xlarge + 4 x r5.2xlarge

• Example 2: 1 x r5.12xlarge + 1 x r5.4xlarge + 1 x
r5.2xlarge + 5 x r5.xlarge + 2 x r5.large

C5 • Example 1: 1 x c5.9xlarge + 2 x c5.4xlarge + 1 x

c5.xlarge
• Example 2: 4 x c5.4xlarge + 1 x c5.xlarge + 2 x
c5.large

M5 • Example 1: 4 x m5.4xlarge + 4 x m5.2xlarge

• Example 2: 1 x m5.12xlarge + 1 x m5.4xlarge + 1 x
m5.2xlarge + 5 x m5.xlarge + 2 x m5.large

Other instance families support only a single instance type on the same Dedicated Host. For more
information about the instance families and instance type configurations supported on Dedicated Hosts
see Amazon EC2 Dedicated Host Pricing.

Dedicated Hosts restrictions

Before you allocate Dedicated Hosts, take note of the following limitations and restrictions:

• To run RHEL, SUSE Linux, and SQL Server on Dedicated Hosts, you must bring your own AMIs. RHEL,
SUSE Linux, and SQL Server AMIs that are offered by AWS or that are available on AWS Marketplace
can't be used with Dedicated Hosts. For more information on how to create your own AMI, see Bring
your own license (p. 466).
• Up to two On-Demand Dedicated Hosts per instance family, per Region can be allocated. It is possible
to request a limit increase: Request to Raise Allocation Limit on Amazon EC2 Dedicated Hosts.
• The instances that run on a Dedicated Host can only be launched in a VPC.
• Auto Scaling groups are supported when using a launch template that specifies a host resource group.
For more information, see Creating a Launch Template for an Auto Scaling Group in the Amazon EC2
Auto Scaling User Guide.
• Amazon RDS instances are not supported.
• The AWS Free Usage tier is not available for Dedicated Hosts.
• Instance placement control refers to managing instance launches onto Dedicated Hosts. You cannot
launch Dedicated Hosts into placement groups.

Pricing and billing

The price for a Dedicated Host varies by payment option.

Payment Options
• On-Demand Dedicated Hosts (p. 468)
• Dedicated Host Reservations (p. 468)
• Savings Plans (p. 468)
• Pricing for Windows Server on Dedicated Hosts (p. 468)

467
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

On-Demand Dedicated Hosts

On-Demand billing is automatically activated when you allocate a Dedicated Host to your account.

The On-Demand price for a Dedicated Host varies by instance family and Region. You pay per second
(with a minimum of 60 seconds) for active Dedicated Host, regardless of the quantity or the size of
instances that you choose to launch on it. For more information about On-Demand pricing, see Amazon
EC2 Dedicated Hosts On-Demand Pricing.

You can release an On-Demand Dedicated Host at any time to stop accruing charges for it. For
information about releasing a Dedicated Host, see Release Dedicated Hosts (p. 482).

Dedicated Host Reservations

Dedicated Host Reservations provide a billing discount compared to running On-Demand Dedicated
Hosts. Reservations are available in three payment options:

• No Upfront—No Upfront Reservations provide you with a discount on your Dedicated Host usage over
a term and do not require an upfront payment. Available for a one-year term only.
• Partial Upfront—A portion of the reservation must be paid upfront and the remaining hours in the
term are billed at a discounted rate. Available in one-year and three-year terms.
• All Upfront—Provides the lowest effective price. Available in one-year and three-year terms and
covers the entire cost of the term upfront, with no additional future charges.

You must have active Dedicated Hosts in your account before you can purchase reservations. Each
reservation covers a single, specific Dedicated Host in your account. Reservations are applied to the
instance family on the host, not the instance size. If you have three Dedicated Hosts with different
instances sizes (m4.xlarge, m4.medium, and m4.large) you can associate a single m4 reservation with
all those Dedicated Hosts. The instance family and Region of the reservation must match that of the
Dedicated Hosts you want to associate it with.

When a reservation is associated with a Dedicated Host, the Dedicated Host can't be released until the
reservation's term is over.

For more information about reservation pricing, see Amazon EC2 Dedicated Hosts Pricing.

Savings Plans
Savings Plans are a flexible pricing model that offers significant savings over On-Demand Instances. With
Savings Plans, you make a commitment to a consistent amount of usage, in USD per hour, for a term of
one or three years. This provides you with the flexibility to use the Dedicated Hosts that best meet your
needs and continue to save money, instead of making a commitment to a specific Dedicated Host. For
more information, see the AWS Savings Plans User Guide.

Pricing for Windows Server on Dedicated Hosts

Subject to Microsoft licensing terms, you can bring your existing Windows Server and SQL Server licenses
to Dedicated Hosts. There is no additional charge for software usage if you choose to bring your own
licenses.

In addition, you can also use Windows Server AMIs provided by Amazon to run the latest versions of
Windows Server on Dedicated Hosts. This is common for scenarios where you have existing SQL Server
licenses eligible to run on Dedicated Hosts, but need Windows Server to run the SQL Server workload.

468
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

Windows Server AMIs provided by Amazon are supported on current generation instance types (p. 204)
only. For more information, see Amazon EC2 Dedicated Hosts Pricing.

Work with Dedicated Hosts

To use a Dedicated Host, you first allocate hosts for use in your account. You then launch instances onto
the hosts by specifying host tenancy for the instance. You must select a specific host for the instance to
launch on to, or you can allow it to launch on to any host that has auto-placement enabled and matches
its instance type. When an instance is stopped and restarted, the Host affinity setting determines
whether it's restarted on the same, or a different, host.

If you no longer need an On-Demand host, you can stop the instances running on the host, direct them
to launch on a different host, and then release the host.

Dedicated Hosts are also integrated with AWS License Manager. With License Manager, you can create a
host resource group, which is a collection of Dedicated Hosts that are managed as a single entity. When
creating a host resource group, you specify the host management preferences, such as auto-allocate and
auto-release, for the Dedicated Hosts. This allows you to launch instances onto Dedicated Hosts without
manually allocating and managing those hosts. For more information, see Host Resource Groups in the
AWS License Manager User Guide.

Contents
• Allocate Dedicated Hosts (p. 469)
• Launch instances onto a Dedicated Host (p. 472)
• Launch instances into a host resource group (p. 473)
• Understand auto-placement and affinity (p. 474)
• Modify Dedicated Host auto-placement (p. 475)
• Modify the supported instance types (p. 476)
• Modify instance tenancy and affinity (p. 478)
• View Dedicated Hosts (p. 479)
• Tag Dedicated Hosts (p. 480)
• Monitor Dedicated Hosts (p. 481)
• Release Dedicated Hosts (p. 482)
• Purchase Dedicated Host Reservations (p. 483)
• View Dedicated Host reservations (p. 485)
• Tag Dedicated Host Reservations (p. 485)

Allocate Dedicated Hosts

To begin using Dedicated Hosts, you must allocate Dedicated Hosts in your account using the Amazon
EC2 console or the command line tools. After you allocate the Dedicated Host, the Dedicated Host
capacity is made available in your account immediately and you can start launching instances onto the
Dedicated Host.

Support for multiple instance sizes of the same instance family on the same Dedicated Host is available
for the following instance families: c5, m5, r5, c5n, r5n, and m5n. Other instance families support only
one instance size on the same Dedicated Host.

Due to a hardware limitation with N-type Dedicated Hosts, such as C5n, M5n, and R5n, you cannot mix
smaller instance sizes (large, xlarge, and 2xlarge) with larger instance sizes (4xlarge, 9xlarge,
18xlarge, and .metal). If you require smaller and larger instance sizes on N-type hosts at the same
time, you must allocate separate hosts for the smaller and larger instance sizes.

469
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

You can allocate a Dedicated Host using the following methods.

New console

To allocate a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts and then choose Allocate Dedicated Host.
3. For Instance family, choose the instance family for the Dedicated Host.
4. Specify whether the Dedicated Host supports multiple instance sizes within the selected
instance family, or a specific instance type only. Do one of the following.

• To configure the Dedicated Host to support multiple instance types in the selected instance
family, for Support multiple instance types, choose Enable. Enabling this allows you to
launch different instance sizes from the same instance family onto the Dedicated Host.
For example, if you choose the m5 instance family and choose this option, you can launch
m5.xlarge and m5.4xlarge instances onto the Dedicated Host.
• To configure the Dedicated Host to support a single instance type within the selected instance
family, clear Support multiple instance types, and then for Instance type, choose the
instance type to support. This allows you to launch a single instance type on the Dedicated
Host. For example, if you choose this option and specify m5.4xlarge as the supported
instance type, you can launch only m5.4xlarge instances onto the Dedicated Host.
5. For Availability Zone, choose the Availability Zone in which to allocate the Dedicated Host.
6. To allow the Dedicated Host to accept untargeted instance launches that match its instance
type, for Instance auto-placement, choose Enable. For more information about auto-
placement, see Understand auto-placement and affinity (p. 474).
7. To enable host recovery for the Dedicated Host, for Host recovery, choose Enable. For more
information, see Host recovery (p. 490).
8. For Quantity, enter the number of Dedicated Hosts to allocate.
9. (Optional) Choose Add new tag and enter a tag key and a tag value.
10. Choose Allocate.

Old console

To allocate a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts, Allocate Dedicated Host.
3. For Instance family, choose the instance family for the Dedicated Host.
4. Specify whether the Dedicated Host supports multiple instance sizes within the selected
instance family, or a specific instance type only. Do one of the following.

• To configure the Dedicated Host to support multiple instance types in the selected instance
family, select Support multiple instance types. Enabling this allows you to launch different
instance sizes from the same instance family onto the Dedicated Host. For example, if you
choose the m5 instance family and choose this option, you can launch m5.xlarge and
m5.4xlarge instances onto the Dedicated Host. The instance family must be powered by the
Nitro System.
• To configure the Dedicated Host to support a single instance type within the selected instance
family, clear Support multiple instance types, and then for Instance type, choose the
instance type to support. This allows you to launch a single instance type on the Dedicated
Host. For example, if you choose this option and specify m5.4xlarge as the supported
instance type, you can launch only m5.4xlarge instances onto the Dedicated Host.

470
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

5. For Availability Zone, choose the Availability Zone in which to allocate the Dedicated Host.
6. To allow the Dedicated Host to accept untargeted instance launches that match its instance
type, for Instance auto-placement, choose Enable. For more information about auto-
placement, see Understand auto-placement and affinity (p. 474).
7. To enable host recovery for the Dedicated Host, for Host recovery choose Enable. For more
information, see Host recovery (p. 490).
8. For Quantity, enter the number of Dedicated Hosts to allocate.
9. (Optional) Choose Add Tag and enter a tag key and a tag value.
10. Choose Allocate host.

AWS CLI

To allocate a Dedicated Host

Use the allocate-hosts AWS CLI command. The following command allocates a Dedicated Host that
supports multiple instance types from the m5 instance family in us-east-1a Availability Zone. The
host also has host recovery enabled and it has auto-placement disabled.

aws ec2 allocate-hosts --instance-family "m5" --availability-zone "us-east-1a" --auto-

placement "off" --host-recovery "on" --quantity 1

The following command allocates a Dedicated Host that supports untargeted m4.large instance
launches in the eu-west-1a Availability Zone, enables host recovery, and applies a tag with a key of
purpose and a value of production.

aws ec2 allocate-hosts --instance-type "m4.large" --availability-zone "eu-west-1a"

--auto-placement "on" --host-recovery "on" --quantity 1 --tag-specifications
'ResourceType=dedicated-host,Tags=[{Key=purpose,Value=production}]'

PowerShell

To allocate a Dedicated Host

Use the New-EC2Host AWS Tools for Windows PowerShell command. The following command
allocates a Dedicated Host that supports multiple instance types from the m5 instance family in
us-east-1a Availability Zone. The host also has host recovery enabled and it has auto-placement
disabled.

PS C:\> New-EC2Host -InstanceFamily m5 -AvailabilityZone us-east-1a -AutoPlacement Off

-HostRecovery On -Quantity 1

The following commands allocate a Dedicated Host that supports untargeted m4.large instance
launches in the eu-west-1a Availability Zone, enable host recovery, and apply a tag with a key of
purpose and a value of production.

The TagSpecification parameter used to tag a Dedicated Host on creation requires an object
that specifies the type of resource to be tagged, the tag key, and the tag value. The following
commands create the required object.

PS C:\> $tag = @{ Key="purpose"; Value="production" }

PS C:\> $tagspec = new-object Amazon.EC2.Model.TagSpecification
PS C:\> $tagspec.ResourceType = "dedicated-host"
PS C:\> $tagspec.Tags.Add($tag)

471
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

The following command allocates the Dedicated Host and applies the tag specified in the $tagspec
object.

PS C:\> New-EC2Host -InstanceType m4.large -AvailabilityZone eu-west-1a -

AutoPlacement On -HostRecovery On -Quantity 1 -TagSpecification $tagspec

Launch instances onto a Dedicated Host

After you have allocated a Dedicated Host, you can launch instances onto it. You can't launch instances
with host tenancy if you do not have active Dedicated Hosts with enough available capacity for the
instance type that you are launching.
Note
The instances launched onto Dedicated Hosts can only be launched in a VPC. For more
information, see Introduction to VPC.

Before you launch your instances, take note of the limitations. For more information, see Dedicated
Hosts restrictions (p. 467).

You can launch an instance onto a Dedicated Host using the following methods.

Console

To launch an instance onto a specific Dedicated Host from the Dedicated Hosts page

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Dedicated Hosts in the navigation pane.
3. On the Dedicated Hosts page, select a host and choose Actions, Launch Instance(s) onto Host.
4. Select an AMI from the list. SQL Server, SUSE, and RHEL AMIs provided by Amazon EC2 can't be
used with Dedicated Hosts.
5. On the Choose an Instance Type page, select the instance type to launch and then choose Next:
Configure Instance Details.

If the Dedicated Host supports a single instance type only, the supported instance type is
selected by default and can't be changed.

If the Dedicated Host supports multiple instance types, you must select an instance type within
the supported instance family based on the available instance capacity of the Dedicated Host.
We recommend that you launch the larger instance sizes first, and then fill the remaining
instance capacity with the smaller instance sizes as needed.
6. On the Configure Instance Details page, configure the instance settings to suit your needs, and
then for Affinity, choose one of the following options:

• Off—The instance launches onto the specified host, but it is not guaranteed to restart on the
same Dedicated Host if stopped.
• Host—If stopped, the instance always restarts on this specific host.

For more information about Affinity, see Understand auto-placement and affinity (p. 474).

The Tenancy and Host options are pre-configured based on the host that you selected.
7. Choose Review and Launch.
8. On the Review Instance Launch page, choose Launch.
9. When prompted, select an existing key pair or create a new one, and then choose Launch
Instances.

472
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

To launch an instance onto a Dedicated Host using the Launch Instance wizard

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, Launch Instance.
3. Select an AMI from the list. SQL Server, SUSE, and RHEL AMIs provided by Amazon EC2 can't be
used with Dedicated Hosts.
4. Select the type of instance to launch and choose Next: Configure Instance Details.
5. On the Configure Instance Details page, configure the instance settings to suit your needs, and
then configure the following settings, which are specific to a Dedicated Host:

• Tenancy—Choose Dedicated Host - Launch this instance on a Dedicated Host.

• Host—Choose either Use auto-placement to launch the instance on any Dedicated Host that
has auto-placement enabled, or select a specific Dedicated Host in the list. The list displays
only Dedicated Hosts that support the selected instance type.
• Affinity—Choose one of the following options:
• Off—The instance launches onto the specified host, but it is not guaranteed to restart on it
if stopped.
• Host—If stopped, the instance always restarts on the specified host.

For more information, see Understand auto-placement and affinity (p. 474).

If you are unable to see these settings, check that you have selected a VPC in the Network
menu.
6. Choose Review and Launch.
7. On the Review Instance Launch page, choose Launch.
8. When prompted, select an existing key pair or create a new one, and then choose Launch
Instances.

AWS CLI

To launch an instance onto a Dedicated Host

Use the run-instances AWS CLI command and specify the instance affinity, tenancy, and host in the
Placement request parameter.
PowerShell

To launch an instance onto a Dedicated Host

Use the New-EC2Instance AWS Tools for Windows PowerShell command and specify the instance
affinity, tenancy, and host in the Placement request parameter.

Launch instances into a host resource group

When you launch an instance into a host resource group that has a Dedicated Host with available
instance capacity, Amazon EC2 launches the instance onto that host. If the host resource group does not
have a host with available instance capacity, Amazon EC2 automatically allocates a new host in the host
resource group, and then launches the instance onto that host. For more information, see Host Resource
Groups in the AWS License Manager User Guide.

Requirements and limits

• You must associate a core- or socket-based license configuration with the AMI.

473
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

• You can't use SQL Server, SUSE, or RHEL AMIs provided by Amazon EC2 with Dedicated Hosts.
• You can't target a specific host by choosing a host ID, and you can't enable instance affinity when
launching an instance into a host resource group.

You can launch an instance into a host resource group using the following methods.

Console

To launch an instance into a host resource group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, Launch Instance.
3. Select an AMI.
4. Select the type of instance to launch and choose Next: Configure Instance Details.
5. On the Configure Instance Details page, configure the instance settings to suit your needs, and
then do the following:

a. For Tenancy, choose Dedicated Host.

b. For Host resource group, choose Launch instance into a host resource group.
c. For Host resource group name, choose the host resource group in which to launch the
instance.
6. Choose Review and Launch.
7. On the Review Instance Launch page, choose Launch.
8. When prompted, select an existing key pair or create a new one, and then choose Launch
Instances.

AWS CLI

To launch an instance into a host resource group

Use the run-instances AWS CLI command, and in the Placement request parameter, omit the
Tenancy option and specify the host resource group ARN.
PowerShell

To launch an instance into a host resource group

Use the New-EC2Instance AWS Tools for Windows PowerShell command, and in the Placement
request parameter, omit the Tenancy option and specify the host resource group ARN.

Understand auto-placement and affinity

Placement control for Dedicated Hosts happens on both the instance level and host level.

Auto-placement

Auto-placement is configured at the host level. It allows you to manage whether instances that you
launch are launched onto a specific host, or onto any available host that has matching configurations.

When the auto-placement of a Dedicated Host is disabled, it only accepts Host tenancy instance launches
that specify its unique host ID. This is the default setting for new Dedicated Hosts.

When the auto-placement of a Dedicated Host is enabled, it accepts any untargeted instance launches
that match its instance type configuration.

474
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

When launching an instance, you need to configure its tenancy. Launching an instance onto a Dedicated
Host without providing a specific HostId enables it to launch on any Dedicated Host that has auto-
placement enabled and that matches its instance type.

Host affinity

Host affinity is configured at the instance level. It establishes a launch relationship between an instance
and a Dedicated Host.

When affinity is set to Host, an instance launched onto a specific host always restarts on the same host
if stopped. This applies to both targeted and untargeted launches.

When affinity is set to Off, and you stop and restart the instance, it can be restarted on any available
host. However, it tries to launch back onto the last Dedicated Host on which it ran (on a best-effort
basis).

Modify Dedicated Host auto-placement

You can modify the auto-placement settings of a Dedicated Host after you have allocated it to your AWS
account, using one of the following methods.

New console

To modify the auto-placement of a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. Select a host and choose Actions, Modify host.
4. For Instance auto-placement, choose Enable to enable auto-placement, or clear Enable
to disable auto-placement. For more information, see Understand auto-placement and
affinity (p. 474).
5. Choose Save.

Old console

To modify the auto-placement of a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Dedicated Hosts in the navigation pane.
3. On the Dedicated Hosts page, select a host and choose Actions, Modify Auto-Placement.
4. On the Modify Auto-placement window, for Allow instance auto-placement, choose Yes to
enable auto-placement, or choose No to disable auto-placement. For more information, see
Understand auto-placement and affinity (p. 474).
5. Choose Save.

AWS CLI

To modify the auto-placement of a Dedicated Host

Use the modify-hosts AWS CLI command. The following example enables auto-placement for the
specified Dedicated Host.

aws ec2 modify-hosts --auto-placement on --host-ids h-012a3456b7890cdef

475
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

PowerShell

To modify the auto-placement of a Dedicated Host

Use the Edit-EC2Host AWS Tools for Windows PowerShell command. The following example enables
auto-placement for the specified Dedicated Host.

PS C:\> Edit-EC2Host --AutoPlacement 1 --HostId h-012a3456b7890cdef

Modify the supported instance types

Support for multiple instance types on the same Dedicated Host is available for the following instance
families: c5, m5, r5, c5n, r5n, and m5n. Other instance families support only a single instance type on
the same Dedicated Host.

You can allocate a Dedicated Host using the following methods.

You can modify a Dedicated Host to change the instance types that it supports. If it currently supports
a single instance type, you can modify it to support multiple instance types within that instance family.
Similarly, if it currently supports multiple instance types, you can modify it to support a specific instance
type only.

To modify a Dedicated Host to support multiple instance types, you must first stop all running instances
on the host. The modification takes approximately 10 minutes to complete. The Dedicated Host
transitions to the pending state while the modification is in progress. You can't start stopped instances
or launch new instances on the Dedicated Host while it is in the pending state.

To modify a Dedicated Host that supports multiple instance types to support only a single instance type,
the host must either have no running instances, or the running instances must be of the instance type
that you want the host to support. For example, to modify a host that supports multiple instance types
in the m5 instance family to support only m5.large instances, the Dedicated Host must either have no
running instances, or it must have only m5.large instances running on it.

You can modify the supported instance types using one of the following methods.

New console

To modify the supported instance types for a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the Navigation pane, choose Dedicated Host.
3. Select the Dedicated Host to modify and choose Actions, Modify host.
4. Do one of the following, depending on the current configuration of the Dedicated Host:

• If the Dedicated Host currently supports a specific instance type, Support multiple instance
types is not enabled, and Instance type lists the supported instance type. To modify the host
to support multiple types in the current instance family, for Support multiple instance types,
choose Enable.

You must first stop all instances running on the host before modifying it to support multiple
instance types.
• If the Dedicated Host currently supports multiple instance types in an instance family,
Enabled is selected for Support multiple instance types. To modify the host to support
a specific instance type, for Support multiple instance types, clear Enable, and then for
Instance type, select the specific instance type to support.

You can't change the instance family supported by the Dedicated Host.

476
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

5. Choose Save.

Old console

To modify the supported instance types for a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the Navigation pane, choose Dedicated Host.
3. Select the Dedicated Host to modify and choose Actions, Modify Supported Instance Types.
4. Do one of the following, depending on the current configuration of the Dedicated Host:

• If the Dedicated Host currently supports a specific instance type, No is selected for Support
multiple instance types. To modify the host to support multiple types in the current instance
family, for Support multiple instance types, select Yes.

You must first stop all instances running on the host before modifying it to support multiple
instance types.
• If the Dedicated Host currently supports multiple instance types in an instance family, Yes is
selected for Support multiple instance types, and Instance family displays the supported
instance family. To modify the host to support a specific instance type, for Support multiple
instance types, select No, and then for Instance type, select the specific instance type to
support.

You can't change the instance family supported by the Dedicated Host.
5. Choose Save.

AWS CLI

To modify the supported instance types for a Dedicated Host

Use the modify-hosts AWS CLI command.

The following command modifies a Dedicated Host to support multiple instance types within the m5
instance family.

aws ec2 modify-hosts --instance-family m5 --host-ids h-012a3456b7890cdef

The following command modifies a Dedicated Host to support m5.xlarge instances only.

aws ec2 modify-hosts --instance-type m5.xlarge --instance-family --host-

ids h-012a3456b7890cdef

PowerShell

To modify the supported instance types for a Dedicated Host

Use the Edit-EC2Host AWS Tools for Windows PowerShell command.

The following command modifies a Dedicated Host to support multiple instance types within the m5
instance family.

PS C:\> Edit-EC2Host --InstanceFamily m5 --HostId h-012a3456b7890cdef

The following command modifies a Dedicated Host to support m5.xlarge instances only.

PS C:\> Edit-EC2Host --InstanceType m5.xlarge --HostId h-012a3456b7890cdef

477
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

Modify instance tenancy and affinity

You can change the tenancy of an instance from dedicated to host, or from host to dedicated, after
you have launched it. You can also modify the affinity between the instance and the host. To modify
either instance tenancy or affinity, the instance must be in the stopped state.

You can modify an instance's tenancy and affinity using the following methods.

Console

To modify instance tenancy or affinity

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Instances, and select the instance to modify.
3. Choose Instance state, Stop.
4. Open the context (right-click) menu on the instance and choose Instance Settings, Modify
Instance Placement.
5. On the Modify Instance Placement page, configure the following:

• Tenancy—Choose one of the following:

• Run a dedicated hardware instance—Launches the instance as a Dedicated Instance. For
more information, see Dedicated Instances (p. 496).
• Launch the instance on a Dedicated Host—Launches the instance onto a Dedicated Host
with configurable affinity.
• Affinity—Choose one of the following:
• This instance can run on any one of my hosts—The instance launches onto any available
Dedicated Host in your account that supports its instance type.
• This instance can only run on the selected host—The instance is only able to run on the
Dedicated Host selected for Target Host.
• Target Host—Select the Dedicated Host that the instance must run on. If no target host is
listed, you might not have available, compatible Dedicated Hosts in your account.

For more information, see Understand auto-placement and affinity (p. 474).
6. Choose Save.

AWS CLI

To modify instance tenancy or affinity

Use the modify-instance-placement AWS CLI command. The following example changes the
specified instance's affinity from default to host, and specifies the Dedicated Host that the
instance has affinity with.

aws ec2 modify-instance-placement --instance-id i-1234567890abcdef0 --affinity host --

host-id h-012a3456b7890cdef

PowerShell

To modify instance tenancy or affinity

Use the Edit-EC2InstancePlacement AWS Tools for Windows PowerShell command. The following
example changes the specified instance's affinity from default to host, and specifies the
Dedicated Host that the instance has affinity with.

478
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

PS C:\> Edit-EC2InstancePlacement -InstanceId i-1234567890abcdef0 -Affinity host -

HostId h-012a3456b7890cdef

View Dedicated Hosts

You can view details about a Dedicated Host and the individual instances on it using the following
methods.

New console

To view the details of a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. On the Dedicated Hosts page, select a host.
4. For information about the host, choose Details.

Available vCPUs indicates the vCPUs that are available on the Dedicated Host for new instance
launches. For example, a Dedicated Host that supports multiple instance types within the c5
instance family, and that has no instances running on it, has 72 available vCPUs. This means that
you can launch different combinations of instance types onto the Dedicated Host to consume
the 72 available vCPUs.

For information about instances running on the host, choose Running instances.

Old console

To view the details of a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. On the Dedicated Hosts page, select a host.
4. For information about the host, choose Description. Available vCPUs indicates the vCPUs that
are available on the Dedicated Host for new instance launches. For example, a Dedicated Host
that supports multiple instance types within the c5 instance family, and that has no instances
running on it, has 72 available vCPUs. This means that you can launch different combinations of
instance types onto the Dedicated Host to consume the 72 available vCPUs.

For information about instances running on the host, choose Instances.

AWS CLI

To view the capacity of a Dedicated Host

Use the describe-hosts AWS CLI command.

The following example uses the describe-hosts (AWS CLI) command to view the available instance
capacity for a Dedicated Host that supports multiple instance types within the c5 instance family.
The Dedicated Host already has two c5.4xlarge instances and four c5.2xlarge instances running
on it.

$ aws ec2 describe-hosts --host-id h-012a3456b7890cdef

"AvailableInstanceCapacity": [

479
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

{ "AvailableCapacity": 2,
"InstanceType": "c5.xlarge",
"TotalCapacity": 18 },
{ "AvailableCapacity": 4,
"InstanceType": "c5.large",
"TotalCapacity": 36 }
],
"AvailableVCpus": 8

PowerShell

To view the instance capacity of a Dedicated Host

Use the Get-EC2Host AWS Tools for Windows PowerShell command.

PS C:\> Get-EC2Host -HostId h-012a3456b7890cdef

Tag Dedicated Hosts

You can assign custom tags to your existing Dedicated Hosts to categorize them in different ways, for
example, by purpose, owner, or environment. This helps you to quickly find a specific Dedicated Host
based on the custom tags that you assigned. Dedicated Host tags can also be used for cost allocation
tracking.

You can also apply tags to Dedicated Hosts at the time of creation. For more information, see Allocate
Dedicated Hosts (p. 469).

You can tag a Dedicated Host using the following methods.

New console

To tag a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. Select the Dedicated Host to tag, and then choose Actions, Manage tags.
4. In the Manage tags screen, choose Add tag, and then specify the key and value for the tag.
5. (Optional) Choose Add tag to add additional tags to the Dedicated Host.
6. Choose Save changes.

Old console

To tag a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. Select the Dedicated Host to tag, and then choose Tags.
4. Choose Add/Edit Tags.
5. In the Add/Edit Tags dialog box, choose Create Tag, and then specify the key and value for the
tag.
6. (Optional) Choose Create Tag to add additional tags to the Dedicated Host.
7. Choose Save.

480
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

AWS CLI

To tag a Dedicated Host

Use the create-tags AWS CLI command.

The following command tags the specified Dedicated Host with Owner=TeamA.

aws ec2 create-tags --resources h-abc12345678909876 --tags Key=Owner,Value=TeamA

PowerShell

To tag a Dedicated Host

Use the New-EC2Tag AWS Tools for Windows PowerShell command.

The New-EC2Tag command needs a Tag object, which specifies the key and value pair to be used
for the Dedicated Host tag. The following commands create a Tag object named $tag, with a key
and value pair of Owner and TeamA respectively.

PS C:\> $tag = New-Object Amazon.EC2.Model.Tag

PS C:\> $tag.Key = "Owner"
PS C:\> $tag.Value = "TeamA"

The following command tags the specified Dedicated Host with the $tag object.

PS C:\> New-EC2Tag -Resource h-abc12345678909876 -Tag $tag

Monitor Dedicated Hosts

Amazon EC2 constantly monitors the state of your Dedicated Hosts. Updates are communicated on the
Amazon EC2 console. You can view information about a Dedicated Host using the following methods.

Console

To view the state of a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. Locate the Dedicated Host in the list and review the value in the State column.

AWS CLI

To view the state of a Dedicated Host

Use the describe-hosts AWS CLI command and then review the state property in the hostSet
response element.

aws ec2 describe-hosts --host-id h-012a3456b7890cdef

PowerShell

To view the state of a Dedicated Host

Use the Get-EC2Host AWS Tools for Windows PowerShell command and then review the state
property in the hostSet response element.

481
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

PS C:\> Get-EC2Host -HostId h-012a3456b7890cdef

The following table explains the possible Dedicated Host states.

State Description

available AWS hasn't detected an issue with the Dedicated Host. No maintenance or
repairs are scheduled. Instances can be launched onto this Dedicated Host.

released The Dedicated Host has been released. The host ID is no longer in use.
Released hosts can't be reused.

under-assessment AWS is exploring a possible issue with the Dedicated Host. If action must be
taken, you are notified via the AWS Management Console or email. Instances
can't be launched onto a Dedicated Host in this state.

pending The Dedicated Host cannot be used for new instance launches. It is either
being modified to support multiple instance types (p. 476), or a host
recovery (p. 490) is in progress.

permanent-failure An unrecoverable failure has been detected. You receive an eviction notice
through your instances and by email. Your instances might continue to run.
If you stop or terminate all instances on a Dedicated Host with this state,
AWS retires the host. AWS does not restart instances in this state. Instances
can't be launched onto Dedicated Hosts in this state.

released- AWS permanently releases Dedicated Hosts that have failed and no longer
permanent-failure have running instances on them. The Dedicated Host ID is no longer
available for use.

Release Dedicated Hosts

Any running instances on the Dedicated Host must be stopped before you can release the host. These
instances can be migrated to other Dedicated Hosts in your account so that you can continue to use
them. These steps apply only to On-Demand Dedicated Hosts.

You can release a Dedicated Host using the following methods.

New console

To release a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. On the Dedicated Hosts page, select the Dedicated Host to release.
4. Choose Actions, Release host.
5. To confirm, choose Release.

Old console

To release a Dedicated Host

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Dedicated Hosts in the navigation pane.

482
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

3. On the Dedicated Hosts page, select the Dedicated Host to release.

4. Choose Actions, Release Hosts.
5. Choose Release to confirm.

AWS CLI

To release a Dedicated Host

Use the release-hosts AWS CLI command.

aws ec2 release-hosts --host-ids h-012a3456b7890cdef

PowerShell

To release a Dedicated Host

Use the Remove-EC2Hosts AWS Tools for Windows PowerShell command.

PS C:\> Remove-EC2Hosts -HostId h-012a3456b7890cdef

After you release a Dedicated Host, you can't reuse the same host or host ID again, and you are no longer
charged On-Demand billing rates for it. The state of the Dedicated Host is changed to released, and
you are not able to launch any instances onto that host.
Note
If you have recently released Dedicated Hosts, it can take some time for them to stop counting
towards your limit. During this time, you might experience LimitExceeded errors when trying
to allocate new Dedicated Hosts. If this is the case, try allocating new hosts again after a few
minutes.

The instances that were stopped are still available for use and are listed on the Instances page. They
retain their host tenancy setting.

Purchase Dedicated Host Reservations

You can purchase reservations using the following methods:

Console

To purchase reservations

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Dedicated Hosts, Dedicated Host Reservations, Purchase Dedicated Host Reservation.
3. On the Purchase Dedicated Host Reservation screen, you can search for available offerings
using the default settings, or you can specify custom values for the following:

• Host instance family—The options listed correspond with the Dedicated Hosts in your
account that are not already assigned to a reservation.
• Availability Zone—The Availability Zone of the Dedicated Hosts in your account that aren't
already assigned to a reservation.
• Payment option—The payment option for the offering.
• Term—The term of the reservation, which can be one or three years.
4. Choose Find offering and select an offering that matches your requirements.

483
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

5. Choose the Dedicated Hosts to associate with the reservation, and then choose Review.
6. Review your order and choose Order.

AWS CLI

To purchase reservations

1. Use the describe-host-reservation-offerings AWS CLI command to list the available offerings
that match your needs. The following example lists the offerings that support instances in the
m4 instance family and have a one-year term.
Note
The term is specified in seconds. A one-year term includes 31,536,000 seconds, and a
three-year term includes 94,608,000 seconds.

aws ec2 describe-host-reservation-offerings --filter Name=instance-family,Values=m4

--max-duration 31536000

The command returns a list of offerings that match your criteria. Note the offeringId of the
offering to purchase.
2. Use the purchase-host-reservation AWS CLI command to purchase the offering and provide
the offeringId noted in the previous step. The following example purchases the specified
reservation and associates it with a specific Dedicated Host that is already allocated in the AWS
account, and it applies a tag with a key of purpose and a value of production.

aws ec2 purchase-host-reservation --offering-id hro-03f707bf363b6b324 --

host-id-set h-013abcd2a00cbd123 --tag-specifications 'ResourceType=host-
reservation,Tags={Key=purpose,Value=production}'

PowerShell

To purchase reservations

1. Use the Get-EC2HostReservationOffering AWS Tools for Windows PowerShell command to list
the available offerings that match your needs. The following examples list the offerings that
support instances in the m4 instance family and have a one-year term.
Note
The term is specified in seconds. A one-year term includes 31,536,000 seconds, and a
three-year term includes 94,608,000 seconds.

PS C:\> $filter = @{Name="instance-family"; Value="m4"}

PS C:\> Get-EC2HostReservationOffering -filter $filter -MaxDuration 31536000

The command returns a list of offerings that match your criteria. Note the offeringId of the
offering to purchase.
2. Use the New-EC2HostReservation AWS Tools for Windows PowerShell command to purchase
the offering and provide the offeringId noted in the previous step. The following example
purchases the specified reservation and associates it with a specific Dedicated Host that is
already allocated in the AWS account.

PS C:\> New-EC2HostReservation -OfferingId hro-03f707bf363b6b324 -

HostIdSet h-013abcd2a00cbd123

484
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

View Dedicated Host reservations

You can view information about the Dedicated Hosts that are associated with your reservation, including:

• The term of the reservation

• The payment option
• The start and end dates

You can view details of your Dedicated Host reservations using the following methods.

Console

To view the details of a Dedicated Host reservation

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Dedicated Hosts in the navigation pane.
3. On the Dedicated Hosts page, choose Dedicated Host Reservations, and then select the
reservation from the list provided.
4. Choose Details for information about the reservation.
5. Choose Hosts for information about the Dedicated Hosts with which the reservation is
associated.

AWS CLI

To view the details of a Dedicated Host reservation

Use the describe-host-reservations AWS CLI command.

aws ec2 describe-host-reservations

PowerShell

To view the details of a Dedicated Host reservation

Use the Get-EC2HostReservation AWS Tools for Windows PowerShell command.

PS C:\> Get-EC2HostReservation

Tag Dedicated Host Reservations

You can assign custom tags to your Dedicated Host Reservations to categorize them in different ways,
for example, by purpose, owner, or environment. This helps you to quickly find a specific Dedicated Host
Reservation based on the custom tags that you assigned.

You can tag a Dedicated Host Reservation using the command line tools only.

AWS CLI

To tag a Dedicated Host Reservation

Use the create-tags AWS CLI command.

aws ec2 create-tags --resources hr-1234563a4ffc669ae --tags Key=Owner,Value=TeamA

485
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

PowerShell

To tag a Dedicated Host Reservation

Use the New-EC2Tag AWS Tools for Windows PowerShell command.

The New-EC2Tag command needs a Tag parameter, which specifies the key and value pair to be
used for the Dedicated Host Reservation tag. The following commands create the Tag parameter.

PS C:\> $tag = New-Object Amazon.EC2.Model.Tag

PS C:\> $tag.Key = "Owner"
PS C:\> $tag.Value = "TeamA"

PS C:\> New-EC2Tag -Resource hr-1234563a4ffc669ae -Tag $tag

Work with shared Dedicated Hosts

Dedicated Host sharing enables Dedicated Host owners to share their Dedicated Hosts with other
AWS accounts or within an AWS organization. This enables you to create and manage Dedicated Hosts
centrally, and share the Dedicated Host across multiple AWS accounts or within your AWS organization.

In this model, the AWS account that owns the Dedicated Host (owner) shares it with other AWS accounts
(consumers). Consumers can launch instances onto Dedicated Hosts that are shared with them in
the same way that they would launch instances onto Dedicated Hosts that they allocate in their own
account. The owner is responsible for managing the Dedicated Host and the instances that they launch
onto it. Owners can't modify instances that consumers launch onto shared Dedicated Hosts. Consumers
are responsible for managing the instances that they launch onto Dedicated Hosts shared with them.
Consumers can't view or modify instances owned by other consumers or by the Dedicated Host owner,
and they can't modify Dedicated Hosts that are shared with them.

A Dedicated Host owner can share a Dedicated Host with:

• Specific AWS accounts inside or outside of its AWS organization

• An organizational unit inside its AWS organization
• Its entire AWS organization

Contents
• Prerequisites for sharing Dedicated Hosts (p. 487)
• Limitations for sharing Dedicated Hosts (p. 487)
• Related services (p. 487)
• Share across Availability Zones (p. 487)
• Share a Dedicated Host (p. 487)
• Unshare a shared Dedicated Host (p. 488)
• Identify a shared Dedicated Host (p. 489)
• View instances running on a shared Dedicated Host (p. 489)
• Shared Dedicated Host permissions (p. 490)
• Billing and metering (p. 490)
• Dedicated Host limits (p. 490)
• Host recovery and Dedicated Host sharing (p. 490)

486
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

Prerequisites for sharing Dedicated Hosts

• To share a Dedicated Host, you must own it in your AWS account. You can't share a Dedicated Host that
has been shared with you.
• To share a Dedicated Host with your AWS organization or an organizational unit in your AWS
organization, you must enable sharing with AWS Organizations. For more information, see Enable
Sharing with AWS Organizations in the AWS RAM User Guide.

Limitations for sharing Dedicated Hosts

You can't share Dedicated Hosts that have been allocated for the following instance types:
u-6tb1.metal, u-9tb1.metal, u-12tb1.metal, u-18tb1.metal, and u-24tb1.metal.

Related services
AWS Resource Access Manager
Dedicated Host sharing integrates with AWS Resource Access Manager (AWS RAM). AWS RAM is a service
that enables you to share your AWS resources with any AWS account or through AWS Organizations.
With AWS RAM, you share resources that you own by creating a resource share. A resource share specifies
the resources to share, and the consumers with whom to share them. Consumers can be individual AWS
accounts, or organizational units or an entire organization from AWS Organizations.

For more information about AWS RAM, see the AWS RAM User Guide.

Share across Availability Zones

To ensure that resources are distributed across the Availability Zones for a Region, we independently
map Availability Zones to names for each account. This could lead to Availability Zone naming
differences across accounts. For example, the Availability Zone us-east-1a for your AWS account might
not have the same location as us-east-1a for another AWS account.

To identify the location of your Dedicated Hosts relative to your accounts, you must use the Availability
Zone ID (AZ ID). The Availability Zone ID is a unique and consistent identifier for an Availability Zone
across all AWS accounts. For example, use1-az1 is an Availability Zone ID for the us-east-1 Region
and it is the same location in every AWS account.

To view the Availability Zone IDs for the Availability Zones in your account

1. Open the AWS RAM console at https://console.aws.amazon.com/ram.

2. The Availability Zone IDs for the current Region are displayed in the Your AZ ID panel on the right-
hand side of the screen.

Share a Dedicated Host

When an owner shares a Dedicated Host, it enables consumers to launch instances on the host.
Consumers can launch as many instances onto the shared host as its available capacity allows.
Important
Note that you are responsible for ensuring that you have appropriate license rights to share any
BYOL licenses on your Dedicated Hosts.

If you share a Dedicated Host with auto-placement enabled, keep the following in mind as it could lead
to unintended Dedicated Host usage:

• If consumers launch instances with Dedicated Host tenancy and they do not have capacity on a
Dedicated Host that they own in their account, the instance is automatically launched onto the shared
Dedicated Host.

487
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

To share a Dedicated Host, you must add it to a resource share. A resource share is an AWS RAM resource
that lets you share your resources across AWS accounts. A resource share specifies the resources to share,
and the consumers with whom they are shared. You can add the Dedicated Host to an existing resource,
or you can add it to a new resource share.

If you are part of an organization in AWS Organizations and sharing within your organization is enabled,
consumers in your organization are automatically granted access to the shared Dedicated Host.
Otherwise, consumers receive an invitation to join the resource share and are granted access to the
shared Dedicated Host after accepting the invitation.
Note
After you share a Dedicated Host, it could take a few minutes for consumers to have access to it.

You can share a Dedicated Host that you own by using one of the following methods.

Amazon EC2 console

To share a Dedicated Host that you own using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. Choose the Dedicated Host to share and choose Actions, Share host.
4. Select the resource share to which to add the Dedicated Host and choose Share host.

It could take a few minutes for consumers to get access to the shared host.

AWS RAM console

To share a Dedicated Host that you own using the AWS RAM console

See Creating a Resource Share in the AWS RAM User Guide.

AWS CLI

To share a Dedicated Host that you own using the AWS CLI

Use the create-resource-share command.

Unshare a shared Dedicated Host

The Dedicated Host owner can unshare a shared Dedicated Host at any time. When you unshare a shared
Dedicated Host, the following rules apply:

• Consumers with whom the Dedicated Host was shared can no longer launch new instances onto it.
• Instances owned by consumers that were running on the Dedicated Host at the time of unsharing
continue to run but are scheduled for retirement. Consumers receive retirement notifications for
the instances and they have two weeks to take action on the notifications. However, if the Dedicated
Host is reshared with the consumer within the retirement notice period, the instance retirements are
cancelled.

To unshare a shared Dedicated Host that you own, you must remove it from the resource share. You can
do this by using one of the following methods.

Amazon EC2 console

To unshare a shared Dedicated Host that you own using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

488
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

2. In the navigation pane, choose Dedicated Hosts.

3. Choose the Dedicated Host to unshare and choose the Sharing tab.
4. The Sharing tab lists the resource shares to which the Dedicated Host has been added. Select
the resource share from which to remove the Dedicated Host and choose Remove host from
resource share.

AWS RAM console

To unshare a shared Dedicated Host that you own using the AWS RAM console

See Updating a Resource Share in the AWS RAM User Guide.

Command line

To unshare a shared Dedicated Host that you own using the AWS CLI

Use the disassociate-resource-share command.

Identify a shared Dedicated Host

Owners and consumers can identify shared Dedicated Hosts using one of the following methods.

Amazon EC2 console

To identify a shared Dedicated Host using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts. The screen lists Dedicated Hosts that you own
and Dedicated Hosts that are shared with you. The Owner column shows the AWS account ID of
the Dedicated Host owner.

Command line

To identify a shared Dedicated Host using the AWS CLI

Use the describe-hosts command. The command returns the Dedicated Hosts that you own and
Dedicated Hosts that are shared with you.

View instances running on a shared Dedicated Host

Owners and consumers can view the instances running on a shared Dedicated Host at any time using one
of the following methods.

Amazon EC2 console

To view the instances running on a shared Dedicated Host using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. Select the Dedicated Host for which to view the instances and choose Instances. The tab lists
the instances that are running on the host. Owners see all of the instances running on the host,
including instances launched by consumers. Consumers only see running instances that they
launched onto the host. The Owner column shows the AWS account ID of the account that
launched the instance.

489
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

Command line

To view the instances running on a shared Dedicated Host using the AWS CLI

Use the describe-hosts command. The command returns the instances running on each Dedicated
Host. Owners see all of the instances running on the host. Consumers only see running instances
that they launched on the shared hosts. InstanceOwnerId shows the AWS account ID of the
instance owner.

Shared Dedicated Host permissions

Permissions for owners
Owners are responsible for managing their shared Dedicated Hosts and the instances that they launch
onto them. Owners can view all instances running on the shared Dedicated Host, including those
launched by consumers. However, owners can't take any action on running instances that were launched
by consumers.

Permissions for consumers

Consumers are responsible for managing the instances that they launch onto a shared Dedicated Host.
Consumers can't modify the shared Dedicated Host in any way, and they can't view or modify instances
that were launched by other consumers or the Dedicated Host owner.

Billing and metering

There are no additional charges for sharing Dedicated Hosts.

Owners are billed for Dedicated Hosts that they share. Consumers are not billed for instances that they
launch onto shared Dedicated Hosts.

Dedicated Host Reservations continue to provide billing discounts for shared Dedicated Hosts. Only
Dedicated Host owners can purchase Dedicated Host Reservations for shared Dedicated Hosts that they
own.

Dedicated Host limits

Shared Dedicated Hosts count towards the owner's Dedicated Hosts limits only. Consumer's Dedicated
Hosts limits are not affected by Dedicated Hosts that have been shared with them. Similarly, instances
that consumers launch onto shared Dedicated Hosts do not count towards their instance limits.

Host recovery and Dedicated Host sharing

Host recovery recovers instances launched by the Dedicated Host owner and the consumers with whom it
has been shared. The replacement Dedicated Host is allocated to the owner's account. It is added to the
same resource shares as the original Dedicated Host, and it is shared with the same consumers.

For more information, see Host recovery (p. 490).

Host recovery
Host recovery automatically restarts your instances on to a new replacement host if failures are detected
on your Dedicated Host. Host recovery reduces the need for manual intervention and lowers the
operational burden if there is an unexpected Dedicated Host failure.

Additionally, built-in integration with AWS License Manager automates the tracking and management of
your licenses if a host recovery occurs.
Note
AWS License Manager integration is supported only in Regions in which AWS License Manager is
available.

490
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

Contents
• Host recovery basics (p. 491)
• Supported instance types (p. 492)
• Configure host recovery (p. 492)
• Host recovery states (p. 493)
• Manually recover unsupported instances (p. 493)
• Related services (p. 494)
• Pricing (p. 494)

Host recovery basics

Host recovery uses host-level health checks to assess Dedicated Host availability and to detect
underlying system failures. Examples of problems that can cause host-level health checks to fail include:

• Loss of network connectivity

• Loss of system power
• Hardware or software issues on the physical host

When a system failure is detected on your Dedicated Host, host recovery is initiated and Amazon EC2
automatically allocates a replacement Dedicated Host. The replacement Dedicated Host receives a new
host ID, but retains the same attributes as the original Dedicated Host, including:

• Availability Zone
• Instance type
• Tags
• Auto placement settings

After the replacement Dedicated Host is allocated, the instances are recovered on to the replacement
Dedicated Host. The recovered instances retain the same attributes as the original instances, including:

• Instance ID
• Private IP addresses
• Elastic IP addresses
• EBS volume attachments
• All instance metadata

If instances have a host affinity relationship with the impaired Dedicated Host, the recovered instances
establish host affinity with the replacement Dedicated Host.

When all of the instances have been recovered on to the replacement Dedicated Host, the impaired
Dedicated Host is released, and the replacement Dedicated Host becomes available for use.

When host recovery is initiated, the AWS account owner is notified by email and by an AWS Personal
Health Dashboard event. A second notification is sent after the host recovery has been successfully
completed.

Stopped instances are not recovered on to the replacement Dedicated Host. If you attempt to start
a stopped instance that targets the impaired Dedicated Host, the instance start fails. We recommend
that you modify the stopped instance to either target a different Dedicated Host, or to launch on any
available Dedicated Host with matching configurations and auto-placement enabled.

491
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

Instances with instance storage are not recovered on to the replacement Dedicated Host. As a remedial
measure, the impaired Dedicated Host is marked for retirement and you receive a retirement notification
after the host recovery is complete. Follow the remedial steps described in the retirement notification
within the specified time period to manually recover the remaining instances on the impaired Dedicated
Host.

If you are using AWS License Manager to track your licenses, AWS License Manager allocates new licenses
for the replacement Dedicated Host based on the license configuration limits. If the license configuration
has hard limits that will be breached as a result of the host recovery, the recovery process is not allowed
and you are notified of the host recovery failure through an Amazon SNS notification. If the license
configuration has soft limits that will be breached as a result of the host recovery, the recovery is allowed
to continue and you are notified of the limit breach through an Amazon SNS notification. For more
information, see Using License Configurations in the AWS License Manager User Guide.

Supported instance types

Host recovery is supported for the following instance families: A1, C3, C4, C5, C5n, C6g, Inf1, M3, M4,
M5, M5n, M6g, P3, R3, R4, R5, R5n, R6g, X1, X1e, u-6tb1, u-9tb1, u-12tb1, u-18tb1, and u-24tb1.

To recover instances that are not supported, see Manually recover unsupported instances (p. 493).

Configure host recovery

You can configure host recovery at the time of Dedicated Host allocation, or after allocation using the
Amazon EC2 console or AWS Command Line Interface (CLI).

Contents
• Enable host recovery (p. 492)
• Disable host recovery (p. 492)
• View the host recovery configuration (p. 493)

Enable host recovery

You can enable host recovery at the time of Dedicated Host allocation or after allocation.

For more information about enabling host recovery at the time of Dedicated Host allocation, see Allocate
Dedicated Hosts (p. 469).

To enable host recovery after allocation using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. Select the Dedicated Host for which to enable host recovery, and then choose Actions, Modify Host
Recovery.
4. For Host recovery, choose Enable, and then choose Save.

To enable host recovery after allocation using the AWS CLI

Use the modify-hosts command and specify the host-recovery parameter.

$ aws ec2 modify-hosts --host-recovery on --host-ids h-012a3456b7890cdef

Disable host recovery

You can disable host recovery at any time after the Dedicated Host has been allocated.

492
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

To disable host recovery after allocation using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. Select the Dedicated Host for which to disable host recovery, and then choose Actions, Modify Host
Recovery.
4. For Host recovery, choose Disable, and then choose Save.

To disable host recovery after allocation using the AWS CLI

Use the modify-hosts command and specify the host-recovery parameter.

$ aws ec2 modify-hosts --host-recovery off --host-ids h-012a3456b7890cdef

View the host recovery configuration

You can view the host recovery configuration for a Dedicated Host at any time.

To view the host recovery configuration for a Dedicated Host using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Dedicated Hosts.
3. Select the Dedicated Host, and in the Description tab, review the Host Recovery field.

To view the host recovery configuration for a Dedicated Host using the AWS CLI

Use the describe-hosts command.

$ aws ec2 describe-hosts --host-ids h-012a3456b7890cdef

The HostRecovery response element indicates whether host recovery is enabled or disabled.

Host recovery states

When a Dedicated Host failure is detected, the impaired Dedicated Host enters the under-assessment
state, and all of the instances enter the impaired state. You can't launch instances on to the impaired
Dedicated Host while it is in the under-assessment state.

After the replacement Dedicated Host is allocated, it enters the pending state. It remains in this state
until the host recovery process is complete. You can't launch instances on to the replacement Dedicated
Host while it is in the pending state. Recovered instances on the replacement Dedicated Host remain in
the impaired state during the recovery process.

After the host recovery is complete, the replacement Dedicated Host enters the available state,
and the recovered instances return to the running state. You can launch instances on to the
replacement Dedicated Host after it enters the available state. The original impaired Dedicated Host is
permanently released and it enters the released-permanent-failure state.

If the impaired Dedicated Host has instances that do not support host recovery, such as instances with
instance store-backed volumes, the Dedicated Host is not released. Instead, it is marked for retirement
and enters the permanent-failure state.

Manually recover unsupported instances

Host recovery does not support recovering instances that use instance store volumes. Follow the
instructions below to manually recover any of your instances that could not be automatically recovered.

493
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

Warning
Data on instance store volumes is lost when an instance is stopped, hibernated, or terminated.
This includes instance store volumes that are attached to an instance that has an EBS volume
as the root device. To protect data from instance store volumes, back it up to persistent storage
before the instance is stopped or terminated.

Manually recover EBS-backed instances

For EBS-backed instances that could not be automatically recovered, we recommend that you manually
stop and start the instances to recover them onto a new Dedicated Host. For more information about
stopping your instance, and about the changes that occur in your instance configuration when it's
stopped, see Stop and start your instance (p. 623).

Manually recover instance store-backed instances

For instance store-backed instances that could not be automatically recovered, we recommend that you
do the following:

1. Launch a replacement instance on a new Dedicated Host from your most recent AMI.
2. Migrate all of the necessary data to the replacement instance.
3. Terminate the original instance on the impaired Dedicated Host.

Related services
Dedicated Host integrates with the following AWS services:

• AWS License Manager—Tracks licenses across your Amazon EC2 Dedicated Hosts (supported only
in Regions in which AWS License Manager is available). For more information, see the AWS License
Manager User Guide.

Pricing
There are no additional charges for using host recovery, but the usual Dedicated Host charges apply. For
more information, see Amazon EC2 Dedicated Hosts Pricing.

As soon as host recovery is initiated, you are no longer billed for the impaired Dedicated Host. Billing for
the replacement Dedicated Host begins only after it enters the available state.

If the impaired Dedicated Host was billed using the On-Demand rate, the replacement Dedicated Host
is also billed using the On-Demand rate. If the impaired Dedicated Host had an active Dedicated Host
Reservation, it is transferred to the replacement Dedicated Host.

Track configuration changes

You can use AWS Config to record configuration changes for Dedicated Hosts, and for instances that are
launched, stopped, or terminated on them. You can then use the information captured by AWS Config as
a data source for license reporting.

AWS Config records configuration information for Dedicated Hosts and instances individually, and pairs
this information through relationships. There are three reporting conditions:

• AWS Config recording status—When On, AWS Config is recording one or more AWS resource types,
which can include Dedicated Hosts and Dedicated Instances. To capture the information required for
license reporting, verify that hosts and instances are being recorded with the following fields.
• Host recording status—When Enabled, the configuration information for Dedicated Hosts is recorded.

494
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Hosts

• Instance recording status—When Enabled, the configuration information for Dedicated Instances is
recorded.

If any of these three conditions are disabled, the icon in the Edit Config Recording button is red. To
derive the full benefit of this tool, ensure that all three recording methods are enabled. When all three
are enabled, the icon is green. To edit the settings, choose Edit Config Recording. You are directed to
the Set up AWS Config page in the AWS Config console, where you can set up AWS Config and start
recording for your hosts, instances, and other supported resource types. For more information, see
Setting up AWS Config using the Console in the AWS Config Developer Guide.
Note
AWS Config records your resources after it discovers them, which might take several minutes.

After AWS Config starts recording configuration changes to your hosts and instances, you can get the
configuration history of any host that you have allocated or released and any instance that you have
launched, stopped, or terminated. For example, at any point in the configuration history of a Dedicated
Host, you can look up how many instances are launched on that host, along with the number of sockets
and cores on the host. For any of those instances, you can also look up the ID of its Amazon Machine
Image (AMI). You can use this information to report on licensing for your own server-bound software
that is licensed per-socket or per-core.

You can view configuration histories in any of the following ways:

• By using the AWS Config console. For each recorded resource, you can view a timeline page, which
provides a history of configuration details. To view this page, choose the gray icon in the Config
Timeline column of the Dedicated Hosts page. For more information, see Viewing Configuration
Details in the AWS Config Console in the AWS Config Developer Guide.
• By running AWS CLI commands. First, you can use the list-discovered-resources command to get a
list of all hosts and instances. Then, you can use the get-resource-config-history command to get the
configuration details of a host or instance for a specific time interval. For more information, see View
Configuration Details Using the CLI in the AWS Config Developer Guide.
• By using the AWS Config API in your applications. First, you can use the ListDiscoveredResources action
to get a list of all hosts and instances. Then, you can use the GetResourceConfigHistory action to get
the configuration details of a host or instance for a specific time interval.

For example, to get a list of all of your Dedicated Hosts from AWS Config, run a CLI command such as the
following.

aws configservice list-discovered-resources --resource-type AWS::EC2::Host

To obtain the configuration history of a Dedicated Host from AWS Config, run a CLI command such as
the following.

aws configservice get-resource-config-history --resource-type AWS::EC2::Instance --

resource-id i-1234567890abcdef0

To manage AWS Config settings using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. On the Dedicated Hosts page, choose Edit Config Recording.
3. In the AWS Config console, follow the steps provided to turn on recording. For more information, see
Setting up AWS Config using the Console.

For more information, see Viewing Configuration Details in the AWS Config Console.

495
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Instances

To activate AWS Config using the command line or API

• AWS CLI: Viewing Configuration Details (AWS CLI) in the AWS Config Developer Guide.
• Amazon EC2 API: GetResourceConfigHistory.

Dedicated Instances
Dedicated Instances are Amazon EC2 instances that run in a virtual private cloud (VPC) on hardware
that's dedicated to a single customer. Dedicated Instances that belong to different AWS accounts are
physically isolated at a hardware level, even if those accounts are linked to a single payer account.
However, Dedicated Instances may share hardware with other instances from the same AWS account that
are not Dedicated Instances.
Note
A Dedicated Host is also a physical server that's dedicated for your use. With a Dedicated
Host, you have visibility and control over how instances are placed on the server. For more
information, see Dedicated Hosts (p. 465).

Dedicated Instance basics

Each instance that you launch into a VPC has a tenancy attribute. This attribute has the following values.

default

Your instance runs on shared hardware.

dedicated

Your instance runs on single-tenant hardware.

host

Your instance runs on a Dedicated Host, which is an isolated server with configurations that you can
control.

After you launch an instance, there are some limitations to changing its tenancy.

• You cannot change the tenancy of an instance from default to dedicated or host after you've
launched it.
• You cannot change the tenancy of an instance from dedicated or host to default after you've
launched it.

You can change the tenancy of an instance from dedicated to host, or from host to dedicated after
you've launched it. For more information, see Change the tenancy of an instance (p. 500).

Each VPC has a related instance tenancy attribute. This attribute has the following values.

default

An instance launched into the VPC runs on shared hardware by default, unless you explicitly specify
a different tenancy during instance launch.
dedicated

An instance launched into the VPC is a Dedicated Instance by default, unless you explicitly specify a
tenancy of host during instance launch. You cannot specify a tenancy of default during instance
launch.

496
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Instances

You can change the instance tenancy of a VPC from dedicated to default after you create it. You
cannot change the instance tenancy of a VPC from default to dedicated after it is created.

To create Dedicated Instances, you can do the following:

• Create the VPC with the instance tenancy set to dedicated (all instances launched into this VPC are
Dedicated Instances).
• Create the VPC with the instance tenancy set to default, and specify a tenancy of dedicated for any
instances when you launch them.

Dedicated Instances limitations

Some AWS services or their features won't work with a VPC with the instance tenancy set to dedicated.
Check the service's documentation to confirm if there are any limitations.

Some instance types cannot be launched into a VPC with the instance tenancy set to dedicated. For
more information about supported instances types, see Amazon EC2 Dedicated Instances.

Amazon EBS with Dedicated Instances

When you launch an Amazon EBS-backed Dedicated Instance, the EBS volume doesn't run on single-
tenant hardware.

Reserved Instances with dedicated tenancy

To guarantee that sufficient capacity is available to launch Dedicated Instances, you can purchase
Dedicated Reserved Instances. For more information, see Reserved Instances (p. 325).

When you purchase a Dedicated Reserved Instance, you are purchasing the capacity to launch a
Dedicated Instance into a VPC at a much reduced usage fee; the price break in the usage charge applies
only if you launch an instance with dedicated tenancy. When you purchase a Reserved Instance with
default tenancy, it applies only to a running instance with default tenancy; it would not apply to a
running instance with dedicated tenancy.

You can't use the modification process to change the tenancy of a Reserved Instance after you've
purchased it. However, you can exchange a Convertible Reserved Instance for a new Convertible
Reserved Instance with a different tenancy.

Automatic scaling of Dedicated Instances

You can use Amazon EC2 Auto Scaling to launch Dedicated Instances. For more information, see
Launching Auto Scaling Instances in a VPC in the Amazon EC2 Auto Scaling User Guide.

Automatic recovery of Dedicated Instances

You can configure automatic recovery for a Dedicated Instances if it becomes impaired due to an
underlying hardware failure or a problem that requires AWS involvement to repair. For more information,
see Recover your instance (p. 648).

Dedicated Spot Instances

You can run a Dedicated Spot Instance by specifying a tenancy of dedicated when you create a Spot
Instance request. For more information, see Specify a tenancy for your Spot Instances (p. 385).

Pricing for Dedicated Instances

Pricing for Dedicated Instances is different to pricing for On-Demand Instances. For more information,
see the Amazon EC2 Dedicated Instances product page.

497
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Instances

Burstable performance instances with Dedicated Instances

You can leverage the benefits of running on dedicated tenancy hardware with the section called
“Burstable performance instances” (p. 224). T3 Dedicated Instances launch in unlimited mode by default,
and they provide a baseline level of CPU performance with the ability to burst to a higher CPU level
when required by your workload. The T3 baseline performance and ability to burst are governed by
CPU credits. Because of the burstable nature of the T3 instance types, we recommend that you monitor
how your T3 instances use the CPU resources of the dedicated hardware for the best performance.
T3 Dedicated Instances are intended for customers with diverse workloads that display random CPU
behavior, but that ideally have average CPU usage at or below the baseline usages. For more information,
see the section called “CPU credits and baseline utilization” (p. 225).

Amazon EC2 has systems in place to identify and correct variability in performance. However, it is still
possible to experience short term variability if you launch multiple T3 Dedicated Instances that have
correlated CPU usage patterns. For these more demanding or correlated workloads, we recommend
using M5 or M5a Dedicated Instances rather than T3 Dedicated Instances.

Work with Dedicated Instances

You can create a VPC with an instance tenancy of dedicated to ensure that all instances launched
into the VPC are Dedicated Instances. Alternatively, you can specify the tenancy of the instance during
launch.

Topics
• Create a VPC with an instance tenancy of dedicated (p. 498)
• Launch Dedicated Instances into a VPC (p. 499)
• Display tenancy information (p. 499)
• Change the tenancy of an instance (p. 500)
• Change the tenancy of a VPC (p. 500)

Create a VPC with an instance tenancy of dedicated

When you create a VPC, you have the option of specifying its instance tenancy. If you're using the
Amazon VPC console, you can create a VPC using the VPC wizard or the Your VPCs page.

To create a VPC with an instance tenancy of dedicated (VPC Wizard)

1. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.

2. From the dashboard, choose Launch VPC Wizard.
3. Select a VPC configuration, and then choose Select.
4. For Hardware tenancy, choose Dedicated.
5. Choose Create VPC.

To create a VPC with an instance tenancy of dedicated (Create VPC dialog box)

1. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.

2. In the navigation pane, choose Your VPCs, and then choose Create VPC.
3. For Tenancy, choose Dedicated. Specify the CIDR block, and choose Create VPC.

To set the tenancy option when you create a VPC using the command line

• create-vpc (AWS CLI)

498
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Instances

• New-EC2Vpc (AWS Tools for Windows PowerShell)

If you launch an instance into a VPC that has an instance tenancy of dedicated, your instance is
automatically a Dedicated Instance, regardless of the tenancy of the instance.

Launch Dedicated Instances into a VPC

You can launch a Dedicated Instance using the Amazon EC2 launch instance wizard.

To launch a Dedicated Instance into a default tenancy VPC using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Launch Instance.
3. On the Choose an Amazon Machine Image (AMI) page, select an AMI and choose Select.
4. On the Choose an Instance Type page, select the instance type and choose Next: Configure
Instance Details.
Note
Ensure that you choose an instance type that's supported as a Dedicated Instance. For more
information, see Amazon EC2 Dedicated Instances.
5. On the Configure Instance Details page, select a VPC and subnet. For Tenancy, choose Dedicated -
Run a dedicated instance, and then choose Next: Add Storage.
6. Continue as prompted by the wizard. When you've finished reviewing your options on the Review
Instance Launch page, choose Launch to choose a key pair and launch the Dedicated Instance.

For more information about launching an instance with a tenancy of host, see Launch instances onto a
Dedicated Host (p. 472).

To set the tenancy option for an instance during launch using the command line

• run-instances (AWS CLI)

• New-EC2Instance (AWS Tools for Windows PowerShell)

Display tenancy information

To display tenancy information for your VPC using the console

1. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.

2. In the navigation pane, choose Your VPCs.
3. Check the instance tenancy of your VPC in the Tenancy column.
4.
If the Tenancy column is not displayed, choose the settings icon ( ) in the top-right corner,
toggle on Tenancy, and choose Confirm.

To display tenancy information for your instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Check the tenancy of your instance in the Tenancy column.
4. If the Tenancy column is not displayed, do one of the following:

•
Choose the settings icon ( ) in the top-right corner, toggle on Tenancy, and choose Confirm.

499
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Dedicated Instances

• Select the instance. On the Details tab near the bottom of the page, under Host and placement
group, check the value for Tenancy.

To describe the tenancy of your VPC using the command line

• describe-vpcs (AWS CLI)

• Get-EC2Vpc (AWS Tools for Windows PowerShell)

To describe the tenancy of your instance using the command line

• describe-instances (AWS CLI)

• Get-EC2Instance (AWS Tools for Windows PowerShell)

To describe the tenancy value of a Reserved Instance using the command line

• describe-reserved-instances (AWS CLI)

• Get-EC2ReservedInstance (AWS Tools for Windows PowerShell)

To describe the tenancy value of a Reserved Instance offering using the command line

• describe-reserved-instances-offerings (AWS CLI)

• Get-EC2ReservedInstancesOffering (AWS Tools for Windows PowerShell)

Change the tenancy of an instance

Depending on your instance type and platform, you can change the tenancy of a stopped Dedicated
Instance to host after launching it. The next time the instance starts, it's started on a Dedicated Host
that's allocated to your account. For more information about allocating and working with Dedicated
Hosts, and the instance types that can be used with Dedicated Hosts, see Work with Dedicated
Hosts (p. 469). Similarly, you can change the tenancy of a stopped Dedicated Host instance to
dedicated after launching it. The next time the instance starts, it's started on single-tenant hardware
that we control.

To change the tenancy of an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and select your instance.
3. Choose Instance state, Stop instance, Stop.
4. Choose Actions, Instance settings, Modify instance placement.
5. For Tenancy, choose whether to run your instance on dedicated hardware or on a Dedicated Host.
Choose Save.

To modify the tenancy value of an instance using the command line

• modify-instance-placement (AWS CLI)

• Edit-EC2InstancePlacement (AWS Tools for Windows PowerShell)

Change the tenancy of a VPC

You can change the instance tenancy attribute of a VPC from dedicated to default. Modifying the
instance tenancy of the VPC does not affect the tenancy of any existing instances in the VPC. The next

500
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

time you launch an instance in the VPC, it has a tenancy of default, unless you specify otherwise during
launch.

You cannot change the tenancy attribute of a VPC from default to dedicated after it is created.

You can modify the instance tenancy attribute of a VPC using the AWS CLI, an AWS SDK, or the Amazon
EC2 API only.

To modify the instance tenancy attribute of a VPC using the AWS CLI

• Use the modify-vpc-tenancy command to specify the ID of the VPC and instance tenancy value. The
only supported value is default.

aws ec2 modify-vpc-tenancy --vpc-id vpc-1a2b3c4d --instance-tenancy default

On-Demand Capacity Reservations

On-Demand Capacity Reservations enable you to reserve capacity for your Amazon EC2 instances in
a specific Availability Zone for any duration. This gives you the ability to create and manage Capacity
Reservations independently from the billing discounts offered by Savings Plans or regional Reserved
Instances. By creating Capacity Reservations, you ensure that you always have access to EC2 capacity
when you need it, for as long as you need it. You can create Capacity Reservations at any time, without
entering into a one-year or three-year term commitment, and the capacity is available immediately.
When you no longer need it, cancel the Capacity Reservation to stop incurring charges.

When you create a Capacity Reservation, you specify:

• The Availability Zone in which to reserve the capacity

• The number of instances for which to reserve capacity
• The instance attributes, including the instance type, tenancy, and platform/OS

Capacity Reservations can only be used by instances that match their attributes. By default, they are
automatically used by running instances that match the attributes. If you don't have any running
instances that match the attributes of the Capacity Reservation, it remains unused until you launch an
instance with matching attributes.

In addition, you can use Savings Plans and regional Reserved Instances with your Capacity Reservations
to benefit from billing discounts. AWS automatically applies your discount when the attributes of a
Capacity Reservation match the attributes of a Savings Plan or regional Reserved Instance. For more
information, see Billing discounts (p. 504).

Contents
• Differences between Capacity Reservations, Reserved Instances, and Savings Plans (p. 502)
• Supported platforms (p. 502)
• Capacity Reservation limits (p. 503)
• Capacity Reservation limitations and restrictions (p. 503)
• Capacity Reservation pricing and billing (p. 503)
• Work with Capacity Reservations (p. 504)
• Capacity Reservations in Local Zones (p. 513)
• Capacity Reservations in Wavelength Zones (p. 514)
• Work with shared Capacity Reservations (p. 514)
• CloudWatch metrics for On-Demand Capacity Reservations (p. 519)

501
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

Differences between Capacity Reservations, Reserved Instances,

and Savings Plans
The following table highlights key differences between Capacity Reservations, Reserved Instances, and
Savings Plans:

  Capacity Zonal Reserved Regional Reserved Savings Plans

Reservations Instances Instances

Term No commitment Requires a fixed one-year or three-year commitment

required. Can be
created and canceled
as needed.

Capacity Capacity reserved in a specific Availability No capacity reserved.

benefit Zone.

Billing No billing discount. † Provides a billing discount.

discount

Instance Your On-Demand Default is 20 per Default is 20 per No limit.

Limits Instance limits per Availability Zone. Region. You can
Region apply. You can request a request a limit
limit increase. increase.

† You can combine Capacity Reservations with Savings Plans or Regional Reserved Instances to receive a
discount.

For more information, see the following:

• Reserved Instances (p. 325)

• Savings Plans User Guide

Supported platforms
You must create the Capacity Reservation with the correct platform to ensure that it properly matches
with your instances. Capacity Reservations support the following platforms:

• Linux/UNIX
• Linux with SQL Server Standard
• Linux with SQL Server Web
• Linux with SQL Server Enterprise
• Red Hat Enterprise Linux
• SUSE Linux

When you purchase a Capacity Reservation, you must specify the platform that represents the operating
system for your instance.

• For SUSE Linux and RHEL distributions, excluding BYOL, you must choose the specific platform. For
example, the SUSE Linux or Red Hat Enterprise Linux platform.
• For all other Linux distributions (including Ubuntu), choose the Linux/UNIX platform.
• If you bring your existing RHEL subscription (BYOL), you must choose the Linux/UNIX platform.

502
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

For more information about the supported Windows platforms, see Supported platforms in the Amazon
EC2 User Guide for Windows Instances.

Capacity Reservation limits

The number of instances for which you are allowed to reserve capacity is based on your account's On-
Demand Instance limit. You can reserve capacity for as many instances as that limit allows, minus the
number of instances that are already running.

Capacity Reservation limitations and restrictions

Before you create Capacity Reservations, take note of the following limitations and restrictions.

• Active and unused Capacity Reservations count toward your On-Demand Instance limits
• Capacity Reservations are not transferable from one AWS account to another. However, you can share
Capacity Reservations with other AWS accounts. For more information, see Work with shared Capacity
Reservations (p. 514).
• Zonal Reserved Instance billing discounts do not apply to Capacity Reservations
• Capacity Reservations can't be created in placement groups
• Capacity Reservations can't be used with Dedicated Hosts

Capacity Reservation pricing and billing

The price for a Capacity Reservation varies by payment option.

Pricing
When the Capacity Reservation is active, you are charged the equivalent On-Demand rate whether you
run the instances or not. If you do not use the reservation, this shows up as unused reservation on your
EC2 bill. When you run an instance that matches the attributes of a reservation, you just pay for the
instance and nothing for the reservation. There are no upfront or additional charges.

For example, if you create a Capacity Reservation for 20 m4.large Linux instances and run 15
m4.large Linux instances in the same Availability Zone, you will be charged for 15 active instances and
for 5 unused instances in the reservation.

Billing discounts for Savings Plans and regional Reserved Instances apply to Capacity Reservations. For
more information, see Billing discounts (p. 504).

For more information, see Amazon EC2 Pricing.

Billing
Capacity Reservations are billed at per-second granularity. This means that you are charged for partial
hours. For example, if a reservation remains active in your account for 24 hours and 15 minutes, you will
be billed for 24.25 reservation hours.

The following example shows how a Capacity Reservation is billed. The Capacity Reservation is created
for one m4.large Linux instance, which has an On-Demand rate of $0.10 per usage hour. In this
example, the Capacity Reservation is active in the account for five hours. The Capacity Reservation is
unused for the first hour, so it is billed for one unused hour at the m4.large instance type's standard
On-Demand rate. In hours two through five, the Capacity Reservation is occupied by an m4.large
instance. During this time, the Capacity Reservation accrues no charges, and the account is instead billed
for the m4.large instance occupying it. In the sixth hour, the Capacity Reservation is canceled and the
m4.large instance runs normally outside of the reserved capacity. For that hour, it is charged at the On-
Demand rate of the m4.large instance type.

503
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

Billing discounts
Billing discounts for Savings Plans and regional Reserved Instances apply to Capacity Reservations. AWS
automatically applies these discounts to Capacity Reservations that have matching attributes. When
a Capacity Reservation is used by an instance, the discount is applied to the instance. Discounts are
preferentially applied to instance usage before covering unused Capacity Reservations.

Billing discounts for zonal Reserved Instances do not apply to Capacity Reservations.

For more information, see the following:

• Reserved Instances (p. 325)

• Savings Plans User Guide

Viewing your bill

You can review the charges and fees to your account on the AWS Billing and Cost Management console.

• The Dashboard displays a spend summary for your account.

• On the Bills page, under Details, expand the Elastic Compute Cloud section and the Region to get
billing information about your Capacity Reservations.

You can view the charges online, or you can download a CSV file. For more information, see Capacity
Reservation Line Items in the AWS Billing and Cost Management User Guide.

Work with Capacity Reservations

To start using Capacity Reservations, you create the capacity reservation in the required Availability
Zone. Then, you can launch instances into the reserved capacity, view its capacity utilization in real time,
and increase or decrease its capacity as needed.

By default, Capacity Reservations automatically match new instances and running instances that have
matching attributes (instance type, platform, and Availability Zone). This means that any instance with
matching attributes automatically runs in the Capacity Reservation. However, you can also target a
Capacity Reservation for specific workloads. This enables you to explicitly control which instances are
allowed to run in that reserved capacity.

You can specify how the reservation ends. You can choose to manually cancel the Capacity Reservation
or end it automatically at a specified time. If you specify an end time, the Capacity Reservation is
canceled within an hour of the specified time. For example, if you specify 5/31/2019, 13:30:55, the
Capacity Reservation is guaranteed to end between 13:30:55 and 14:30:55 on 5/31/2019. After a
reservation ends, you can no longer target instances to the Capacity Reservation. Instances running in
the reserved capacity continue to run uninterrupted. If instances targeting a Capacity Reservation are
stopped, you cannot restart them until you remove their Capacity Reservation targeting preference or
configure them to target a different Capacity Reservation.

504
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

Contents
• Create a Capacity Reservation (p. 505)
• Work with Capacity Reservation groups (p. 506)
• Launch instances into an existing Capacity Reservation (p. 509)
• Modify a Capacity Reservation (p. 510)
• Modify an instance's Capacity Reservation settings (p. 511)
• View a Capacity Reservation (p. 512)
• Cancel a Capacity Reservation (p. 513)

Create a Capacity Reservation

After you create the Capacity Reservation, the capacity is available immediately. The capacity remains
reserved for your use as long as the Capacity Reservation is active, and you can launch instances into it
at any time. If the Capacity Reservation is open, new instances and existing instances that have matching
attributes automatically run in the capacity of the Capacity Reservation. If the Capacity Reservation is
targeted, instances must specifically target it to run in the reserved capacity.

Your request to create a Capacity Reservation could fail if one of the following is true:

• Amazon EC2 does not have sufficient capacity to fulfill the request. Either try again at a later time,
try a different Availability Zone, or try a smaller capacity. If your application is flexible across instance
types and sizes, try different instance attributes.
• The requested quantity exceeds your On-Demand Instance limit for the selected instance family.
Increase your On-Demand Instance limit for the instance family and try again. For more information,
see On-Demand Instance limits (p. 322).

To create a Capacity Reservation using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Capacity Reservations, and then choose Create Capacity Reservation.
3. On the Create a Capacity Reservation page, configure the following settings in the Instance details
section. The instance type, platform, and Availability Zone of the instances that you launch must
match the instance type, platform, and Availability Zone that you specify here or the Capacity
Reservation is not applied. For example, if an open Capacity Reservation doesn't match, an instance
launch that targets that Capacity Reservation explicitly will fail.

a. Instance Type—The type of instance to launch into the reserved capacity.

b. Launch EBS-optimized instances—Specify whether to reserve the capacity for EBS-optimized
instances. This option is selected by default for some instance types. For more information
about EBS-optimized instances, see Amazon Elastic Block Store (p. 1080).
c. Attach instance store at launch—Specify whether instances launched into the Capacity
Reservation use temporary block-level storage. The data on an instance store volume persists
only during the life of the associated instance.
d. Platform—The operating system for your instances. For more information, see Supported
platforms (p. 502). For more information about the supported Windows platforms, see
Supported platforms in the Amazon EC2 User Guide for Windows Instances.
e. Availability Zone—The Availability Zone in which to reserve the capacity.
f. Tenancy—Specify whether to run on shared hardware (default) or a dedicated instance.
g. Quantity—The number of instances for which to reserve capacity. If you specify a quantity that
exceeds your remaining On-Demand Instance limit for the selected instance type, the request is
denied.
4. Configure the following settings in the Reservation details section:

505
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

a. Reservation Ends—Choose one of the following options:

• Manually—Reserve the capacity until you explicitly cancel it.

• Specific time—Cancel the capacity reservation automatically at the specified date and time.
b. Instance eligibility—Choose one of the following options:

• open—(Default) The Capacity Reservation matches any instance that has matching attributes
(instance type, platform, and Availability Zone). If you launch an instance with matching
attributes, it is placed into the reserved capacity automatically.
• targeted—The Capacity Reservation only accepts instances that have matching attributes
(instance type, platform, and Availability Zone), and that explicitly target the reservation.
5. Choose Request reservation.

To create a Capacity Reservation using the AWS CLI

Use the create-capacity-reservation command. For more information, see Supported

platforms (p. 502). For more information about the supported Windows platforms, see Supported
platforms in the Amazon EC2 User Guide for Windows Instances.

For example, the following command creates a Capacity Reservation that reserves capacity for three
m5.2xlarge instances running Red Hat Enterprise Linux AMIs in the us-east-1a Availability Zone.

aws ec2 create-capacity-reservation --instance-type m5.2xlarge --instance-platform Red Hat

Enterprise Linux --availability-zone us-east-1a --instance-count 3

Work with Capacity Reservation groups

You can use AWS Resource Groups to create logical collections of Capacity Reservations, called resource
groups. A resource group is a logical grouping of AWS resources that are all in the same AWS Region. You
can include multiple Capacity Reservations that have different attributes (instance type, platform, and
Availability Zone) in a single resource group.

When you create resource groups for your Capacity Reservations, you can target instances to a group
of Capacity Reservations instead of an individual Capacity Reservation. Instances that target a group of
Capacity Reservations match with any Capacity Reservation in the group that has matching attributes
(instance type, platform, and Availability Zone) and available capacity. If the group does not have a
Capacity Reservation with matching attributes and available capacity, the instances run using On-
Demand capacity. If a matching Capacity Reservation is added to the targeted group at a later stage, the
instance is automatically matched with and moved into its reserved capacity.

To prevent unintended use of Capacity Reservations in a group, configure the Capacity Reservations in
the group to accept only instances that explicitly target the capacity reservation. To do this, set Instance
eligibility to targeted (old console) or Only instances that specify this reservation (new console) when
creating the Capacity Reservation using the Amazon EC2 console. When using the AWS CLI, specify --
instance-match-criteria targeted when creating the Capacity Reservation. Doing this ensures
that only instances that explicitly target the group, or a Capacity Reservation in the group, can run in the
group.

If a Capacity Reservation in a group is canceled or expires while it has running instances, the instances
are automatically moved to another Capacity Reservation in the group that has matching attributes
and available capacity. If there are no remaining Capacity Reservations in the group that have matching
attributes and available capacity, the instances run in On-Demand capacity. If a matching Capacity
Reservation is added to the targeted group at a later stage, the instance is automatically moved into its
reserved capacity.

To create a group for your Capacity Reservations

506
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

Use the create-group AWS CLI command. For name, provide a descriptive name for the group, and for
configuration, specify two Type request parameters:

• AWS::EC2::CapacityReservationPool to ensure that the resource group can be targeted for

instance launches
• AWS::ResourceGroups::Generic with allowed-resource-types set to
AWS::EC2::CapacityReservation to ensure that the resource group accepts Capacity Reservations
only

For example, the following command creates a group named MyCRGroup.

$ aws resource-groups create-group --name MyCRGroup --configuration

'{"Type":"AWS::EC2::CapacityReservationPool"}' '{"Type":"AWS::ResourceGroups::Generic",
"Parameters": [{"Name": "allowed-resource-types", "Values":
["AWS::EC2::CapacityReservation"]}]}'

The following shows example output.

{
"GroupConfiguration": {
"Status": "UPDATE_COMPLETE",
"Configuration": [
{
"Type": "AWS::EC2::CapacityReservationPool"
},
{
"Type": "AWS::ResourceGroups::Generic",
"Parameters": [
{
"Values": [
"AWS::EC2::CapacityReservation"
],
"Name": "allowed-resource-types"
}
]
}
]
},
"Group": {
"GroupArn": "arn:aws:resource-groups:sa-east-1:123456789012:group/MyCRGroup",
"Name": "MyCRGroup"
}
}

To add a Capacity Reservation to a group

Use the group-resources AWS CLI command. For group, specify the name of the group to which to add
the Capacity Reservations, and for resources, specify ARNs of the Capacity Reservations to add. To
add multiple Capacity Reservations, separate the ARNs with a space. To get the ARNs of the Capacity
Reservations to add, use the describe-capacity-reservations AWS CLI command and specify the IDs of
the Capacity Reservations.

For example, the following command adds two Capacity Reservations to a group named MyCRGroup.

$ aws resource-groups group-resources --group MyCRGroup --resource-arns arn:aws:ec2:sa-

east-1:123456789012:capacity-reservation/cr-1234567890abcdef1 arn:aws:ec2:sa-
east-1:123456789012:capacity-reservation/cr-54321abcdef567890

The following shows example output.

507
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

{
"Failed": [],
"Succeeded": [
"arn:aws:ec2:sa-east-1:123456789012:capacity-reservation/cr-1234567890abcdef1",
"arn:aws:ec2:sa-east-1:123456789012:capacity-reservation/cr-54321abcdef567890"
]
}

To view the Capacity Reservations in a specific group

Use the list-group-resources AWS CLI command. For group, specify the name of the group.

For example, the following command lists the Capacity Reservations in a group named MyCRGroup.

$ aws resource-groups list-group-resources --group MyCRGroup

The following shows example output.

{
"QueryErrors": [],
"ResourceIdentifiers": [
{
"ResourceType": "AWS::EC2::CapacityReservation",
"ResourceArn": "arn:aws:ec2:sa-east-1:123456789012:capacity-reservation/
cr-1234567890abcdef1"
},
{
"ResourceType": "AWS::EC2::CapacityReservation",
"ResourceArn": "arn:aws:ec2:sa-east-1:123456789012:capacity-reservation/
cr-54321abcdef567890"
}
]
}

To view the groups to which a specific Capacity Reservation has been added (AWS CLI)

Use the get-groups-for-capacity-reservation AWS CLI command.

For example, the following command lists the groups to which Capacity Reservation
cr-1234567890abcdef1 has been added.

$ aws ec2 get-groups-for-capacity-reservation --capacity-reservation-

id cr-1234567890abcdef1

The following shows example output.

{
"CapacityReservationGroups": [
{
"OwnerId": "123456789012",
"GroupArn": "arn:aws:resource-groups:sa-east-1:123456789012:group/MyCRGroup"
}
]
}

To view the groups to which a specific Capacity Reservation has been added (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

508
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

2. In the navigation pane, choose Capacity Reservations, select the Capacity Reservation to view, and
then choose View.

The groups to which the Capacity Reservation has been added are listed in the Groups card.

To remove a Capacity Reservation from a group

Use the ungroup-resources AWS CLI command. For group, specify the ARN of the group from which to
remove the Capacity Reservation, and for resources specify the ARNs of the Capacity Reservations to
remove. To remove multiple Capacity Reservations, separate the ARNs with a space.

The following example removes two Capacity Reservations from a group named MyCRGroup.

$ aws resource-groups ungroup-resources --group MyCRGroup --resource-arns arn:aws:ec2:sa-

east-1:123456789012:capacity-reservation/cr-0e154d26a16094dd arn:aws:ec2:sa-
east-1:123456789012:capacity-reservation/cr-54321abcdef567890

The following shows example output.

{
"Failed": [],
"Succeeded": [
"arn:aws:ec2:sa-east-1:123456789012:capacity-reservation/cr-0e154d26a16094dd",
"arn:aws:ec2:sa-east-1:123456789012:capacity-reservation/cr-54321abcdef567890"
]
}

To delete a group

Use the delete-group AWS CLI command. For group, provide the name of the group to delete.

For example, the following command deletes a group named MyCRGroup.

$ aws resource-groups delete-group --group MyCRGroup

The following shows example output.

{
"Group": {
"GroupArn": "arn:aws:resource-groups:sa-east-1:123456789012:group/MyCRGroup",
"Name": "MyCRGroup"
}
}

Launch instances into an existing Capacity Reservation

When you launch an instance, you can specify whether to launch the instance into any open Capacity
Reservation, into a specific Capacity Reservation, or into a group of Capacity Reservations. You can only
launch an instance into a Capacity Reservation that has matching attributes (instance type, platform, and
Availability Zone) and sufficient capacity. Alternatively, you can configure the instance to avoid running in
a Capacity Reservation, even if you have an open Capacity Reservation that has matching attributes and
available capacity.

Launching an instance into a Capacity Reservation reduces its available capacity by the number of
instances launched. For example, if you launch three instances, the available capacity of the Capacity
Reservation is reduced by three.

509
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

To launch instances into an existing Capacity Reservation using the console

1. Open the Launch Instance wizard by choosing Launch Instances from Dashboard or Instances.
2. Select an Amazon Machine Image (AMI) and an instance type.
3. Complete the Configure Instance Details page. For Capacity Reservation, choose one of the
following options:

• None — Prevents the instances from launching into a Capacity Reservation. The instances run in
On-Demand capacity.
• Open — Launches the instances into any Capacity Reservation that has matching attributes and
sufficient capacity for the number of instances you selected. If there is no matching Capacity
Reservation with sufficient capacity, the instance uses On-Demand capacity.
• Target by ID — Launches the instances into the selected Capacity Reservation. If the selected
Capacity Reservation does not have sufficient capacity for the number of instances you selected,
the instance launch fails.
• Target by group — Launches the instances into any Capacity Reservation with matching
attributes and available capacity in the selected Capacity Reservation group. If the selected
group does not have a Capacity Reservation with matching attributes and available capacity, the
instances launch into On-Demand capacity.
4. Complete the remaining steps to launch the instances.

To launch an instance into an existing Capacity Reservation using the AWS CLI

Use the run-instances command and specify the --capacity-reservation-specification

parameter.

The following example launches a t2.micro instance into any open Capacity Reservation that has
matching attributes and available capacity:

aws ec2 run-instances --image-id ami-abc12345 --count 1 --instance-type t2.micro --key-

name MyKeyPair --subnet-id subnet-1234567890abcdef1 --capacity-reservation-specification
CapacityReservationPreference=open

The following example launches a t2.micro instance into a targeted Capacity Reservation:

aws ec2 run-instances --image-id ami-abc12345 --count 1 --instance-type t2.micro --key-

name MyKeyPair --subnet-id subnet-1234567890abcdef1 --capacity-reservation-specification
CapacityReservationTarget={CapacityReservationId=cr-a1234567}

The following example launches a t2.micro instance into a Capacity Reservation group:

aws ec2 run-instances --image-id ami-abc12345 --count 1 --instance-type t2.micro --key-

name MyKeyPair --subnet-id subnet-1234567890abcdef1 --capacity-reservation-specification
CapacityReservationTarget={CapacityReservationResourceGroupArn=arn:aws:resource-groups:us-
west-1:123456789012:group/my-cr-group}

Modify a Capacity Reservation

You can change the attributes of an active Capacity Reservation after you have created it. You cannot
modify a Capacity Reservation after it has expired or after you have explicitly canceled it.

When modifying a Capacity Reservation, you can only increase or decrease the quantity and change
the way in which it is released. You cannot change the instance type, EBS optimization, instance store
settings, platform, Availability Zone, or instance eligibility of a Capacity Reservation. If you need to

510
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

modify any of these attributes, we recommend that you cancel the reservation, and then create a new
one with the required attributes.

If you specify a new quantity that exceeds your remaining On-Demand Instance limit for the selected
instance type, the update fails.

To modify a Capacity Reservation using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Capacity Reservations, select the Capacity Reservation to modify, and then choose Edit.
3. Modify the Quantity or Reservation ends options as needed, and choose Save changes.

To modify a Capacity Reservation using the AWS CLI

Use the modify-capacity-reservations command:

For example, the following command modifies a Capacity Reservation to reserve capacity for eight
instances.

aws ec2 modify-capacity-reservation --capacity-reservation-id cr-1234567890abcdef0 --

instance-count 8

Modify an instance's Capacity Reservation settings

You can modify the following Capacity Reservation settings for a stopped instance at any time:

• Start in any Capacity Reservation that has matching attributes (instance type, platform, and
Availability Zone) and available capacity.
• Start the instance in a specific Capacity Reservation.
• Start the in any Capacity Reservation that has matching attributes and available capacity in a Capacity
Reservation group
• Prevent the instance from starting in a Capacity Reservation.

To modify an instance's Capacity Reservation settings using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Instances and select the instance to modify. Stop the instance if it is not already stopped.
3. Choose Actions, Modify Capacity Reservation Settings.
4. For Capacity Reservation, choose one of the following options:

• Open — Launches the instances into any Capacity Reservation that has matching attributes and
sufficient capacity for the number of instances you selected. If there is no matching Capacity
Reservation with sufficient capacity, the instance uses On-Demand capacity.
• None — Prevents the instances from launching into a Capacity Reservation. The instances run in
On-Demand capacity.
• Specify Capacity Reservation — Launches the instances into the selected Capacity Reservation.
If the selected Capacity Reservation does not have sufficient capacity for the number of instances
you selected, the instance launch fails.
• Specify Capacity Reservation group — Launches the instances into any Capacity Reservation
with matching attributes and available capacity in the selected Capacity Reservation group. If
the selected group does not have a Capacity Reservation with matching attributes and available
capacity, the instances launch into On-Demand capacity.

511
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

To modify an instance's Capacity Reservation settings using the AWS CLI

Use the modify-instance-capacity-reservation-attributes command.

For example, the following command changes an instance's Capacity Reservation setting to open or
none.

aws ec2 modify-instance-capacity-reservation-attributes --instance-id i-1234567890abcdef0

--capacity-reservation-specification CapacityReservationPreference=none|open

For example, the following command modifies an instance to target a specific Capacity Reservation.

aws ec2 modify-instance-capacity-reservation-attributes --instance-

id i-1234567890abcdef0 --capacity-reservation-specification
CapacityReservationTarget={CapacityReservationId=cr-1234567890abcdef0}

For example, the following command modifies an instance to target a specific Capacity Reservation
group.

aws ec2 modify-instance-capacity-reservation-attributes --instance-

id i-1234567890abcdef0 --capacity-reservation-specification
CapacityReservationTarget={CapacityReservationResourceGroupArn=arn:aws:resource-groups:us-
west-1:123456789012:group/my-cr-group}

View a Capacity Reservation

Capacity Reservations have the following possible states:

• active—The capacity is available for use.

• expired—The Capacity Reservation expired automatically at the date and time specified in your
reservation request. The reserved capacity is no longer available for your use.
• cancelled—The Capacity Reservation was manually canceled. The reserved capacity is no longer
available for your use.
• pending—The Capacity Reservation request was successful but the capacity provisioning is still
pending.
• failed—The Capacity Reservation request has failed. A request can fail due to invalid request
parameters, capacity constraints, or instance limit constraints. You can view a failed request for 60
minutes.

To view your Capacity Reservations using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Capacity Reservations and select a Capacity Reservation to view.
3. Choose View launched instances for this reservation.

To view your Capacity Reservations using the AWS CLI

Use the describe-capacity-reservations command:

For example, the following command describes all Capacity Reservations.

aws ec2 describe-capacity-reservations

512
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

Cancel a Capacity Reservation

You can cancel a Capacity Reservation at any time if you no longer need the reserved capacity. When you
cancel a Capacity Reservation, the capacity is released immediately, and it is no longer reserved for your
use.

You can cancel empty Capacity Reservations and Capacity Reservations that have running instances. If
you cancel a Capacity Reservation that has running instances, the instances continue to run normally
outside of the capacity reservation at standard On-Demand Instance rates or at a discounted rate if you
have a matching Savings Plan or regional Reserved Instance.

After you cancel a Capacity Reservation, instances that target it can no longer launch. Modify these
instances so that they either target a different Capacity Reservation, launch into any open Capacity
Reservation with matching attributes and sufficient capacity, or avoid launching into a Capacity
Reservation. For more information, see Modify an instance's Capacity Reservation settings (p. 511).

To cancel a Capacity Reservation using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Capacity Reservations and select the Capacity Reservation to cancel.
3. Choose Cancel reservation, Cancel reservation.

To cancel a Capacity Reservation using the AWS CLI

Use the cancel-capacity-reservation command:

For example, the following command cancels a Capacity Reservation with an ID of

cr-1234567890abcdef0.

aws ec2 cancel-capacity-reservation --capacity-reservation-id cr-1234567890abcdef0

Capacity Reservations in Local Zones

A Local Zone is an extension of an AWS Region that is geographically close to your users. Resources
created in a Local Zone can serve local users with very low-latency communications. For more
information, see AWS Local Zones.

You can extend a VPC from its parent AWS Region into a Local Zone by creating a new subnet in that
Local Zone. When you create a subnet in a Local Zone, your VPC is extended to that Local Zone. The
subnet in the Local Zone operates the same as the other subnets in your VPC.

By using Local Zones, you can place Capacity Reservations in multiple locations that are closer to your
users. You create and use Capacity Reservations in Local Zones in the same way that you create and use
Capacity Reservations in regular Availability Zones. The same features and instance matching behavior
apply. For more information about the pricing models that are supported in Local Zones, see AWS Local
Zones FAQs.

Limitations

• You can't share Capacity Reservations that are created in a Local Zone.
• You can't use Capacity Reservation groups in a Local Zone.

To use a Capacity Reservation in a Local Zone

1. Enable the Local Zone for use in your AWS account. For more information, see Enable Local Zones in
the Amazon EC2 User Guide for Linux Instances.

513
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

2. Create a Capacity Reservation in the Local Zone. For Availability Zone, choose the Local Zone.
The Local Zone is represented by an AWS Region code followed by an identifier that indicates
the location, for example us-west-2-lax-1a. For more information, see Create a Capacity
Reservation (p. 505).
3. Create a subnet in the Local Zone. For Availability Zone, choose the Local Zone. For more
information, see Creating a subnet in your VPC in the Amazon VPC User Guide.
4. Launch an instance. For Subnet, choose the subnet in the Local Zone (for example subnet-123abc
| us-west-2-lax-1a), and for Capacity Reservation, choose the specification (either open or
target it by ID) that's required for the Capacity Reservation that you created in the Local Zone. For
more information, see Launch instances into an existing Capacity Reservation (p. 509).

Capacity Reservations in Wavelength Zones

AWS Wavelength enables developers to build applications that deliver ultra-low latencies to mobile
devices and end users. Wavelength deploys standard AWS compute and storage services to the edge of
telecommunication carriers' 5G networks. You can extend a Amazon Virtual Private Cloud (VPC) to one or
more Wavelength Zones. You can then use AWS resources like Amazon EC2 instances to run applications
that require ultra-low latency and a connection to AWS services in the Region. For more information, see
AWS Wavelength Zones.

When you create On-Demand Capacity Reservations, you can choose the Wavelength Zone and you can
launch instances into a Capacity Reservation in a Wavelength Zone by specifying the subnet associated
with the Wavelength Zone. A Wavelength Zone is represented by an AWS Region code followed by an
identifier that indicates the location, for example us-east-1-wl1-bos-wlz-1.

Wavelength Zones are not available in every Region. For information about the Regions that support
Wavelength Zones, see Available Wavelength Zones in the AWS Wavelength Developer Guide.

Considerations

• You can't share Capacity Reservations that are created in a Wavelength Zone.
• You can't use Capacity Reservation groups in a Wavelength Zone.

To use a Capacity Reservation in a Wavelength Zone

1. Enable the Wavelength Zone for use in your AWS account. For more information, see Enable
Wavelength Zones in the Amazon EC2 User Guide for Linux Instances.
2. Create a Capacity Reservation in the Wavelength Zone. For Availability Zone, choose the
Wavelength. The Wavelength is represented by an AWS Region code followed by an identifier that
indicates the location, for example us-east-1-wl1-bos-wlz-1. For more information, see Create
a Capacity Reservation (p. 505).
3. Create a subnet in the Wavelength Zone. For Availability Zone, choose the Wavelength Zone. For
more information, see Creating a subnet in your VPC in the Amazon VPC User Guide.
4. Launch an instance. For Subnet, choose the subnet in the Wavelength Zone (for example
subnet-123abc | us-east-1-wl1-bos-wlz-1), and for Capacity Reservation, choose the
specification (either open or target it by ID) that's required for the Capacity Reservation that you
created in the Wavelength. For more information, see Launch instances into an existing Capacity
Reservation (p. 509).

Work with shared Capacity Reservations

Capacity Reservation sharing enables Capacity Reservation owners to share their reserved capacity with
other AWS accounts or within an AWS organization. This enables you to create and manage Capacity

514
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

Reservations centrally, and share the reserved capacity across multiple AWS accounts or within your AWS
organization.

In this model, the AWS account that owns the Capacity Reservation (owner) shares it with other AWS
accounts (consumers). Consumers can launch instances into Capacity Reservations that are shared with
them in the same way that they launch instances into Capacity Reservations that they own in their own
account. The Capacity Reservation owner is responsible for managing the Capacity Reservation and the
instances that they launch into it. Owners cannot modify instances that consumers launch into Capacity
Reservations that they have shared. Consumers are responsible for managing the instances that they
launch into Capacity Reservations shared with them. Consumers cannot view or modify instances owned
by other consumers or by the Capacity Reservation owner.

A Capacity Reservation owner can share a Capacity Reservation with:

• Specific AWS accounts inside or outside of its AWS organization

• An organizational unit inside its AWS organization
• Its entire AWS organization

Contents
• Prerequisites for sharing Capacity Reservations (p. 515)
• Related services (p. 515)
• Share across Availability Zones (p. 516)
• Share a Capacity Reservation (p. 516)
• Stop sharing a Capacity Reservation (p. 517)
• Identify a shared Capacity Reservation (p. 517)
• View shared Capacity Reservation usage (p. 518)
• Shared Capacity Reservation permissions (p. 518)
• Billing and metering (p. 518)
• Instance limits (p. 518)

Prerequisites for sharing Capacity Reservations

• To share a Capacity Reservation, you must own it in your AWS account. You cannot share a Capacity
Reservation that has been shared with you.
• You can only share Capacity Reservations for shared tenancy instances. You cannot share Capacity
Reservations for dedicated tenancy instances.
• Capacity Reservation sharing is not available to new AWS accounts or AWS accounts that have a
limited billing history. New accounts that are linked to a qualified payer account or are linked through
an AWS organization are exempt from this restriction.
• To share a Capacity Reservation with your AWS organization or an organizational unit in your AWS
organization, you must enable sharing with AWS Organizations. For more information, see Enable
Sharing with AWS Organizations in the AWS RAM User Guide.

Related services
Capacity Reservation sharing integrates with AWS Resource Access Manager (AWS RAM). AWS RAM
is a service that enables you to share your AWS resources with any AWS account or through AWS
Organizations. With AWS RAM, you share resources that you own by creating a resource share. A resource
share specifies the resources to share, and the consumers with whom to share them. Consumers can be
individual AWS accounts, or organizational units or an entire organization from AWS Organizations.

515
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

For more information about AWS RAM, see the AWS RAM User Guide.

Share across Availability Zones

To ensure that resources are distributed across the Availability Zones for a Region, we independently
map Availability Zones to names for each account. This could lead to Availability Zone naming
differences across accounts. For example, the Availability Zone us-east-1a for your AWS account might
not have the same location as us-east-1a for another AWS account.

To identify the location of your Capacity Reservations relative to your accounts, you must use the
Availability Zone ID (AZ ID). The AZ ID is a unique and consistent identifier for an Availability Zone across
all AWS accounts. For example, use1-az1 is an AZ ID for the us-east-1 Region and it is the same
location in every AWS account.

To view the AZ IDs for the Availability Zones in your account

1. Open the AWS RAM console at https://console.aws.amazon.com/ram.

2. The AZ IDs for the current Region are displayed in the Your AZ ID panel on the right-hand side of the
screen.

Share a Capacity Reservation

When you share a Capacity Reservation that you own with other AWS accounts, you enable them to
launch instances into your reserved capacity. If you share an open Capacity Reservation, keep the
following in mind as it could lead to unintended Capacity Reservation usage:

• If consumers have running instances that match the attributes of the Capacity Reservation, have the
CapacityReservationPreference parameter set to open, and are not yet running in reserved
capacity, they automatically use the shared Capacity Reservation.
• If consumers launch instances that have matching attributes (instance type, platform, and Availability
Zone) and have the CapacityReservationPreference parameter set to open, they automatically
launch into the shared Capacity Reservation.

To share a Capacity Reservation, you must add it to a resource share. A resource share is an AWS RAM
resource that lets you share your resources across AWS accounts. A resource share specifies the resources
to share, and the consumers with whom they are shared. When you share a Capacity Reservation using
the Amazon EC2 console, you add it to an existing resource share. To add the Capacity Reservation to a
new resource share, you must create the resource share using the AWS RAM console.

If you are part of an organization in AWS Organizations and sharing within your organization is enabled,
consumers in your organization are automatically granted access to the shared Capacity Reservation.
Otherwise, consumers receive an invitation to join the resource share and are granted access to the
shared Capacity Reservation after accepting the invitation.

You can share a Capacity Reservation that you own using the Amazon EC2 console, AWS RAM console, or
the AWS CLI.

To share a Capacity Reservation that you own using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Capacity Reservations.
3. Choose the Capacity Reservation to share and choose Actions, Share reservation.
4. Select the resource share to which to add the Capacity Reservation and choose Share Capacity
Reservation.

It could take a few minutes for consumers to get access to the shared Capacity Reservation.

516
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

To share a Capacity Reservation that you own using the AWS RAM console

See Creating a Resource Share in the AWS RAM User Guide.

To share a Capacity Reservation that you own using the AWS CLI

Use the create-resource-share command.

Stop sharing a Capacity Reservation

The Capacity Reservation owner can stop sharing a Capacity Reservation at any time. The following rules
apply:

• Instances owned by consumers that were running in the shared capacity at the time sharing stops
continue to run normally outside of the reserved capacity, and the capacity is restored to the Capacity
Reservation subject to Amazon EC2 capacity availability.
• Consumers with whom the Capacity Reservation was shared can no longer launch new instances into
the reserved capacity.

To stop sharing a Capacity Reservation that you own, you must remove it from the resource share. You
can do this using the Amazon EC2 console, AWS RAM console, or the AWS CLI.

To stop sharing a Capacity Reservation that you own using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Capacity Reservations.
3. Select the Capacity Reservation and choose the Sharing tab.
4. The Sharing tab lists the resource shares to which the Capacity Reservation has been added. Select
the resource share from which to remove the Capacity Reservation and choose Remove from
resource share.

To stop sharing a Capacity Reservation that you own using the AWS RAM console

See Updating a Resource Share in the AWS RAM User Guide.

To stop sharing a Capacity Reservation that you own using the AWS CLI

Use the disassociate-resource-share command.

Identify a shared Capacity Reservation

Owners and consumers can identify shared Capacity Reservations using the Amazon EC2 console and
AWS CLI

To identify a shared Capacity Reservation using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Capacity Reservations. The screen lists Capacity Reservations that
you own and Capacity Reservations that are shared with you. The Owner column shows the AWS
account ID of the Capacity Reservation owner. (me) next to the AWS account ID indicates that you
are the owner.

To identify a shared Capacity Reservation using the AWS CLI

Use the describe-capacity-reservations command. The command returns the Capacity Reservations that
you own and Capacity Reservations that are shared with you. OwnerId shows the AWS account ID of the
Capacity Reservation owner.

517
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

View shared Capacity Reservation usage

The owner of a shared Capacity Reservation can view its usage at any time using the Amazon EC2
console and the AWS CLI.

To view Capacity Reservation usage using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Capacity Reservations.
3. Select the Capacity Reservation for which to view the usage and choose the Usage tab.

The AWS account ID column shows the account IDs of the consumers currently using the Capacity
Reservation. The Launched instances column shows the number of instances each consumer
currently has running in the reserved capacity.

To view Capacity Reservation usage using the AWS CLI

Use the get-capacity-reservation-usage command. AccountId shows the account ID of the account
using the Capacity Reservation. UsedInstanceCount shows the number of instances the consumer
currently has running in the reserved capacity.

Shared Capacity Reservation permissions

Permissions for owners

Owners are responsible for managing and canceling their shared Capacity Reservations. Owners cannot
modify instances running in the shared Capacity Reservation that are owned by other accounts. Owners
remain responsible for managing instances that they launch into the shared Capacity Reservation.

Permissions for consumers

Consumers are responsible for managing their instances that are running the shared Capacity
Reservation. Consumers cannot modify the shared Capacity Reservation in any way, and they cannot
view or modify instances that are owned by other consumers or the Capacity Reservation owner.

Billing and metering

There are no additional charges for sharing Capacity Reservations.

The Capacity Reservation owner is billed for instances that they run inside the Capacity Reservation
and for unused reserved capacity. Consumers are billed for the instances that they run inside the shared
Capacity Reservation.

Instance limits
All Capacity Reservation usage counts toward the Capacity Reservation owner's On-Demand Instance
limits. This includes:

• Unused reserved capacity

• Usage by instances owned by the Capacity Reservation owner
• Usage by instances owned by consumers

Instances launched into the shared capacity by consumers count towards the Capacity Reservation
owner's On-Demand Instance limit. Consumers' instance limits are a sum of their own On-Demand
Instance limits and the capacity available in the shared Capacity Reservations to which they have access.

518
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
On-Demand Capacity Reservations

CloudWatch metrics for On-Demand Capacity Reservations

With CloudWatch metrics, you can efficiently monitor your Capacity Reservations and identify unused
capacity by setting CloudWatch alarms to notify you when usage thresholds are met. This can help you
maintain a constant Capacity Reservation volume and achieve a higher level of utilization.

On-Demand Capacity Reservations send metric data to CloudWatch every five minutes. Metrics are not
supported for Capacity Reservations that are active for less than five minutes.

For more information about viewing metrics in the CloudWatch console, see Using Amazon CloudWatch
Metrics. For more information about creating alarms, see Creating Amazon CloudWatch Alarms.

Contents
• Capacity Reservation usage metrics (p. 519)
• Capacity Reservation metric dimensions (p. 519)
• View CloudWatch metrics for Capacity Reservations (p. 519)

Capacity Reservation usage metrics

The AWS/EC2CapacityReservations namespace includes the following usage metrics you can use to
monitor and maintain on-demand capacity within thresholds you specify for your reservation.

Metric Description

UsedInstanceCount The number of instances that are currently in use.

Unit: Count

AvailableInstanceCount The number of instances that are available.

Unit: Count

TotalInstanceCount The total number of instances you have reserved.

Unit: Count

InstanceUtilization The percentage of reserved capacity instances that are

currently in use.

Unit: Percent

Capacity Reservation metric dimensions

You can use the following dimensions to refine the metrics listed in the previous table.

Dimension Description

CapacityReservationId This globally unique dimension filters the data you request for the
identified capacity reservation only.

View CloudWatch metrics for Capacity Reservations

Metrics are grouped first by the service namespace, and then by the supported dimensions. You can use
the following procedures to view the metrics for your Capacity Reservations.

519
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance lifecycle

To view Capacity Reservation metrics using the CloudWatch console

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. If necessary, change the Region. From the navigation bar, select the Region where your Capacity
Reservation resides. For more information, see Regions and Endpoints.
3. In the navigation pane, choose Metrics.
4. For All metrics, choose EC2 Capacity Reservations.
5. Choose the metric dimension By Capacity Reservation. Metrics will be grouped by
CapacityReservationId.
6. To sort the metrics, use the column heading. To graph a metric, select the check box next to the
metric.

To view Capacity Reservation metrics (AWS CLI)

Use the following list-metrics command:

aws cloudwatch list-metrics --namespace "AWS/EC2CapacityReservations"

Instance lifecycle
An Amazon EC2 instance transitions through different states from the moment you launch it through to
its termination.

The following illustration represents the transitions between instance states. Notice that you can't
stop and start an instance store-backed instance. For more information about instance store-backed
instances, see Storage for the root device (p. 102).

520
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance launch

The following table provides a brief description of each instance state and indicates whether it is billed
or not.
Note
The table indicates billing for instance usage only. Some AWS resources, such as Amazon EBS
volumes and Elastic IP addresses, incur charges regardless of the instance's state. For more
information, see Avoiding Unexpected Charges in the AWS Billing and Cost Management User
Guide.

Instance Description Instance usage billing

state

pending The instance is preparing to Not billed

enter the running state. An
instance enters the pending
state when it launches for
the first time, or when it is
started after being in the
stopped state.

running The instance is running and Billed

ready for use.

stoppingThe instance is preparing Not billed if preparing to stop

to be stopped or stop-
hibernated. Billed if preparing to hibernate

stopped The instance is shut down Not billed

and cannot be used. The
instance can be started at
any time.

The instance is preparing to

shutting- Not billed
down be terminated.

The instance has been

terminated Not billed
permanently deleted and
Note
cannot be started.
Reserved Instances that applied to terminated
instances are billed until the end of their term
according to their payment option. For more
information, see Reserved Instances (p. 325)

Note
Rebooting an instance doesn't start a new instance billing period because the instance stays in
the running state.

Instance launch
When you launch an instance, it enters the pending state. The instance type that you specified at launch
determines the hardware of the host computer for your instance. We use the Amazon Machine Image
(AMI) you specified at launch to boot the instance. After the instance is ready for you, it enters the
running state. You can connect to your running instance and use it the way that you'd use a computer
sitting in front of you.

As soon as your instance transitions to the running state, you're billed for each second, with a one-
minute minimum, that you keep the instance running, even if the instance remains idle and you don't
connect to it.

521
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance stop and start (Amazon
EBS-backed instances only)
For more information, see Launch your instance (p. 525) and Connect to your Linux instance (p. 598).

Instance stop and start (Amazon EBS-backed

instances only)
If your instance fails a status check or is not running your applications as expected, and if the root
volume of your instance is an Amazon EBS volume, you can stop and start your instance to try to fix the
problem.

When you stop your instance, it enters the stopping state, and then the stopped state. We don't
charge usage or data transfer fees for your instance after you stop it, but we do charge for the storage
for any Amazon EBS volumes. While your instance is in the stopped state, you can modify certain
attributes of the instance, including the instance type.

When you start your instance, it enters the pending state, and we move the instance to a new host
computer (though in some cases, it remains on the current host). When you stop and start your instance,
you lose any data on the instance store volumes on the previous host computer.

Your instance retains its private IPv4 address, which means that an Elastic IP address associated with the
private IPv4 address or network interface is still associated with your instance. If your instance has an
IPv6 address, it retains its IPv6 address.

Each time you transition an instance from stopped to running, we charge per second when the
instance is running, with a minimum of one minute every time you start your instance.

For more information, see Stop and start your instance (p. 623).

Instance hibernate (Amazon EBS-backed instances

only)
When you hibernate an instance, we signal the operating system to perform hibernation (suspend-to-
disk), which saves the contents from the instance memory (RAM) to your Amazon EBS root volume. We
persist the instance's Amazon EBS root volume and any attached Amazon EBS data volumes. When you
start your instance, the Amazon EBS root volume is restored to its previous state and the RAM contents
are reloaded. Previously attached data volumes are reattached and the instance retains its instance ID.

When you hibernate your instance, it enters the stopping state, and then the stopped state. We don't
charge usage for a hibernated instance when it is in the stopped state, but we do charge while it is in
the stopping state, unlike when you stop an instance (p. 522) without hibernating it. We don't charge
usage for data transfer fees, but we do charge for the storage for any Amazon EBS volumes, including
storage for the RAM data.

When you start your hibernated instance, it enters the pending state, and we move the instance to a
new host computer (though in some cases, it remains on the current host).

Your instance retains its private IPv4 address, which means that an Elastic IP address associated with the
private IPv4 address or network interface is still associated with your instance. If your instance has an
IPv6 address, it retains its IPv6 address.

For more information, see Hibernate your Linux instance (p. 627).

Instance reboot
You can reboot your instance using the Amazon EC2 console, a command line tool, and the Amazon EC2
API. We recommend that you use Amazon EC2 to reboot your instance instead of running the operating
system reboot command from your instance.

522
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance retirement

Rebooting an instance is equivalent to rebooting an operating system. The instance remains on the same
host computer and maintains its public DNS name, private IP address, and any data on its instance store
volumes. It typically takes a few minutes for the reboot to complete, but the time it takes to reboot
depends on the instance configuration.

Rebooting an instance doesn't start a new instance billing period; per second billing continues without a
further one-minute minimum charge.

For more information, see Reboot your instance (p. 639).

Instance retirement
An instance is scheduled to be retired when AWS detects the irreparable failure of the underlying
hardware hosting the instance. When an instance reaches its scheduled retirement date, it is stopped or
terminated by AWS. If your instance root device is an Amazon EBS volume, the instance is stopped, and
you can start it again at any time. If your instance root device is an instance store volume, the instance is
terminated, and cannot be used again.

For more information, see Instance retirement (p. 640).

Instance termination
When you've decided that you no longer need an instance, you can terminate it. As soon as the status of
an instance changes to shutting-down or terminated, you stop incurring charges for that instance.

If you enable termination protection, you can't terminate the instance using the console, CLI, or API.

After you terminate an instance, it remains visible in the console for a short while, and then the entry
is automatically deleted. You can also describe a terminated instance using the CLI and API. Resources
(such as tags) are gradually disassociated from the terminated instance, therefore may no longer be
visible on the terminated instance after a short while. You can't connect to or recover a terminated
instance.

Each Amazon EBS-backed instance supports the InstanceInitiatedShutdownBehavior attribute,

which controls whether the instance stops or terminates when you initiate shutdown from within the
instance itself (for example, by using the shutdown command on Linux). The default behavior is to stop
the instance. You can modify the setting of this attribute while the instance is running or stopped.

Each Amazon EBS volume supports the DeleteOnTermination attribute, which controls whether the
volume is deleted or preserved when you terminate the instance it is attached to. The default is to delete
the root device volume and preserve any other EBS volumes.

For more information, see Terminate your instance (p. 643).

Differences between reboot, stop, hibernate, and

terminate
The following table summarizes the key differences between rebooting, stopping, hibernating, and
terminating your instance.

Characteristic Reboot Stop/start (Amazon Hibernate (Amazon Terminate

EBS-backed EBS-backed instances
instances only) only)

Host The instance stays We move the instance We move the instance None
computer on the same host to a new host to a new host
computer computer (though computer (though in

523
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Differences between reboot, stop, hibernate, and terminate

Characteristic Reboot Stop/start (Amazon Hibernate (Amazon Terminate

EBS-backed EBS-backed instances
instances only) only)
in some cases, it some cases, it remains
remains on the on the current host).
current host).

Private and These addresses The instance keeps its The instance keeps its None
public IPv4 stay the same private IPv4 address. private IPv4 address.
addresses The instance gets The instance gets
a new public IPv4 a new public IPv4
address, unless it has address, unless it has
an Elastic IP address, an Elastic IP address,
which doesn't change which doesn't change
during a stop/start. during a stop/start.

Elastic IP The Elastic IP The Elastic IP address The Elastic IP address The Elastic
addresses address remains remains associated remains associated IP address is
(IPv4) associated with with the instance with the instance disassociated from
the instance the instance

IPv6 The address stays The instance keeps its The instance keeps its None
address the same IPv6 address IPv6 address

Instance The data is The data is erased The data is erased The data is erased
store preserved
volumes

Root device The volume is The volume is The volume is The volume is
volume preserved preserved preserved deleted by default

RAM The RAM is erased The RAM is erased The RAM is saved The RAM is erased
(contents of to a file on the root
memory) volume

Billing The instance You stop incurring You incur charges You stop incurring
billing hour charges for an while the instance is in charges for
doesn't change. instance as soon as the stopping state, an instance
its state changes but stop incurring as soon as its
to stopping. charges when the state changes to
Each time an instance is in the shutting-down.
instance transitions stopped state. Each
from stopped to time an instance
running, we start a transitions from
new instance billing stopped to running,
period, billing a we start a new
minimum of one instance billing period,
minute every time billing a minimum
you start your of one minute every
instance. time you start your
instance.

Operating system shutdown commands always terminate an instance store-backed instance. You can
control whether operating system shutdown commands stop or terminate an Amazon EBS-backed
instance. For more information, see Change the instance initiated shutdown behavior (p. 646).

524
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Launch your instance

An instance is a virtual server in the AWS Cloud. You launch an instance from an Amazon Machine Image
(AMI). The AMI provides the operating system, application server, and applications for your instance.

When you sign up for AWS, you can get started with Amazon EC2 for free using the AWS Free Tier. You
can use the free tier to launch and use a t2.micro instance for free for 12 months (in Regions where
t2.micro is unavailable, you can use a t3.micro instance under the free tier). If you launch an instance
that is not within the free tier, you incur the standard Amazon EC2 usage fees for the instance. For more
information, see Amazon EC2 pricing.

You can launch an instance using the following methods.

Method Documentation

[Amazon EC2 console] Use the launch instance Launch an instance using the Launch Instance
wizard to specify the launch parameters. Wizard (p. 526)

[Amazon EC2 console] Create a launch template Launch an instance from a launch
and launch the instance from the launch template (p. 532)
template.

[Amazon EC2 console] Use an existing instance as Launch an instance using parameters from an
the base. existing instance (p. 545)

[Amazon EC2 console] Use an AMI that you Launch an AWS Marketplace instance (p. 546)
purchased from the AWS Marketplace.

[AWS CLI] Use an AMI that you select. Using Amazon EC2 through the AWS CLI

[AWS Tools for Windows PowerShell] Use an AMI Amazon EC2 from the AWS Tools for Windows
that you select. PowerShell

[AWS CLI] Use EC2 Fleet to provision capacity Launch instances using an EC2 Fleet (p. 548)
across different EC2 instance types and
Availability Zones, and across On-Demand
Instance, Reserved Instance, and Spot Instance
purchase models.

[AWS CloudFormation] Use a AWS AWS::EC2::Instance in the AWS CloudFormation

CloudFormation template to specify an instance. User Guide

[AWS SDK] Use a language-specific AWS SDK to AWS SDK for .NET
launch an instance.
AWS SDK for C++

AWS SDK for Go

AWS SDK for Java

AWS SDK for JavaScript

AWS SDK for PHP V3

AWS SDK for Python

AWS SDK for Ruby V3

525
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

When you launch your instance, you can launch your instance in a subnet that is associated with one of
the following resources:

• An Availability Zone - This option is the default.

• A Local Zone - To launch an instance in a Local Zone, you must opt in to the Local Zone, and then
create a subnet in the zone. For more information, see Local Zones
• A Wavelength Zone - To launch an instance in a Wavelength Zone, you must opt in to the Wavelength
Zone, and then create a subnet in the zone. For information about how to launch an instance in a
Wavelength Zone, see Get started with AWS Wavelength in the AWS Wavelength Developer Guide.
• An Outpost - To launch an instance in an Outpost, you must create an Outpost. For information about
how to create an Outpost, see Get Started with AWS Outposts in the AWS Outposts User Guide.

After you launch your instance, you can connect to it and use it. To begin, the instance state is pending.
When the instance state is running, the instance has started booting. There might be a short time
before you can connect to the instance. Note that bare metal instance types might take longer to launch.
For more information about bare metal instances, see Instances built on the Nitro System (p. 209).

The instance receives a public DNS name that you can use to contact the instance from the internet.
The instance also receives a private DNS name that other instances within the same VPC can use to
contact the instance. For more information about connecting to your instance, see Connect to your Linux
instance (p. 598).

When you are finished with an instance, be sure to terminate it. For more information, see Terminate
your instance (p. 643).

Launch an instance using the Launch Instance Wizard

You can launch an instance using the launch instance wizard. The launch instance wizard specifies all
the launch parameters required for launching an instance. Where the launch instance wizard provides a
default value, you can accept the default or specify your own value. At the very least, you need to select
an AMI and a key pair to launch an instance.

Before you launch your instance, be sure that you are set up. For more information, see Set up to use
Amazon EC2 (p. 28).
Important
When you launch an instance that's not within the AWS Free Tier, you are charged for the time
that the instance is running, even if it remains idle.

Steps to launch an instance:

• Initiate instance launch (p. 526)
• Step 1: Choose an Amazon Machine Image (AMI) (p. 527)
• Step 2: Choose an Instance Type (p. 528)
• Step 3: Configure Instance Details (p. 528)
• Step 4: Add Storage (p. 530)
• Step 5: Add Tags (p. 531)
• Step 6: Configure Security Group (p. 531)
• Step 7: Review Instance Launch and Select Key Pair (p. 531)

Initiate instance launch

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.
2. In the navigation bar at the top of the screen, the current Region is displayed (for example, US
East (Ohio)). Select a Region for the instance that meets your needs. This choice is important

526
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

because some Amazon EC2 resources can be shared between Regions, while others can't. For more
information, see Resource locations (p. 1304).
3. From the Amazon EC2 console dashboard, choose Launch instance.

Step 1: Choose an Amazon Machine Image (AMI)

When you launch an instance, you must select a configuration, known as an Amazon Machine Image
(AMI). An AMI contains the information required to create a new instance. For example, an AMI might
contain the software required to act as a web server, such as Linux, Apache, and your website.

When you launch an instance, you can either select an AMI from the list, or you can select a Systems
Manager parameter that points to an AMI ID. For more information, see Using a Systems Manager
parameter to find an AMI.

On the Choose an Amazon Machine Image (AMI) page, use one of two options to choose an AMI. Either
search the list of AMIs (p. 527), or search by Systems Manager parameter (p. 527).

By searching the list of AMIs

1. Select the type of AMI to use in the left pane:

Quick Start

A selection of popular AMIs to help you get started quickly. To select an AMI that is eligible for
the free tier, choose Free tier only in the left pane. These AMIs are marked Free tier eligible.
My AMIs

The private AMIs that you own, or private AMIs that have been shared with you. To view AMIs
that are shared with you, choose Shared with me in the left pane.
AWS Marketplace

An online store where you can buy software that runs on AWS, including AMIs. For more
information about launching an instance from the AWS Marketplace, see Launch an AWS
Marketplace instance (p. 546).
Community AMIs

The AMIs that AWS community members have made available for others to use. To filter the list
of AMIs by operating system, choose the appropriate check box under Operating system. You
can also filter by architecture and root device type.
2. Check the Root device type listed for each AMI. Notice which AMIs are the type that you need, either
ebs (backed by Amazon EBS) or instance-store (backed by instance store). For more information,
see Storage for the root device (p. 102).
3. Check the Virtualization type listed for each AMI. Notice which AMIs are the type that you need,
either hvm or paravirtual. For example, some instance types require HVM. For more information,
see Linux AMI virtualization types (p. 104).
4. Choose an AMI that meets your needs, and then choose Select.

By Systems Manager parameter

1. Choose Search by Systems Manager parameter (at top right).

2. For Systems Manager parameter, select a parameter. The corresponding AMI ID appears next to
Currently resolves to.
3. Choose Search. The AMIs that match the AMI ID appear in the list.
4. Select the AMI from the list, and choose Select.

527
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Step 2: Choose an Instance Type

On the Choose an Instance Type page, select the hardware configuration and size of the instance
to launch. Larger instance types have more CPU and memory. For more information, see Instance
types (p. 203).

To remain eligible for the free tier, choose the t2.micro instance type (or the t3.micro instance
type in Regions where t2.micro is unavailable). For more information, see Burstable performance
instances (p. 224).

By default, the wizard displays current generation instance types, and selects the first available instance
type based on the AMI that you selected. To view previous generation instance types, choose All
generations from the filter list.
Note
To set up an instance quickly for testing purposes, choose Review and Launch to accept the
default configuration settings, and launch your instance. Otherwise, to configure your instance
further, choose Next: Configure Instance Details.

Step 3: Configure Instance Details

On the Configure Instance Details page, change the following settings as necessary (expand Advanced
Details to see all the settings), and then choose Next: Add Storage:

• Number of instances: Enter the number of instances to launch.

Tip
To ensure faster instance launches, break up large requests into smaller batches. For example,
create five separate launch requests for 100 instances each instead of one launch request for
500 instances.
• (Optional) To help ensure that you maintain the correct number of instances to handle demand on
your application, you can choose Launch into Auto Scaling Group to create a launch configuration
and an Auto Scaling group. Auto Scaling scales the number of instances in the group according to your
specifications. For more information, see the Amazon EC2 Auto Scaling User Guide.
Note
If Amazon EC2 Auto Scaling marks an instance that is in an Auto Scaling group as unhealthy,
the instance is automatically scheduled for replacement where it is terminated and another is
launched, and you lose your data on the original instance. An instance is marked as unhealthy
if you stop or reboot the instance, or if another event marks the instance as unhealthy. For
more information, see Health Checks for Auto Scaling Instances in the Amazon EC2 Auto
Scaling User Guide.
• Purchasing option: Choose Request Spot instances to launch a Spot Instance. This adds and removes
options from this page. Set your maximum price, and optionally update the request type, interruption
behavior, and request validity. For more information, see Create a Spot Instance request (p. 387).
• Network: Select the VPC, or to create a new VPC, choose Create new VPC to go the Amazon VPC
console. When you have finished, return to the wizard and choose Refresh to load your VPC in the list.
• Subnet: You can launch an instance in a subnet associated with an Availability Zone, Local Zone,
Wavelength Zone or Outpost.

To launch the instance in an Availability Zone, select the subnet into which to launch your instance.
You can select No preference to let AWS choose a default subnet in any Availability Zone. To create a
new subnet, choose Create new subnet to go to the Amazon VPC console. When you are done, return
to the wizard and choose Refresh to load your subnet in the list.

To launch the instance in a Local Zone, select a subnet that you created in the Local Zone.

To launch an instance in an Outpost, select a subnet in a VPC that you associated with an Outpost.

528
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

• Auto-assign Public IP: Specify whether your instance receives a public IPv4 address. By default,
instances in a default subnet receive a public IPv4 address and instances in a nondefault subnet do not.
You can select Enable or Disable to override the subnet's default setting. For more information, see
Public IPv4 addresses and external DNS hostnames (p. 810).
• Auto-assign IPv6 IP: Specify whether your instance receives an IPv6 address from the range of the
subnet. Select Enable or Disable to override the subnet's default setting. This option is only available
if you've associated an IPv6 CIDR block with your VPC and subnet. For more information, see Your VPC
and Subnets in the Amazon VPC User Guide.
• Domain join directory: Select the AWS Directory Service directory (domain) to which your Linux
instance is joined after launch. If you select a domain, you must select an IAM role with the required
permissions. For more information, see Seamlessly Join a Linux EC2 Instance to Your AWS Managed
Microsoft AD Directory.
• Placement group: A placement group determines the placement strategy of your instances. Select
an existing placement group, or create a new one. This option is only available if you've selected an
instance type that supports placement groups. For more information, see Placement groups (p. 927).
• Capacity Reservation: Specify whether to launch the instance into shared capacity, any open Capacity
Reservation, a specific Capacity Reservation, or a Capacity Reservation group. For more information,
see Launch instances into an existing Capacity Reservation (p. 509)..
• IAM role: Select an AWS Identity and Access Management (IAM) role to associate with the instance. For
more information, see IAM roles for Amazon EC2 (p. 1032).
• CPU options: Choose Specify CPU options to specify a custom number of vCPUs during launch.
Set the number of CPU cores and threads per core. For more information, see Optimize CPU
options (p. 668).
• Shutdown behavior: Select whether the instance should stop or terminate when shut down. For more
information, see Change the instance initiated shutdown behavior (p. 646).
• Stop - Hibernate behavior: To enable hibernation, select this check box. This option is only available
if your instance meets the hibernation prerequisites. For more information, see Hibernate your Linux
instance (p. 627).
• Enable termination protection: To prevent accidental termination, select this check box. For more
information, see Enable termination protection (p. 645).
• Monitoring: Select this check box to enable detailed monitoring of your instance using Amazon
CloudWatch. Additional charges apply. For more information, see Monitor your instances using
CloudWatch (p. 761).
• EBS-optimized instance: An Amazon EBS-optimized instance uses an optimized configuration stack
and provides additional, dedicated capacity for Amazon EBS I/O. If the instance type supports this
feature, select this check box to enable it. Additional charges apply. For more information, see Amazon
EBS–optimized instances (p. 1217).
• Tenancy: If you are launching your instance into a VPC, you can choose to run your instance on
isolated, dedicated hardware (Dedicated) or on a Dedicated Host (Dedicated host). Additional charges
may apply. For more information, see Dedicated Instances (p. 496) and Dedicated Hosts (p. 465).
• T2/T3 Unlimited: Select this check box to enable applications to burst beyond the baseline for as
long as needed. Additional charges may apply. For more information, see Burstable performance
instances (p. 224).
• File systems: To create a new file system to mount to your instance, choose Create new file system,
enter a name for the new file system, and then choose Create. The file system is created using Amazon
EFS Quick Create, which applies the service recommended settings. The security groups required to
enable access to the file system are automatically created and attached to the instance and the mount
targets of the file system. You can also choose to manually create and attach the required security
groups. For more information, see Create an EFS file system using Amazon EFS Quick Create (p. 1287).

To mount one or more existing Amazon EFS file systems to your instance, choose Add file system and
then choose the file systems to mount and the mount points to use. For more information, see Create
an EFS file system and mount it to your instance (p. 1288).

529
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

• Network interfaces: If you selected a specific subnet, you can specify up to two network interfaces for
your instance:
• For Network Interface, select New network interface to let AWS create a new interface, or select an
existing, available network interface.
• For Primary IP, enter a private IPv4 address from the range of your subnet, or leave Auto-assign to
let AWS choose a private IPv4 address for you.
• For Secondary IP addresses, choose Add IP to assign more than one private IPv4 address to the
selected network interface.
• (IPv6-only) For IPv6 IPs, choose Add IP, and enter an IPv6 address from the range of the subnet, or
leave Auto-assign to let AWS choose one for you.
• Network Card Index: The index of the network card. The primary network interface must be
assigned to network card index 0. Some instance types support multiple network cards.
• Choose Add Device to add a secondary network interface. A secondary network interface can reside
in a different subnet of the VPC, provided it's in the same Availability Zone as your instance.

For more information, see Elastic network interfaces (p. 839). If you specify more than one network
interface, your instance cannot receive a public IPv4 address. Additionally, if you specify an existing
network interface for eth0, you cannot override the subnet's public IPv4 setting using Auto-assign
Public IP. For more information, see Assign a public IPv4 address during instance launch (p. 814).
• Kernel ID: (Only valid for paravirtual (PV) AMIs) Select Use default unless you want to use a specific
kernel.
• RAM disk ID: (Only valid for paravirtual (PV) AMIs) Select Use default unless you want to use a specific
RAM disk. If you have selected a kernel, you may need to select a specific RAM disk with the drivers to
support it.
• Enclave: Select Enable to enable the instance for AWS Nitro Enclaves. For more information, see What
is AWS Nitro Enclaves? in the AWS Nitro Enclaves User Guide.
• Metadata accessible: You can enable or disable access to the instance metadata. For more
information, see Configure the instance metadata service (p. 705).
• Metadata version: If you enable access to the instance metadata, you can choose to require the use of
Instance Metadata Service Version 2 when requesting instance metadata. For more information, see
Configure instance metadata options for new instances (p. 709).
• Metadata token response hop limit: If you enable instance metadata, you can set the allowable
number of network hops for the metadata token. For more information, see Configure the instance
metadata service (p. 705).
• User data: You can specify user data to configure an instance during launch, or to run a configuration
script. To attach a file, select the As file option and browse for the file to attach.

Step 4: Add Storage

The AMI you selected includes one or more volumes of storage, including the root device volume. On the
Add Storage page, you can specify additional volumes to attach to the instance by choosing Add New
Volume. Configure each volume as follows, and then choose Next: Add Tags.

• Type: Select instance store or Amazon EBS volumes to associate with your instance. The types of
volume available in the list depend on the instance type you've chosen. For more information, see
Amazon EC2 instance store (p. 1268) and Amazon EBS volumes (p. 1082).
• Device: Select from the list of available device names for the volume.
• Snapshot: Enter the name or ID of the snapshot from which to restore a volume. You can also search
for available shared and public snapshots by typing text into the Snapshot field. Snapshot descriptions
are case-sensitive.
• Size: For EBS volumes, you can specify a storage size. Even if you have selected an AMI and
instance that are eligible for the free tier, to stay within the free tier, you must stay under 30 GiB

530
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

of total storage. For more information, see Constraints on the size and configuration of an EBS
volume (p. 1101).
• Volume Type: For EBS volumes, select a volume type. For more information, see Amazon EBS volume
types (p. 1084).
• IOPS: If you have selected a Provisioned IOPS SSD volume type, then you can enter the number of I/O
operations per second (IOPS) that the volume can support.
• Delete on Termination: For Amazon EBS volumes, select this check box to delete the volume when
the instance is terminated. For more information, see Preserve Amazon EBS volumes on instance
termination (p. 646).
• Encrypted: If the instance type supports EBS encryption, you can specify the encryption state of
the volume. If you have enabled encryption by default in this Region, the default CMK is selected
for you. You can select a different key or disable encryption. For more information, see Amazon EBS
encryption (p. 1176).

Step 5: Add Tags

On the Add Tags page, specify tags (p. 1311) by providing key and value combinations. You can tag the
instance, the volumes, or both. For Spot Instances, you can tag the Spot Instance request only. Choose
Add another tag to add more than one tag to your resources. Choose Next: Configure Security Group
when you are done.

Step 6: Configure Security Group

On the Configure Security Group page, use a security group to define firewall rules for your instance.
These rules specify which incoming network traffic is delivered to your instance. All other traffic is
ignored. (For more information about security groups, see Amazon EC2 security groups for Linux
instances (p. 1058).) Select or create a security group as follows, and then choose Review and Launch.

• To select an existing security group, choose Select an existing security group, and select your security
group. You can't edit the rules of an existing security group, but you can copy them to a new group by
choosing Copy to new. Then you can add rules as described in the next step.
• To create a new security group, choose Create a new security group. The wizard automatically defines
the launch-wizard-x security group and creates an inbound rule to allow you to connect to your
instance over SSH (port 22).
• You can add rules to suit your needs. For example, if your instance is a web server, open ports 80
(HTTP) and 443 (HTTPS) to allow internet traffic.

To add a rule, choose Add Rule, select the protocol to open to network traffic, and then specify the
source. Choose My IP from the Source list to let the wizard add your computer's public IP address.
However, if you are connecting through an ISP or from behind your firewall without a static IP address,
you need to find out the range of IP addresses used by client computers.
Warning
Rules that enable all IP addresses (0.0.0.0/0) to access your instance over SSH or RDP are
acceptable for this short exercise, but are unsafe for production environments. You should
authorize only a specific IP address or range of addresses to access your instance.

Step 7: Review Instance Launch and Select Key Pair

On the Review Instance Launch page, check the details of your instance, and make any necessary
changes by choosing the appropriate Edit link.

When you are ready, choose Launch.

In the Select an existing key pair or create a new key pair dialog box, you can choose an existing
key pair, or create a new one. For example, choose Choose an existing key pair, then select the key

531
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

pair you created when getting set up. For more information, see Amazon EC2 key pairs and Linux
instances (p. 1044).
Important
If you choose the Proceed without key pair option, you won't be able to connect to the instance
unless you choose an AMI that is configured to allow users another way to log in.

To launch your instance, select the acknowledgment check box, then choose Launch Instances.

(Optional) You can create a status check alarm for the instance (additional fees may apply). (If you're not
sure, you can always add one later.) On the confirmation screen, choose Create status check alarms and
follow the directions. For more information, see Create and edit status check alarms (p. 747).

If the instance fails to launch or the state immediately goes to terminated instead of running, see
Troubleshoot instance launch issues (p. 1326).

Launch an instance from a launch template

You can create a launch template that contains the configuration information to launch an instance.
Launch templates enable you to store launch parameters so that you do not have to specify them every
time you launch an instance. For example, a launch template can contain the AMI ID, instance type, and
network settings that you typically use to launch instances. When you launch an instance using the
Amazon EC2 console, an AWS SDK, or a command line tool, you can specify the launch template to use.

For each launch template, you can create one or more numbered launch template versions. Each version
can have different launch parameters. When you launch an instance from a launch template, you can
use any version of the launch template. If you do not specify a version, the default version is used. You
can set any version of the launch template as the default version—by default, it's the first version of the
launch template.

The following diagram shows a launch template with three versions. The first version specifies the
instance type, AMI ID, subnet, and key pair to use to launch the instance. The second version is based
on the first version and also specifies a security group for the instance. The third version uses different
values for some of the parameters. Version 2 is set as the default version. If you launched an instance
from this launch template, the launch parameters from version 2 would be used if no other version were
specified.

Contents
• Launch template restrictions (p. 533)
• Use launch templates to control launch parameters (p. 533)
• Control the use of launch templates (p. 533)
• Create a launch template (p. 534)
• Manage launch template versions (p. 539)

532
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

• Launch an instance from a launch template (p. 542)

• Use launch templates with Amazon EC2 Auto Scaling (p. 543)
• Use launch templates with EC2 Fleet (p. 544)
• Use launch templates with Spot Fleet (p. 544)
• Delete a launch template (p. 544)

Launch template restrictions

The following rules apply to launch templates and launch template versions:

• You are limited to creating 5,000 launch templates per Region and 10,000 versions per launch
template.
• Launch template parameters are optional. However, you must ensure that your request to launch an
instance includes all required parameters. For example, if your launch template does not include an
AMI ID, you must specify both the launch template and an AMI ID when you launch an instance.
• Launch template parameters are not fully validated when you create the launch template. If you
specify incorrect values for parameters, or if you do not use supported parameter combinations, no
instances can launch using this launch template. Ensure that you specify the correct values for the
parameters and that you use supported parameter combinations. For example, to launch an instance in
a placement group, you must specify a supported instance type.
• You can tag a launch template, but you cannot tag a launch template version.
• Launch template versions are numbered in the order in which they are created. When you create a
launch template version, you cannot specify the version number yourself.

Use launch templates to control launch parameters

A launch template can contain all or some of the parameters to launch an instance. When you launch an
instance using a launch template, you can override parameters that are specified in the launch template.
Or, you can specify additional parameters that are not in the launch template.
Note
You cannot remove launch template parameters during launch (for example, you cannot specify
a null value for the parameter). To remove a parameter, create a new version of the launch
template without the parameter and use that version to launch the instance.

To launch instances, IAM users must have permissions to use the ec2:RunInstances action. You must
also have permissions to create or use the resources that are created or associated with the instance. You
can use resource-level permissions for the ec2:RunInstances action to control the launch parameters
that users can specify. Alternatively, you can grant users permissions to launch an instance using a launch
template. This enables you to manage launch parameters in a launch template rather than in an IAM
policy, and to use a launch template as an authorization vehicle for launching instances. For example,
you can specify that users can only launch instances using a launch template, and that they can only use
a specific launch template. You can also control the launch parameters that users can override in the
launch template. For example policies, see Launch templates (p. 1009).

Control the use of launch templates

By default, IAM users do not have permissions to work with launch templates. You can create an IAM user
policy that grants users permissions to create, modify, describe, and delete launch templates and launch
template versions. You can also apply resource-level permissions to some launch template actions to
control a user's ability to use specific resources for those actions. For more information, see the following
example policies: Example: Work with launch templates (p. 1020).

Take care when granting users permissions to use the ec2:CreateLaunchTemplate and
ec2:CreateLaunchTemplateVersion actions. You cannot use resource-level permissions to control

533
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

which resources users can specify in the launch template. To restrict the resources that are used to
launch an instance, ensure that you grant permissions to create launch templates and launch template
versions only to appropriate administrators.

Create a launch template

Create a new launch template using parameters that you define, or use an existing launch template or an
instance as the basis for a new launch template.

Tasks
• Create a new launch template using parameters you define (p. 534)
• Create a launch template from an existing launch template (p. 538)
• Create a launch template from an instance (p. 538)

Create a new launch template using parameters you define

Console

To create a new launch template using defined parameters using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Launch Templates, and then choose Create launch template.
3. For Launch template name, enter a descriptive name for the launch template.
4. For Template version description, provide a brief description of the launch template version.
5. To tag the launch template on creation, expand Template tags, choose Add tag, and then enter
a tag key and value pair.
6. For Launch template contents, provide the following information:

• AMI: An AMI from which to launch the instance. To search through all available AMIs, choose
Search for AMI. To select a commonly used AMI, choose Quick Start. Or, choose AWS
Marketplace or Community AMIs. You can use an AMI that you own or find a suitable AMI.
• Instance type: Ensure that the instance type is compatible with the AMI that you've specified.
For more information, see Instance types (p. 203).
• Key pair name: The key pair for the instance. For more information, see Amazon EC2 key pairs
and Linux instances (p. 1044).
• Network platform: If applicable, whether to launch the instance into a VPC or EC2-Classic. If
you choose VPC, specify the subnet in the Network interfaces section. If you choose Classic,
ensure that the specified instance type is supported in EC2-Classic and specify the Availability
Zone for the instance.
• Security groups: One or more security groups to associate with the instance. If you add a
network interface to the launch template, omit this setting and specify the security groups
as part of the network interface specification. You cannot launch an instance from a launch
template that specifies security groups and a network interface. For more information, see
Amazon EC2 security groups for Linux instances (p. 1058).
7. For Storage (volumes), specify volumes to attach to the instance besides the volumes specified
by the AMI (Volume 1 (AMI Root)). To add a new volume, choose Add new volume.

• Volume type: The instance store or Amazon EBS volumes with which to associate your
instance. The type of volume depends on the instance type that you've chosen. For more
information, see Amazon EC2 instance store (p. 1268) and Amazon EBS volumes (p. 1082).
• Device name: A device name for the volume.
• Snapshot: The ID of the snapshot from which to create the volume.
• Size: For Amazon EBS volumes, the storage size.

534
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

• Volume type: For Amazon EBS volumes, the volume type. For more information, see Amazon
EBS volume types (p. 1084).
• IOPS: For the Provisioned IOPS SSD volume type, the number of I/O operations per second
(IOPS) that the volume can support.
• Delete on termination: For Amazon EBS volumes, whether to delete the volume when the
instance is terminated. For more information, see Preserve Amazon EBS volumes on instance
termination (p. 646).
• Encrypted: If the instance type supports EBS encryption, you can enable encryption for the
volume. If you have enabled encryption by default in this Region, encryption is enabled for
you. For more information, see Amazon EBS encryption (p. 1176).
• Key: The CMK to use for EBS encryption. You can specify the ARN of any customer master key
(CMK) that you created using the AWS Key Management Service. If you specify a CMK, you
must also use Encrypted to enable encryption.
8. For Resource tags, specify tags (p. 1311) by providing key and value combinations. You can tag
the instance, the volumes, Spot Instance requests, or all three.
9. For Network interfaces, you can specify up to two network interfaces (p. 839) for the
instance.

• Device index: The device number for the network interface, for example, eth0 for the
primary network interface. If you leave the field blank, AWS creates the primary network
interface.
• Network interface: The ID of the network interface, or leave blank to let AWS create a new
network interface.
• Description: (Optional) A description for the new network interface.
• Subnet: The subnet in which to create a new network interface. For the primary network
interface (eth0), this is the subnet in which the instance is launched. If you've entered an
existing network interface for eth0, the instance is launched in the subnet in which the
network interface is located.
• Auto-assign public IP: Whether to automatically assign a public IP address to the network
interface with the device index of eth0. This setting can only be enabled for a single, new
network interface.
• Primary IP: A private IPv4 address from the range of your subnet. Leave blank to let AWS
choose a private IPv4 address for you.
• Secondary IP: A secondary private IPv4 address from the range of your subnet. Leave blank to
let AWS choose one for you.
• (IPv6-only) IPv6 IPs: An IPv6 address from the range of the subnet.
• Security groups: One or more security groups in your VPC with which to associate the
network interface.
• Delete on termination: Whether the network interface is deleted when the instance is
deleted.
• Elastic Fabric Adapter: Indicates whether the network interface is an Elastic Fabric Adapter.
For more information, see Elastic Fabric Adapter.
• Network card index: The index of the network card. The primary network interface must be
assigned to network card index 0. Some instance types support multiple network cards.
10. For Advanced details, expand the section to view the fields and specify any additional
parameters for the instance.

• Purchasing option: The purchasing model. Choose Request Spot Instances to request Spot
Instances at the Spot price, capped at the On-Demand price, and choose Customize to change
the default Spot Instance settings. If you do not request a Spot Instance, EC2 launches an On-
Demand Instance by default. For more information, see Spot Instances (p. 364).

535
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

• IAM instance profile: An AWS Identity and Access Management (IAM) instance profile to
associate with the instance. For more information, see IAM roles for Amazon EC2 (p. 1032).
• Shutdown behavior: Whether the instance should stop or terminate when shut down. For
more information, see Change the instance initiated shutdown behavior (p. 646).
• Stop - Hibernate behavior: Whether the instance is enabled for hibernation. This field is
only valid for instances that meet the hibernation prerequisites. For more information, see
Hibernate your Linux instance (p. 627).
• Termination protection: Whether to prevent accidental termination. For more information,
see Enable termination protection (p. 645).
• Detailed CloudWatch monitoring: Whether to enable detailed monitoring of the instance
using Amazon CloudWatch. Additional charges apply. For more information, see Monitor your
instances using CloudWatch (p. 761).
• Elastic inference: An elastic inference accelerator to attach to your EC2 CPU instance. For
more information, see Working with Amazon Elastic Inference in the Amazon Elastic Inference
Developer Guide.
• T2/T3 Unlimited: Whether to enable applications to burst beyond the baseline for as long as
needed. This field is only valid for T2, T3, and T3a instances. Additional charges may apply.
For more information, see Burstable performance instances (p. 224).
• Placement group name: Specify a placement group in which to launch the instance. Not all
instance types can be launched in a placement group. For more information, see Placement
groups (p. 927).
• EBS-optimized instance: Provides additional, dedicated capacity for Amazon EBS I/O. Not all
instance types support this feature, and additional charges apply. For more information, see
Amazon EBS–optimized instances (p. 1217).
• Capacity Reservation: Specify whether to launch the instance into shared capacity, any open
Capacity Reservation, a specific Capacity Reservation, or a Capacity Reservation group. For
more information, see Launch instances into an existing Capacity Reservation (p. 509).
• Tenancy: Choose whether to run your instance on shared hardware (Shared), isolated,
dedicated hardware (Dedicated), or on a Dedicated Host (Dedicated host). If you choose
to launch the instance onto a Dedicated Host, you can specify whether to launch the
instance into a host resource group or you can target a specific Dedicated Host. Additional
charges may apply. For more information, see Dedicated Instances (p. 496) and Dedicated
Hosts (p. 465).
• RAM disk ID: (Only valid for paravirtual (PV) AMIs) A RAM disk for the instance. If you have
specified a kernel, you may need to specify a specific RAM disk with the drivers to support it.
• Kernel ID: (Only valid for paravirtual (PV) AMIs) A kernel for the instance.
• License configurations: You can launch instances against the specified license configuration
to track your license usage. For more information, see Create a License Configuration in the
AWS License Manager User Guide.
• Metadata accessible: Whether to enable or disable access to the instance metadata. For more
information, see Configure the instance metadata service (p. 705).
• Metadata version: If you enable access to the instance metadata, you can choose to require
the use of Instance Metadata Service Version 2 when requesting instance metadata. For more
information, see Configure instance metadata options for new instances (p. 709).
• Metadata response hop limit: If you enable instance metadata, you can set the allowable
number of network hops for the metadata token. For more information, see Configure the
instance metadata service (p. 705).
• User data: You can specify user data to configure an instance during launch, or to run a
configuration script. For more information, see Run commands on your Linux instance at
launch (p. 698).
11. Choose Create launch template.

536
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

AWS CLI

To create a launch template using the AWS CLI

• Use the create-launch-template command. The following example creates a launch template
that specifies the following:

• A tag for the launch template (purpose=production)

• The instance type (r4.4xlarge) and AMI (ami-8c1be5f6) to launch
• The number of cores (4) and threads per core (2) for a total of 8 vCPUs (4 cores x 2 threads)
• The subnet in which to launch the instance (subnet-7b16de0c)

The template assigns a public IP address and an IPv6 address to the instance and creates a tag
for the instance(Name=webserver).

aws ec2 create-launch-template \

--launch-template-name TemplateForWebServer \
--version-description WebVersion1 \
--tag-specifications 'ResourceType=launch-
template,Tags=[{Key=purpose,Value=production}]' \
--launch-template-data file://template-data.json

The following is an example template-data.json file.

{
"NetworkInterfaces": [{
"AssociatePublicIpAddress": true,
"DeviceIndex": 0,
"Ipv6AddressCount": 1,
"SubnetId": "subnet-7b16de0c"
}],
"ImageId": "ami-8c1be5f6",
"InstanceType": "r4.4xlarge",
"TagSpecifications": [{
"ResourceType": "instance",
"Tags": [{
"Key":"Name",
"Value":"webserver"
}]
}],
"CpuOptions": {
"CoreCount":4,
"ThreadsPerCore":2
}
}

The following is example output.

{
"LaunchTemplate": {
"LatestVersionNumber": 1,
"LaunchTemplateId": "lt-01238c059e3466abc",
"LaunchTemplateName": "TemplateForWebServer",
"DefaultVersionNumber": 1,
"CreatedBy": "arn:aws:iam::123456789012:root",
"CreateTime": "2017-11-27T09:13:24.000Z"
}
}

537
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Create a launch template from an existing launch template

To create a launch template from an existing launch template using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Launch Templates, and then choose Create launch template.
3. For Launch template name, enter a descriptive name for the launch template.
4. For Template version description, provide a brief description of the launch template version.
5. To tag the launch template on creation, expand Template tags, choose Add tag, and then enter a
tag key and value pair.
6. Expand Source template, and for Launch template name choose a launch template on which to
base the new launch template.
7. For Source template version, choose the launch template version on which to base the new launch
template.
8. Adjust any launch parameters as required, and then choose Create launch template.

Create a launch template from an instance

Console

To create a launch template from an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance, and choose Actions, Create template from instance.
4. Provide a name, description, and tags, and adjust the launch parameters as required.
Note
When you create a launch template from an instance, the instance's network interface
IDs and IP addresses are not included in the template.
5. Choose Create launch template.

AWS CLI

You can use the AWS CLI to create a launch template from an existing instance by first getting
the launch template data from an instance, and then creating a launch template using the launch
template data.

To get launch template data from an instance using the AWS CLI

• Use the get-launch-template-data command and specify the instance ID. You can use the output
as a base to create a new launch template or launch template version. By default, the output
includes a top-level LaunchTemplateData object, which cannot be specified in your launch
template data. Use the --query option to exclude this object.

aws ec2 get-launch-template-data \

--instance-id i-0123d646e8048babc \
--query "LaunchTemplateData"

The following is example output.

538
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

"Monitoring": {},
"ImageId": "ami-8c1be5f6",
"BlockDeviceMappings": [
{
"DeviceName": "/dev/xvda",
"Ebs": {
"DeleteOnTermination": true
}
}
],
"EbsOptimized": false,
"Placement": {
"Tenancy": "default",
"GroupName": "",
"AvailabilityZone": "us-east-1a"
},
"InstanceType": "t2.micro",
"NetworkInterfaces": [
{
"Description": "",
"NetworkInterfaceId": "eni-35306abc",
"PrivateIpAddresses": [
{
"Primary": true,
"PrivateIpAddress": "10.0.0.72"
}
],
"SubnetId": "subnet-7b16de0c",
"Groups": [
"sg-7c227019"
],
"Ipv6Addresses": [
{
"Ipv6Address": "2001:db8:1234:1a00::123"
}
],
"PrivateIpAddress": "10.0.0.72"
}
]
}

You can write the output directly to a file, for example:

aws ec2 get-launch-template-data \

--instance-id i-0123d646e8048babc \
--query "LaunchTemplateData" >> instance-data.json

To create a launch template using launch template data

Use the create-launch-template command to create a launch template using the output from the
previous procedure. For more information about creating a launch template using the AWS CLI, see
Create a new launch template using parameters you define (p. 534).

Manage launch template versions

You can create launch template versions for a specific launch template, set the default version, describe a
launch template version, and delete versions that you no longer require.

Tasks
• Create a launch template version (p. 540)

539
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

• Set the default launch template version (p. 540)

• Describe a launch template version (p. 541)
• Delete a launch template version (p. 542)

Create a launch template version

When you create a launch template version, you can specify new launch parameters or use an existing
version as the base for the new version. For more information about the launch parameters, see Create a
launch template (p. 534).

Console

To create a launch template version using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Launch Templates.
3. Select a launch template, and then choose Actions, Modify template (Create new version).
4. For Template version description, enter a description for the launch template version.
5. (Optional) Expand Source template and select a version of the launch template to use as a
base for the new launch template version. The new launch template version inherits the launch
parameters from this launch template version.
6. Modify the launch parameters as required, and choose Create launch template.

AWS CLI

To create a launch template version using the AWS CLI

• Use the create-launch-template-version command. You can specify a source version on which to
base the new version. The new version inherits the launch parameters from this version, and you
can override parameters using --launch-template-data. The following example creates a
new version based on version 1 of the launch template and specifies a different AMI ID.

aws ec2 create-launch-template-version \

--launch-template-id lt-0abcd290751193123 \
--version-description WebVersion2 \
--source-version 1 \
--launch-template-data "ImageId=ami-c998b6b2"

Set the default launch template version

You can set the default version for the launch template. When you launch an instance from a launch
template and do not specify a version, the instance is launched using the parameters of the default
version.

Console

To set the default launch template version using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Launch Templates.
3. Select the launch template and choose Actions, Set default version.
4. For Template version, select the version number to set as the default version and choose Set as
default version.

540
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

AWS CLI

To set the default launch template version using the AWS CLI

• Use the modify-launch-template command and specify the version that you want to set as the
default.

aws ec2 modify-launch-template \

--launch-template-id lt-0abcd290751193123 \
--default-version 2

Describe a launch template version

Using the console, you can view all the versions of the selected launch template, or get a list of the
launch templates whose latest or default version matches a specific version number. Using the AWS CLI,
you can describe all versions, individual versions, or a range of versions of a specified launch template.
You can also describe all the latest versions or all the default versions of all the launch templates in your
account.

Console

To describe a launch template version using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Launch Templates.
3. You can view a version of a specific launch template, or get a list of the launch templates whose
latest or default version matches a specific version number.
• To view a version of a launch template: Select the launch template. On the Versions tab,
from Version, select a version to view its details.
• To get a list of all the launch templates whose latest version matches a specific version
number: From the search bar, choose Latest version, and then choose a version number.
• To get a list of all the launch templates whose default version matches a specific version
number: From the search bar, choose Default version, and then choose a version number.

AWS CLI

To describe a launch template version using the AWS CLI

• Use the describe-launch-template-versions command and specify the version numbers. In the
following example, versions 1 and 3 are specified.

aws ec2 describe-launch-template-versions \

--launch-template-id lt-0abcd290751193123 \
--versions 1 3

To describe all the latest and default launch template versions in your account using the
AWS CLI

• Use the describe-launch-template-versions command and specify $Latest, $Default, or

both. You must omit the launch template ID and name in the call. You cannot specify version
numbers.

aws ec2 describe-launch-template-versions \

541
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

--versions "$Latest,$Default"

Delete a launch template version

If you no longer require a launch template version, you can delete it. You cannot replace the version
number after you delete it. You cannot delete the default version of the launch template; you must first
assign a different version as the default.

Console

To delete a launch template version using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Launch Templates.
3. Select the launch template and choose Actions, Delete template version.
4. Select the version to delete and choose Delete.

AWS CLI

To delete a launch template version using the AWS CLI

• Use the delete-launch-template-versions command and specify the version numbers to delete.

aws ec2 delete-launch-template-versions \

--launch-template-id lt-0abcd290751193123 \
--versions 1

Launch an instance from a launch template

You can use the parameters contained in a launch template to launch an instance. You have the option to
override or add launch parameters before you launch the instance.

Instances that are launched using a launch template are automatically assigned two tags with the keys
aws:ec2launchtemplate:id and aws:ec2launchtemplate:version. You cannot remove or edit
these tags.

Console

To launch an instance from a launch template using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Launch Templates.
3. Select the launch template and choose Actions, Launch instance from template.
4. For Source template version, select the launch template version to use.
5. For Number of instances, specify the number of instances to launch.
6. (Optional) You can override or add launch template parameters by changing and adding
parameters in the Instance details section.
7. Choose Launch instance from template.

542
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

AWS CLI

To launch an instance from a launch template using the AWS CLI

• Use the run-instances command and specify the --launch-template parameter. Optionally
specify the launch template version to use. If you don't specify the version, the default version is
used.

aws ec2 run-instances \

--launch-template LaunchTemplateId=lt-0abcd290751193123,Version=1

• To override a launch template parameter, specify the parameter in the run-instances command.
The following example overrides the instance type that's specified in the launch template (if any).

aws ec2 run-instances \

--launch-template LaunchTemplateId=lt-0abcd290751193123 \
--instance-type t2.small

• If you specify a nested parameter that's part of a complex structure, the instance is launched using
the complex structure as specified in the launch template plus any additional nested parameters
that you specify.

In the following example, the instance is launched with the tag Owner=TeamA as well as any other
tags that are specified in the launch template. If the launch template has an existing tag with a
key of Owner, the value is replaced with TeamA.

aws ec2 run-instances \

--launch-template LaunchTemplateId=lt-0abcd290751193123 \
--tag-specifications "ResourceType=instance,Tags=[{Key=Owner,Value=TeamA}]"

In the following example, the instance is launched with a volume with the device name /dev/
xvdb as well as any other block device mappings that are specified in the launch template. If the
launch template has an existing volume defined for /dev/xvdb, its values are replaced with the
specified values.

aws ec2 run-instances \

--launch-template LaunchTemplateId=lt-0abcd290751193123 \
--block-device-mappings "DeviceName=/dev/xvdb,Ebs={VolumeSize=20,VolumeType=gp2}"

If the instance fails to launch or the state immediately goes to terminated instead of running, see
Troubleshoot instance launch issues (p. 1326).

Use launch templates with Amazon EC2 Auto Scaling

You can create an Auto Scaling group and specify a launch template to use for the group. When Amazon
EC2 Auto Scaling launches instances in the Auto Scaling group, it uses the launch parameters defined
in the associated launch template. For more information, see Creating an Auto Scaling Group Using a
Launch Template in the Amazon EC2 Auto Scaling User Guide.

Before you can create an Auto Scaling group using a launch template, you must create a launch template
that includes the parameters required to launch an instance in an Auto Scaling group, such as the ID
of the AMI. The console provides guidance to help you create a template that you can use with Auto
Scaling.

543
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

To create a launch template to use with Auto Scaling using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Launch Templates, and then choose Create launch template.
3. For Launch template name, enter a descriptive name for the launch template.
4. For Template version description, provide a brief description of the launch template version.
5. Under Auto Scaling guidance, select the checkbox to have Amazon EC2 provide guidance to help
create a template to use with Auto Scaling.
6. Modify the launch parameters as required. Because you selected Auto Scaling guidance, some fields
are required and some fields are not available. For considerations to keep in mind when creating
a launch template, and for information about how to configure the launch parameters for Auto
Scaling, see Creating a Launch Template for an Auto Scaling Group in the Amazon EC2 Auto Scaling
User Guide.
7. Choose Create launch template.
8. (Optional) To create an Auto Scaling group using this launch template, in the Next steps page,
choose Create Auto Scaling group.

To create or update an Amazon EC2 Auto Scaling group with a launch template using the
AWS CLI

• Use the create-auto-scaling-group or the update-auto-scaling-group command and specify the --

launch-template parameter.

Use launch templates with EC2 Fleet

You can create an EC2 Fleet request and specify a launch template in the instance configuration. When
Amazon EC2 fulfills the EC2 Fleet request, it uses the launch parameters defined in the associated launch
template. You can override some of the parameters that are specified in the launch template.

For more information, see Create an EC2 Fleet (p. 570).

To create an EC2 Fleet with a launch template using the AWS CLI

• Use the create-fleet command. Use the --launch-template-configs parameter to specify the
launch template and any overrides for the launch template.

Use launch templates with Spot Fleet

You can create a Spot Fleet request and specify a launch template in the instance configuration. When
Amazon EC2 fulfills the Spot Fleet request, it uses the launch parameters defined in the associated
launch template. You can override some of the parameters that are specified in the launch template.

For more information, see Spot Fleet requests (p. 400).

To create a Spot Fleet request with a launch template using the AWS CLI

• Use the request-spot-fleet command. Use the LaunchTemplateConfigs parameter to specify the
launch template and any overrides for the launch template.

Delete a launch template

If you no longer require a launch template, you can delete it. Deleting a launch template deletes all of its
versions.

544
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Console

To delete a launch template (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Launch Templates.
3. Select the launch template and choose Actions, Delete template.
4. Enter Delete to confirm deletion, and then choose Delete.

AWS CLI

To delete a launch template (AWS CLI)

• Use the delete-launch-template (AWS CLI) command and specify the launch template.

aws ec2 delete-launch-template --launch-template-id lt-01238c059e3466abc

Launch an instance using parameters from an existing instance

The Amazon EC2 console provides a Launch more like this wizard option that enables you to use
a current instance as a base for launching other instances. This option automatically populates the
Amazon EC2 launch wizard with certain configuration details from the selected instance.
Note
The Launch more like this wizard option does not clone your selected instance; it only
replicates some configuration details. To create a copy of your instance, first create an AMI from
it, then launch more instances from the AMI.
Alternatively, create a launch template (p. 532) to store the launch parameters for your
instances.

The following configuration details are copied from the selected instance into the launch wizard:

• AMI ID
• Instance type
• Availability Zone, or the VPC and subnet in which the selected instance is located
• Public IPv4 address. If the selected instance currently has a public IPv4 address, the new instance
receives a public IPv4 address - regardless of the selected instance's default public IPv4 address
setting. For more information about public IPv4 addresses, see Public IPv4 addresses and external DNS
hostnames (p. 810).
• Placement group, if applicable
• IAM role associated with the instance, if applicable
• Shutdown behavior setting (stop or terminate)
• Termination protection setting (true or false)
• CloudWatch monitoring (enabled or disabled)
• Amazon EBS-optimization setting (true or false)
• Tenancy setting, if launching into a VPC (shared or dedicated)
• Kernel ID and RAM disk ID, if applicable
• User data, if specified
• Tags associated with the instance, if applicable
• Security groups associated with the instance

545
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

The following configuration details are not copied from your selected instance. Instead, the wizard
applies their default settings or behavior:

• Number of network interfaces: The default is one network interface, which is the primary network
interface (eth0).
• Storage: The default storage configuration is determined by the AMI and the instance type.

New console

To use your current instance as a template

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance you want to use, and then choose Actions, Images and templates, Launch
more like this.
4. The launch wizard opens on the Review Instance Launch page. You can make any necessary
changes by choosing the appropriate Edit link.

When you are ready, choose Launch to select a key pair and launch your instance.
5. If the instance fails to launch or the state immediately goes to terminated instead of
running, see Troubleshoot instance launch issues (p. 1326).

Old console

To use your current instance as a template

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance you want to use, and then choose Actions, Launch More Like This.
4. The launch wizard opens on the Review Instance Launch page. You can make any necessary
changes by choosing the appropriate Edit link.

When you are ready, choose Launch to select a key pair and launch your instance.
5. If the instance fails to launch or the state immediately goes to terminated instead of
running, see Troubleshoot instance launch issues (p. 1326).

Launch an AWS Marketplace instance

You can subscribe to an AWS Marketplace product and launch an instance from the product's AMI
using the Amazon EC2 launch wizard. For more information about paid AMIs, see Paid AMIs (p. 121). To
cancel your subscription after launch, you first have to terminate all instances running from it. For more
information, see Manage your AWS Marketplace subscriptions (p. 125).

To launch an instance from the AWS Marketplace using the launch wizard

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the Amazon EC2 dashboard, choose Launch instance.
3. On the Choose an Amazon Machine Image (AMI) page, choose the AWS Marketplace category on
the left. Find a suitable AMI by browsing the categories, or using the search functionality. Choose
Select to choose your product.
4. A dialog displays an overview of the product you've selected. You can view the pricing information,
as well as any other information that the vendor has provided. When you're ready, choose Continue.

546
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Note
You are not charged for using the product until you have launched an instance with the
AMI. Take note of the pricing for each supported instance type, as you will be prompted to
select an instance type on the next page of the wizard. Additional taxes may also apply to
the product.
5. On the Choose an Instance Type page, select the hardware configuration and size of the instance to
launch. When you're done, choose Next: Configure Instance Details.
6. On the next pages of the wizard, you can configure your instance, add storage, and add tags. For
more information about the different options you can configure, see Launch an instance using the
Launch Instance Wizard (p. 526). Choose Next until you reach the Configure Security Group page.

The wizard creates a new security group according to the vendor's specifications for the product. The
security group may include rules that allow all IPv4 addresses (0.0.0.0/0) access on SSH (port 22)
on Linux or RDP (port 3389) on Windows. We recommend that you adjust these rules to allow only a
specific address or range of addresses to access your instance over those ports.

When you are ready, choose Review and Launch.

7. On the Review Instance Launch page, check the details of the AMI from which you're about to
launch the instance, as well as the other configuration details you set up in the wizard. When you're
ready, choose Launch to select or create a key pair, and launch your instance.
8. Depending on the product you've subscribed to, the instance may take a few minutes or more to
launch. You are first subscribed to the product before your instance can launch. If there are any
problems with your credit card details, you will be asked to update your account details. When the
launch confirmation page displays, choose View Instances to go to the Instances page.
Note
You are charged the subscription price as long as your instance is running, even if it is idle. If
your instance is stopped, you may still be charged for storage.
9. When your instance is in the running state, you can connect to it. To do this, select your instance
in the list and choose Connect. Follow the instructions in the dialog. For more information about
connecting to your instance, see Connect to your Linux instance (p. 598).
Important
Check the vendor's usage instructions carefully, as you may need to use a specific user name
to log in to the instance. For more information about accessing your subscription details,
see Manage your AWS Marketplace subscriptions (p. 125).
10. If the instance fails to launch or the state immediately goes to terminated instead of running, see
Troubleshoot instance launch issues (p. 1326).

Launch an AWS Marketplace AMI instance using the API and CLI
To launch instances from AWS Marketplace products using the API or command line tools, first ensure
that you are subscribed to the product. You can then launch an instance with the product's AMI ID using
the following methods:

Method Documentation

AWS CLI Use the run-instances command, or see the following topic for more
information: Launching an Instance.

AWS Tools for Windows Use the New-EC2Instance command, or see the following topic for
PowerShell more information: Launch an Amazon EC2 Instance Using Windows
PowerShell

Query API Use the RunInstances request.

547
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Launch instances using an EC2 Fleet

An EC2 Fleet contains the configuration information to launch a fleet—or group—of instances. In a single
API call, a fleet can launch multiple instance types across multiple Availability Zones, using the On-
Demand Instance, Reserved Instance, and Spot Instance purchasing options together. Using EC2 Fleet,
you can:

• Define separate On-Demand and Spot capacity targets and the maximum amount you’re willing to pay
per hour
• Specify the instance types that work best for your applications
• Specify how Amazon EC2 should distribute your fleet capacity within each purchasing option

You can also set a maximum amount per hour that you’re willing to pay for your fleet, and EC2 Fleet
launches instances until it reaches the maximum amount. When the maximum amount you're willing to
pay is reached, the fleet stops launching instances even if it hasn’t met the target capacity.

The EC2 Fleet attempts to launch the number of instances that are required to meet the target capacity
specified in your request. If you specified a total maximum price per hour, it fulfills the capacity until it
reaches the maximum amount that you’re willing to pay. The fleet can also attempt to maintain its target
Spot capacity if your Spot Instances are interrupted. For more information, see How Spot Instances
work (p. 369).

You can specify an unlimited number of instance types per EC2 Fleet. Those instance types can be
provisioned using both On-Demand and Spot purchasing options. You can also specify multiple
Availability Zones, specify different maximum Spot prices for each instance, and choose additional
Spot options for each fleet. Amazon EC2 uses the specified options to provision capacity when the fleet
launches.

While the fleet is running, if Amazon EC2 reclaims a Spot Instance because of a price increase or instance
failure, EC2 Fleet can try to replace the instances with any of the instance types that you specify. This
makes it easier to regain capacity during a spike in Spot pricing. You can develop a flexible and elastic
resourcing strategy for each fleet. For example, within specific fleets, your primary capacity can be On-
Demand supplemented with less-expensive Spot capacity if available.

If you have Reserved Instances and you specify On-Demand Instances in your fleet, EC2 Fleet uses your
Reserved Instances. For example, if your fleet specifies an On-Demand Instance as c4.large, and you
have Reserved Instances for c4.large, you receive the Reserved Instance pricing.

There is no additional charge for using EC2 Fleet. You pay only for the EC2 instances that the fleet
launches for you.

Contents
• EC2 Fleet limitations (p. 549)
• EC2 Fleet limits (p. 549)

548
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

• Burstable performance instances (p. 549)

• EC2 Fleet configuration strategies (p. 550)
• Work with EC2 Fleets (p. 561)
• Use Amazon EventBridge to monitor EC2 Fleet events (p. 591)

EC2 Fleet limitations

The following limitations apply to EC2 Fleet:

• EC2 Fleet is available only through the API or AWS CLI.

• An EC2 Fleet request can't span AWS Regions. You need to create a separate EC2 Fleet for each Region.
• An EC2 Fleet request can't span different subnets from the same Availability Zone.

EC2 Fleet limits

The usual Amazon EC2 limits apply to instances launched by an EC2 Fleet, such as Spot request price
limits, instance limits, and volume limits. In addition, the following limits apply:

• The number of active EC2 Fleets per AWS Region: 1,000 * †

• The number of Spot Instance pools (unique combination of instance type and subnet): 300* ‡
• The size of the user data in a launch specification: 16 KB †
• The target capacity per EC2 Fleet: 10,000
• The target capacity across all EC2 Fleets in a Region: 100,000 *
• An EC2 Fleet request can't span Regions.
• An EC2 Fleet request can't span different subnets from the same Availability Zone.

If you need more than the default limits for target capacity, complete the AWS Support Center Create
case form to request a limit increase. For Limit type, choose EC2 Fleet, choose a Region, and then choose
Target Fleet Capacity per Fleet (in units) or Target Fleet Capacity per Region (in units), or both.

* These limits apply to both your EC2 Fleets and your Spot Fleets.

† These are hard limits. You cannot request a limit increase for these limits.

‡ This limit only applies to fleets of type request or maintain. This limit does not apply to instant
fleets.

Burstable performance instances

If you launch your Spot Instances using a burstable performance instance type (p. 224), and if you plan
to use your burstable performance Spot Instances immediately and for a short duration, with no idle
time for accruing CPU credits, we recommend that you launch them in Standard mode (p. 236) to avoid
paying higher costs. If you launch burstable performance Spot Instances in Unlimited mode (p. 228)
and burst CPU immediately, you'll spend surplus credits for bursting. If you use the instance for a short
duration, the instance doesn't have time to accrue CPU credits to pay down the surplus credits, and you
are charged for the surplus credits when you terminate the instance.

Unlimited mode is suitable for burstable performance Spot Instances only if the instance runs long
enough to accrue CPU credits for bursting. Otherwise, paying for surplus credits makes burstable
performance Spot Instances more expensive than using other instances. For more information, see When
to use unlimited mode versus fixed CPU (p. 229).

549
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Launch credits are meant to provide a productive initial launch experience for T2 instances by providing
sufficient compute resources to configure the instance. Repeated launches of T2 instances to access new
launch credits is not permitted. If you require sustained CPU, you can earn credits (by idling over some
period), use Unlimited mode (p. 228) for T2 Spot Instances, or use an instance type with dedicated CPU.

EC2 Fleet configuration strategies

An EC2 Fleet is a group of On-Demand Instances and Spot Instances.

The EC2 Fleet attempts to launch the number of instances that are required to meet the target capacity
that you specify in the fleet request. The fleet can comprise only On-Demand Instances, only Spot
Instances, or a combination of both On-Demand Instances and Spot Instances. The request for Spot
Instances is fulfilled if there is available capacity and the maximum price per hour for your request
exceeds the Spot price. The fleet also attempts to maintain its target capacity if your Spot Instances are
interrupted.

You can also set a maximum amount per hour that you’re willing to pay for your fleet, and EC2 Fleet
launches instances until it reaches the maximum amount. When the maximum amount you're willing to
pay is reached, the fleet stops launching instances even if it hasn’t met the target capacity.

A Spot Instance pool is a set of unused EC2 instances with the same instance type, operating system,
Availability Zone, and network platform. When you create an EC2 Fleet, you can include multiple launch
specifications, which vary by instance type, Availability Zone, subnet, and maximum price. The fleet
selects the Spot Instance pools that are used to fulfill the request, based on the launch specifications
included in your request, and the configuration of the request. The Spot Instances come from the
selected pools.

An EC2 Fleet enables you to provision large amounts of EC2 capacity that makes sense for your
application based on number of cores or instances, or amount of memory. For example, you can specify
an EC2 Fleet to launch a target capacity of 200 instances, of which 130 are On-Demand Instances and
the rest are Spot Instances.

Use the appropriate configuration strategies to create an EC2 Fleet that meets your needs.

Contents
• Planning an EC2 Fleet (p. 550)
• EC2 Fleet request types (p. 551)
• Allocation strategies for Spot Instances (p. 551)
• Configure EC2 Fleet for On-Demand backup (p. 553)
• Capacity Rebalancing (p. 554)
• Maximum price overrides (p. 555)
• Control spending (p. 555)
• EC2 Fleet instance weighting (p. 556)
• Tutorial: Use EC2 Fleet with instance weighting (p. 557)
• Tutorial: Use EC2 Fleet with On-Demand as the primary capacity (p. 560)

Planning an EC2 Fleet

When planning your EC2 Fleet, we recommend that you do the following:

• Determine whether you want to create an EC2 Fleet that submits a synchronous or asynchronous one-
time request for the desired target capacity, or one that maintains a target capacity over time. For
more information, see EC2 Fleet request types (p. 551).

550
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

• Determine the instance types that meet your application requirements.

• If you plan to include Spot Instances in your EC2 Fleet, review Spot Best Practices before you create
the fleet. Use these best practices when you plan your fleet so that you can provision the instances at
the lowest possible price.
• Determine the target capacity for your EC2 Fleet. You can set target capacity in instances or in custom
units. For more information, see EC2 Fleet instance weighting (p. 556).
• Determine what portion of the EC2 Fleet target capacity must be On-Demand capacity and Spot
capacity. You can specify 0 for On-Demand capacity or Spot capacity, or both.
• Determine your price per unit, if you are using instance weighting. To calculate the price per unit,
divide the price per instance hour by the number of units (or weight) that this instance represents. If
you are not using instance weighting, the default price per unit is the price per instance hour.
• Determine the maximum amount per hour that you’re willing to pay for your fleet. For more
information, see Control spending (p. 555).
• Review the possible options for your EC2 Fleet. For more information, see the EC2 Fleet JSON
configuration file reference (p. 567). For EC2 Fleet configuration examples, see EC2 Fleet example
configurations (p. 579).

EC2 Fleet request types

There are three types of EC2 Fleet requests:

instant

If you configure the request type as instant, EC2 Fleet places a synchronous one-time request for
your desired capacity. In the API response, it returns the instances that launched, along with errors
for those instances that could not be launched.
request

If you configure the request type as request, EC2 Fleet places an asynchronous one-time request
for your desired capacity. Thereafter, if capacity is diminished because of Spot interruptions, the
fleet does not attempt to replenish Spot Instances, nor does it submit requests in alternative Spot
Instance pools if capacity is unavailable.
maintain

(Default) If you configure the request type as maintain, EC2 Fleet places an asynchronous request
for your desired capacity, and maintains capacity by automatically replenishing any interrupted Spot
Instances.

All three types of requests benefit from an allocation strategy. For more information, see Allocation
strategies for Spot Instances (p. 551).

Allocation strategies for Spot Instances

The allocation strategy for your EC2 Fleet determines how it fulfills your request for Spot Instances from
the possible Spot Instance pools represented by its launch specifications. The following are the allocation
strategies that you can specify in your fleet:

lowest-price

The Spot Instances come from the pool with the lowest price. This is the default strategy.
diversified

The Spot Instances are distributed across all pools.

551
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

capacity-optimized

The Spot Instances come from the pool with optimal capacity for the number of instances that are
launching.
InstancePoolsToUseCount

The Spot Instances are distributed across the number of Spot pools that you specify. This parameter
is valid only when used in combination with lowest-price.

Maintaining target capacity

After Spot Instances are terminated due to a change in the Spot price or available capacity of a
Spot Instance pool, an EC2 Fleet of type maintain launches replacement Spot Instances. If the
allocation strategy is lowest-price, the fleet launches replacement instances in the pool where the
Spot price is currently the lowest. If the allocation strategy is lowest-price in combination with
InstancePoolsToUseCount, the fleet selects the Spot pools with the lowest price and launches Spot
Instances across the number of Spot pools that you specify. If the allocation strategy is capacity-
optimized, the fleet launches replacement instances in the pool that has the most available Spot
Instance capacity. If the allocation strategy is diversified, the fleet distributes the replacement Spot
Instances across the remaining pools.

Configure EC2 Fleet for cost optimization

To optimize the costs for your use of Spot Instances, specify the lowest-price allocation strategy so
that EC2 Fleet automatically deploys the least expensive combination of instance types and Availability
Zones based on the current Spot price.

For On-Demand Instance target capacity, EC2 Fleet always selects the cheapest instance type based on
the public On-Demand price, while continuing to follow the allocation strategy (either lowest-price,
capacity-optimized, or diversified) for Spot Instances.

Configure EC2 Fleet for cost optimization and diversification

To create a fleet of Spot Instances that is both cheap and diversified, use the lowest-price allocation
strategy in combination with InstancePoolsToUseCount. EC2 Fleet automatically deploys the least
expensive combination of instance types and Availability Zones based on the current Spot price across
the number of Spot pools that you specify. This combination can be used to avoid the most expensive
Spot Instances.

Configure EC2 Fleet for capacity optimization

With Spot Instances, pricing changes slowly over time based on long-term trends in supply and demand,
but capacity fluctuates in real time. The capacity-optimized strategy automatically launches
Spot Instances into the most available pools by looking at real-time capacity data and predicting
which are the most available. This works well for workloads such as big data and analytics, image and
media rendering, machine learning, and high performance computing that may have a higher cost of
interruption associated with restarting work and checkpointing. By offering the possibility of fewer
interruptions, the capacity-optimized strategy can lower the overall cost of your workload.

Choose the appropriate allocation strategy

You can optimize your fleet based on your use case.

If your fleet is small or runs for a short time, the probability that your Spot Instances will be interrupted
is low, even with all of the instances in a single Spot Instance pool. Therefore, the lowest-price
strategy is likely to meet your needs while providing the lowest cost.

If your fleet is large or runs for a long time, you can improve the availability of your fleet by distributing
the Spot Instances across multiple pools. For example, if your EC2 Fleet specifies 10 pools and a target

552
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

capacity of 100 instances, the fleet launches 10 Spot Instances in each pool. If the Spot price for one
pool exceeds your maximum price for this pool, only 10% of your fleet is affected. Using this strategy
also makes your fleet less sensitive to increases in the Spot price in any one pool over time.

With the diversified strategy, the EC2 Fleet does not launch Spot Instances into any pools with a
Spot price that is equal to or higher than the On-Demand price.

To create a cheap and diversified fleet, use the lowest-price strategy in combination with
InstancePoolsToUseCount. You can use a low or high number of Spot pools across which to allocate
your Spot Instances. For example, if you run batch processing, we recommend specifying a low number
of Spot pools (for example, InstancePoolsToUseCount=2) to ensure that your queue always has
compute capacity while maximizing savings. If you run a web service, we recommend specifying a high
number of Spot pools (for example, InstancePoolsToUseCount=10) to minimize the impact if a Spot
Instance pool becomes temporarily unavailable.

If your fleet runs workloads that may have a higher cost of interruption associated with restarting work
and checkpointing, then use the capacity-optimized strategy. This strategy offers the possibility of
fewer interruptions, which can lower the overall cost of your workload.

Configure EC2 Fleet for On-Demand backup

If you have urgent, unpredictable scaling needs, such as a news website that must scale during a major
news event or game launch, we recommend that you specify alternative instance types for your On-
Demand Instances, in the event that your preferred option does not have sufficient available capacity.
For example, you might prefer c5.2xlarge On-Demand Instances, but if there is insufficient available
capacity, you'd be willing to use some c4.2xlarge instances during peak load. In this case, EC2 Fleet
attempts to fulfill all of your target capacity using c5.2xlarge instances, but if there is insufficient
capacity, it automatically launches c4.2xlarge instances to fulfill the target capacity.

Prioritize instance types for On-Demand capacity

When EC2 Fleet attempts to fulfill your On-Demand capacity, it defaults to launching the lowest-
priced instance type first. If AllocationStrategy is set to prioritized, EC2 Fleet uses priority to
determine which instance type to use first in fulfilling On-Demand capacity. The priority is assigned to
the launch template override, and the highest priority is launched first.

For example, you have configured three launch template overrides, each with a different instance type:
c3.large, c4.large, and c5.large. The On-Demand price for c5.large is less than the price for
c4.large. c3.large is the cheapest. If you do not use priority to determine the order, the fleet fulfills
On-Demand capacity by starting with c3.large, and then c5.large. Because you often have unused
Reserved Instances for c4.large, you can set the launch template override priority so that the order is
c4.large, c3.large, and then c5.large.

Use Capacity Reservations for On-Demand Instances

You can configure a fleet to use On-Demand Capacity Reservations first when launching On-Demand
Instances by setting the usage strategy for Capacity Reservations to use-capacity-reservations-
first. You can use this setting in conjunction with the allocation strategy for On-Demand Instances
(lowest-price or prioritized).

When unused Capacity Reservations are used to fulfil On-Demand capacity:

• The fleet uses unused Capacity Reservations to fulfill On-Demand capacity up to the target On-
Demand capacity.
• If multiple instance pools have unused Capacity Reservations, the On-Demand allocation strategy
(lowest-price or prioritized) is applied.
• If the number of unused Capacity Reservations is less than the On-Demand target capacity, the
remaining On-Demand target capacity is launched according to the On-Demand allocation strategy
(lowest-price or prioritized).

553
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

You can only use unused On-Demand Capacity Reservations for fleets of type instant.

For examples of how to configure a fleet to use Capacity Reservations to fulfil On-Demand capacity,
see EC2 Fleet example configurations (p. 579). For more information, see On-Demand Capacity
Reservations (p. 501) and the On-Demand Capacity Reservation FAQs.

Capacity Rebalancing

You can configure EC2 Fleet to launch a replacement Spot Instance when Amazon EC2 emits a rebalance
recommendation to notify you that a Spot Instance is at an elevated risk of interruption. Capacity
Rebalancing helps you maintain workload availability by proactively augmenting your fleet with a new
Spot Instance before a running instance is interrupted by Amazon EC2. For more information, see EC2
instance rebalance recommendations (p. 449).

To configure EC2 Fleet to launch a replacement Spot Instance, use the create-fleet (AWS CLI) command
and the relevant parameters in the MaintenanceStrategies structure. For more information, see the
example launch configuration (p. 590).

Limitations

• Only available for fleets of type maintain.

• When the fleet is running, you can't modify the Capacity Rebalancing setting. To change the Capacity
Rebalancing setting, you must delete the fleet and create a new fleet.

Considerations

If you configure an EC2 Fleet for Capacity Rebalancing, consider the following:

EC2 Fleet can launch new replacement Spot Instances until fulfilled capacity is double target
capacity

When an EC2 Fleet is configured for Capacity Rebalancing, the fleet attempts to launch a new
replacement Spot Instance for every Spot Instance that receives a rebalance recommendation.
After a Spot Instance receives a rebalance recommendation, it is no longer counted as part of the
fulfilled capacity, and EC2 Fleet does not automatically terminate the instance. This gives you the
opportunity to perform rebalancing actions (p. 450) on the instance. Thereafter, you can terminate
the instance, or you can leave it running.

If your fleet reaches double its target capacity, it stops launching new replacement instances even if
the replacement instances themselves receive a rebalance recommendation.

For example, you create an EC2 Fleet with a target capacity of 100 Spot Instances. All the Spot
Instances receive a rebalance recommendation, which causes EC2 Fleet to launch 100 replacement
Spot Instances. This raises the number of fulfilled Spot Instances to 200, which is double the target
capacity. Some of the replacement instances receive a rebalance recommendation, but no more
replacement instances are launched because the fleet cannot exceed double its target capacity.

Note that you are charged for all of the instances while they are running.
We recommend that you manually terminate Spot Instances that receive a rebalance
recommendation

If you configure your EC2 Fleet for Capacity Rebalancing, we recommend that you monitor the
rebalance recommendation signal that is received by the Spot Instances in the fleet. By monitoring
the signal, you can quickly perform rebalancing actions (p. 450) on the affected instances
before Amazon EC2 interrupts them, and then you can manually terminate them. If you do not
terminate the instances, you continue paying for them while they are running. EC2 Fleet does not
automatically terminate the instances that receive a rebalance recommendation.

554
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

You can set up notifications using Amazon EventBridge or instance metadata. For more information,
see Monitor rebalance recommendation signals (p. 450).
EC2 Fleet does not count instances that receive a rebalance recommendation when calculating
fulfilled capacity during scale in or out

If your EC2 Fleet is configured for Capacity Rebalancing, and you change the target capacity to either
scale in or scale out, the fleet does not count the instances that are marked for rebalance as part of
the fulfilled capacity, as follows:
• Scale in – If you decrease your desired target capacity, the fleet terminates instances that are
not marked for rebalance until the desired capacity is reached. The instances that are marked for
rebalance are not counted towards the fulfilled capacity.

For example, you create an EC2 Fleet with a target capacity of 100 Spot Instances. 10 instances
receive a rebalance recommendation, so the fleet launches 10 new replacement instances,
resulting in a fulfilled capacity of 110 instances. You then reduce the target capacity to 50 (scale
in), but the fulfilled capacity is actually 60 instances because the 10 instances that are marked for
rebalance are not terminated by the fleet. You need to manually terminate these instances, or you
can leave them running.
• Scale out – If you increase your desired target capacity, the fleet launches new instances until the
desired capacity is reached. The instances that are marked for rebalance are not counted towards
the fulfilled capacity.

For example, you create an EC2 Fleet with a target capacity of 100 Spot Instances. 10 instances
receive a rebalance recommendation, so the fleet launches 10 new replacement instances,
resulting in a fulfilled capacity of 110 instances. You then increase the target capacity to 200
(scale out), but the fulfilled capacity is actually 210 instances because the 10 instances that are
marked for rebalance are not counted by the fleet as part of the target capacity. You need to
manually terminate these instances, or you can leave them running.
Provide as many Spot Instance pools in the request as possible

Configure your EC2 Fleet to use multiple instance types and Availability Zones. This provides the
flexibility to launch Spot Instances in various Spot Instance pools. For more information, see Be
flexible about instance types and Availability Zones (p. 368).
Configure your EC2 Fleet to use the most optimal Spot Instance pools

Use the capacity-optimized allocation strategy to ensure that replacement Spot Instances
are launched in the most optimal Spot Instance pools. For more information, see Use the capacity
optimized allocation strategy (p. 369).

Maximum price overrides

Each EC2 Fleet can either include a global maximum price, or use the default (the On-Demand price). The
fleet uses this as the default maximum price for each of its launch specifications.

You can optionally specify a maximum price in one or more launch specifications. This price is specific
to the launch specification. If a launch specification includes a specific price, the EC2 Fleet uses this
maximum price, overriding the global maximum price. Any other launch specifications that do not
include a specific maximum price still use the global maximum price.

Control spending

EC2 Fleet stops launching instances when it has met one of the following parameters: the
TotalTargetCapacity or the MaxTotalPrice (the maximum amount you’re willing to pay). To
control the amount you pay per hour for your fleet, you can specify the MaxTotalPrice. When the
maximum total price is reached, EC2 Fleet stops launching instances even if it hasn’t met the target
capacity.

555
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

The following examples show two different scenarios. In the first, EC2 Fleet stops launching instances
when it has met the target capacity. In the second, EC2 Fleet stops launching instances when it has
reached the maximum amount you’re willing to pay (MaxTotalPrice).

Example: Stop launching instances when target capacity is reached

Given a request for m4.large On-Demand Instances, where:

• On-Demand Price: $0.10 per hour

• OnDemandTargetCapacity: 10
• MaxTotalPrice: $1.50

EC2 Fleet launches 10 On-Demand Instances because the total of $1.00 (10 instances x $0.10) does not
exceed the MaxTotalPrice of $1.50 for On-Demand Instances.

Example: Stop launching instances when maximum total price is reached

Given a request for m4.large On-Demand Instances, where:

• On-Demand Price: $0.10 per hour

• OnDemandTargetCapacity: 10
• MaxTotalPrice: $0.80

If EC2 Fleet launches the On-Demand target capacity (10 On-Demand Instances), the total cost per
hour would be $1.00. This is more than the amount ($0.80) specified for MaxTotalPrice for On-
Demand Instances. To prevent spending more than you're willing to pay, EC2 Fleet launches only 8 On-
Demand Instances (below the On-Demand target capacity) because launching more would exceed the
MaxTotalPrice for On-Demand Instances.

EC2 Fleet instance weighting

When you create an EC2 Fleet, you can define the capacity units that each instance type would
contribute to your application's performance. You can then adjust your maximum price for each launch
specification by using instance weighting.

By default, the price that you specify is per instance hour. When you use the instance weighting feature,
the price that you specify is per unit hour. You can calculate your price per unit hour by dividing your
price for an instance type by the number of units that it represents. EC2 Fleet calculates the number of
instances to launch by dividing the target capacity by the instance weight. If the result isn't an integer,
the fleet rounds it up to the next integer, so that the size of your fleet is not below its target capacity.
The fleet can select any pool that you specify in your launch specification, even if the capacity of the
instances launched exceeds the requested target capacity.

The following table includes examples of calculations to determine the price per unit for an EC2 Fleet
with a target capacity of 10.

Instance Instance Target Number of Price per Price per unit hour
type weight capacity instances instance
launched hour

r3.xlarge 2 10 5 $0.05 $0.025

(10 divided (.05 divided by 2)

by 2)

556
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Instance Instance Target Number of Price per Price per unit hour
type weight capacity instances instance
launched hour

r3.8xlarge 8 10 2 $0.10 $0.0125

(10 divided (.10 divided by 8)

by 8, result
rounded up)

Use EC2 Fleet instance weighting as follows to provision the target capacity that you want in the pools
with the lowest price per unit at the time of fulfillment:

1. Set the target capacity for your EC2 Fleet either in instances (the default) or in the units of your
choice, such as virtual CPUs, memory, storage, or throughput.
2. Set the price per unit.
3. For each launch specification, specify the weight, which is the number of units that the instance type
represents toward the target capacity.

Instance weighting example

Consider an EC2 Fleet request with the following configuration:

• A target capacity of 24
• A launch specification with an instance type r3.2xlarge and a weight of 6
• A launch specification with an instance type c3.xlarge and a weight of 5

The weights represent the number of units that instance type represents toward the target capacity. If
the first launch specification provides the lowest price per unit (price for r3.2xlarge per instance hour
divided by 6), the EC2 Fleet would launch four of these instances (24 divided by 6).

If the second launch specification provides the lowest price per unit (price for c3.xlarge per instance
hour divided by 5), the EC2 Fleet would launch five of these instances (24 divided by 5, result rounded
up).

Instance weighting and allocation strategy

Consider an EC2 Fleet request with the following configuration:

• A target capacity of 30 Spot Instances

• A launch specification with an instance type c3.2xlarge and a weight of 8
• A launch specification with an instance type m3.xlarge and a weight of 8
• A launch specification with an instance type r3.xlarge and a weight of 8

The EC2 Fleet would launch four instances (30 divided by 8, result rounded up). With the lowest-
price strategy, all four instances come from the pool that provides the lowest price per unit. With
the diversified strategy, the fleet launches one instance in each of the three pools, and the fourth
instance in whichever of the three pools provides the lowest price per unit.

Tutorial: Use EC2 Fleet with instance weighting

This tutorial uses a fictitious company called Example Corp to illustrate the process of requesting an EC2
Fleet using instance weighting.

557
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Objective
Example Corp, a pharmaceutical company, wants to use the computational power of Amazon EC2 for
screening chemical compounds that might be used to fight cancer.

Planning
Example Corp first reviews Spot Best Practices. Next, Example Corp determines the requirements for
their EC2 Fleet.

Instance types

Example Corp has a compute- and memory-intensive application that performs best with at least 60 GB
of memory and eight virtual CPUs (vCPUs). They want to maximize these resources for the application at
the lowest possible price. Example Corp decides that any of the following EC2 instance types would meet
their needs:

Instance type Memory (GiB) vCPUs

r3.2xlarge 61 8

r3.4xlarge 122 16

r3.8xlarge 244 32

Target capacity in units

With instance weighting, target capacity can equal a number of instances (the default) or a combination
of factors such as cores (vCPUs), memory (GiBs), and storage (GBs). By considering the base for their
application (60 GB of RAM and eight vCPUs) as one unit, Example Corp decides that 20 times this
amount would meet their needs. So the company sets the target capacity of their EC2 Fleet request to
20.

Instance weights

After determining the target capacity, Example Corp calculates instance weights. To calculate the
instance weight for each instance type, they determine the units of each instance type that are required
to reach the target capacity as follows:

• r3.2xlarge (61.0 GB, 8 vCPUs) = 1 unit of 20

• r3.4xlarge (122.0 GB, 16 vCPUs) = 2 units of 20
• r3.8xlarge (244.0 GB, 32 vCPUs) = 4 units of 20

Therefore, Example Corp assigns instance weights of 1, 2, and 4 to the respective launch configurations
in their EC2 Fleet request.

Price per unit hour

Example Corp uses the On-Demand price per instance hour as a starting point for their price. They could
also use recent Spot prices, or a combination of the two. To calculate the price per unit hour, they divide
their starting price per instance hour by the weight. For example:

Instance type On-Demand price Instance weight Price per unit hour

r3.2xLarge $0.7 1 $0.7

r3.4xLarge $1.4 2 $0.7

558
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Instance type On-Demand price Instance weight Price per unit hour

r3.8xLarge $2.8 4 $0.7

Example Corp could use a global price per unit hour of $0.7 and be competitive for all three instance
types. They could also use a global price per unit hour of $0.7 and a specific price per unit hour of $0.9 in
the r3.8xlarge launch specification.

Verify permissions

Before creating an EC2 Fleet, Example Corp verifies that it has an IAM role with the required permissions.
For more information, see EC2 Fleet prerequisites (p. 562).

Create a launch template

Next, Example Corp creates a launch template. The launch template ID is used in the following step. For
more information, see Create a launch template (p. 534).

Create the EC2 Fleet

Example Corp creates a file, config.json, with the following configuration for its EC2 Fleet. In the
following example, replace the resource identifiers with your own resource identifiers.

{
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-07b3bc7625cdab851",
"Version": "1"
},
"Overrides": [
{
"InstanceType": "r3.2xlarge",
"SubnetId": "subnet-482e4972",
"WeightedCapacity": 1
},
{
"InstanceType": "r3.4xlarge",
"SubnetId": "subnet-482e4972",
"WeightedCapacity": 2
},
{
"InstanceType": "r3.8xlarge",
"MaxPrice": "0.90",
"SubnetId": "subnet-482e4972",
"WeightedCapacity": 4
}
]
}
],
"TargetCapacitySpecification": {
"TotalTargetCapacity": 20,
"DefaultTargetCapacityType": "spot"
}
}

Example Corp creates the EC2 Fleet using the following create-fleet command.

aws ec2 create-fleet \

--cli-input-json file://config.json

559
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

For more information, see Create an EC2 Fleet (p. 570).

Fulfillment
The allocation strategy determines which Spot Instance pools your Spot Instances come from.

With the lowest-price strategy (which is the default strategy), the Spot Instances come from the pool
with the lowest price per unit at the time of fulfillment. To provide 20 units of capacity, the EC2 Fleet
launches either 20 r3.2xlarge instances (20 divided by 1), 10 r3.4xlarge instances (20 divided by 2),
or 5 r3.8xlarge instances (20 divided by 4).

If Example Corp used the diversified strategy, the Spot Instances would come from all three pools.
The EC2 Fleet would launch 6 r3.2xlarge instances (which provide 6 units), 3 r3.4xlarge instances
(which provide 6 units), and 2 r3.8xlarge instances (which provide 8 units), for a total of 20 units.

Tutorial: Use EC2 Fleet with On-Demand as the primary capacity

This tutorial uses a fictitious company called ABC Online to illustrate the process of requesting an EC2
Fleet with On-Demand as the primary capacity, and Spot capacity if available.

Objective
ABC Online, a restaurant delivery company, wants to be able to provision Amazon EC2 capacity across
EC2 instance types and purchasing options to achieve their desired scale, performance, and cost.

Plan
ABC Online requires a fixed capacity to operate during peak periods, but would like to benefit from
increased capacity at a lower price. ABC Online determines the following requirements for their EC2
Fleet:

• On-Demand Instance capacity – ABC Online requires 15 On-Demand Instances to ensure that they can
accommodate traffic at peak periods.
• Spot Instance capacity – ABC Online would like to improve performance, but at a lower price, by
provisioning 5 Spot Instances.

Verify permissions
Before creating an EC2 Fleet, ABC Online verifies that it has an IAM role with the required permissions.
For more information, see EC2 Fleet prerequisites (p. 562).

Create a launch template

Next, ABC Online creates a launch template. The launch template ID is used in the following step. For
more information, see Create a launch template (p. 534).

Create the EC2 Fleet

ABC Online creates a file, config.json, with the following configuration for its EC2 Fleet. In the
following example, replace the resource identifiers with your own resource identifiers.

{
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-07b3bc7625cdab851",
"Version": "2"
}

}
],

560
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

"TargetCapacitySpecification": {
"TotalTargetCapacity": 20,
"OnDemandTargetCapacity":15,
"DefaultTargetCapacityType": "spot"
}
}

ABC Online creates the EC2 Fleet using the following create-fleet command.

aws ec2 create-fleet \

--cli-input-json file://config.json

For more information, see Create an EC2 Fleet (p. 570).

Fulfillment
The allocation strategy determines that the On-Demand capacity is always fulfilled, while the balance of
the target capacity is fulfilled as Spot if there is capacity and availability.

Work with EC2 Fleets

To start using an EC2 Fleet, you create a request that includes the total target capacity, On-Demand
capacity, Spot capacity, one or more launch specifications for the instances, and the maximum price that
you are willing to pay. The fleet request must include a launch template that defines the information
that the fleet needs to launch an instance, such as an AMI, instance type, subnet or Availability Zone, and
one or more security groups. You can specify launch specification overrides for the instance type, subnet,
Availability Zone, and maximum price you're willing to pay, and you can assign weighted capacity to each
launch specification override.

If your fleet includes Spot Instances, Amazon EC2 can attempt to maintain your fleet target capacity as
Spot prices change.

An EC2 Fleet request of type maintain or request remains active until it expires or you delete it. When
you delete a fleet of type maintain or request, you can specify whether deletion terminates the
instances in that fleet.

Contents
• EC2 Fleet request states (p. 561)
• EC2 Fleet prerequisites (p. 562)
• EC2 Fleet health checks (p. 565)
• Generate an EC2 Fleet JSON configuration file (p. 565)
• Create an EC2 Fleet (p. 570)
• Tag an EC2 Fleet (p. 572)
• Monitor your EC2 Fleet (p. 574)
• Modify an EC2 Fleet (p. 575)
• Delete an EC2 Fleet (p. 576)
• EC2 Fleet example configurations (p. 579)

EC2 Fleet request states

An EC2 Fleet request can be in one of the following states:

submitted

The EC2 Fleet request is being evaluated and Amazon EC2 is preparing to launch the target number
of instances. The request can include On-Demand Instances, Spot Instances, or both.

561
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

active

The EC2 Fleet request has been validated and Amazon EC2 is attempting to maintain the target
number of running instances. The request remains in this state until it is modified or deleted.
modifying

The EC2 Fleet request is being modified. The request remains in this state until the modification is
fully processed or the request is deleted. Only a maintain fleet type can be modified. This state
does not apply to other request types.
deleted_running

The EC2 Fleet request is deleted and does not launch additional instances. Its existing instances
continue to run until they are interrupted or terminated manually. The request remains in this state
until all instances are interrupted or terminated. Only an EC2 Fleet of type maintain or request
can have running instances after the EC2 Fleet request is deleted. A deleted instant fleet with
running instances is not supported. This state does not apply to instant fleets.
deleted_terminating

The EC2 Fleet request is deleted and its instances are terminating. The request remains in this state
until all instances are terminated.
deleted

The EC2 Fleet is deleted and has no running instances. The request is deleted two days after its
instances are terminated.

The following illustration represents the transitions between the EC2 Fleet request states. If you exceed
your fleet limits, the request is deleted immediately.

EC2 Fleet prerequisites

To create an EC2 Fleet, the following prerequisites must be in place:

• Launch template (p. 563)
• Service-linked role for EC2 Fleet (p. 563)
• Grant access to CMKs for use with encrypted AMIs and EBS snapshots (p. 563)
• Permissions for EC2 Fleet IAM users (p. 564)

562
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Launch template

A launch template includes information about the instances to launch, such as the instance type,
Availability Zone, and the maximum price that you are willing to pay. For more information, see Launch
an instance from a launch template (p. 532).

Service-linked role for EC2 Fleet

The AWSServiceRoleForEC2Fleet role grants the EC2 Fleet permission to request, launch, terminate,
and tag instances on your behalf. Amazon EC2 uses this service-linked role to complete the following
actions:

• ec2:RunInstances – Launch instances.

• ec2:RequestSpotInstances – Request Spot Instances.
• ec2:TerminateInstances – Terminate instances.
• ec2:DescribeImages – Describe Amazon Machine Images (AMIs) for the Spot Instances.
• ec2:DescribeInstanceStatus – Describe the status of the Spot Instances.
• ec2:DescribeSubnets – Describe the subnets for Spot Instances.
• ec2:CreateTags – Add tags to the EC2 Fleet, instances, and volumes.

Ensure that this role exists before you use the AWS CLI or an API to create an EC2 Fleet.
Note
An instant EC2 Fleet does not require this role.

To create the role, use the IAM console as follows.

To create the AWSServiceRoleForEC2Fleet role for EC2 Fleet

1. Open the IAM console at https://console.aws.amazon.com/iam/.

2. In the navigation pane, choose Roles, and then choose Create role.
3. For Select type of trusted entity, choose AWS service.
4. For Choose the service that will use this role, choose EC2 - Fleet, and then choose Next:
Permissions, Next: Tags, and Next: Review.
5. On the Review page, choose Create role.

If you no longer need to use EC2 Fleet, we recommend that you delete the AWSServiceRoleForEC2Fleet
role. After this role is deleted from your account, you can create the role again if you create another fleet.

For more information, see Using service-linked roles in the IAM User Guide.

Grant access to CMKs for use with encrypted AMIs and EBS snapshots

If you specify an encrypted AMI (p. 161) or an encrypted Amazon EBS snapshot (p. 1176) in your EC2
Fleet and you use a customer-managed customer master key (CMK) for encryption, you must grant the
AWSServiceRoleForEC2Fleet role permission to use the CMK so that Amazon EC2 can launch instances
on your behalf. To do this, you must add a grant to the CMK, as shown in the following procedure.

When providing permissions, grants are an alternative to key policies. For more information, see Using
grants and Using key policies in AWS KMS in the AWS Key Management Service Developer Guide.

To grant the AWSServiceRoleForEC2Fleet role permissions to use the CMK

• Use the create-grant command to add a grant to the CMK and to specify the principal (the
AWSServiceRoleForEC2Fleet service-linked role) that is given permission to perform the operations

563
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

that the grant permits. The CMK is specified by the key-id parameter and the ARN of the
CMK. The principal is specified by the grantee-principal parameter and the ARN of the
AWSServiceRoleForEC2Fleet service-linked role.

aws kms create-grant \

--region us-east-1 \
--key-id arn:aws:kms:us-
east-1:444455556666:key/1234abcd-12ab-34cd-56ef-1234567890ab \
--grantee-principal arn:aws:iam::111122223333:role/AWSServiceRoleForEC2Fleet \
--operations "Decrypt" "Encrypt" "GenerateDataKey"
"GenerateDataKeyWithoutPlaintext" "CreateGrant" "DescribeKey" "ReEncryptFrom"
"ReEncryptTo"

Permissions for EC2 Fleet IAM users

If your IAM users will create or manage an EC2 Fleet, be sure to grant them the required permissions as
follows.

To grant an IAM user permissions for EC2 Fleet

1. Open the IAM console at https://console.aws.amazon.com/iam/.

2. In the navigation pane, choose Policies.
3. Choose Create policy.
4. On the Create policy page, choose the JSON tab, replace the text with the following, and choose
Review policy.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:*"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": [
"iam:ListRoles",
"iam:PassRole",
"iam:ListInstanceProfiles"
],
"Resource": "*"
}
]
}

The ec2:* grants an IAM user permission to call all Amazon EC2 API actions. To limit the user to
specific Amazon EC2 API actions, specify those actions instead.

An IAM user must have permission to call the iam:ListRoles action to enumerate
existing IAM roles, the iam:PassRole action to specify the EC2 Fleet role, and the
iam:ListInstanceProfiles action to enumerate existing instance profiles.

(Optional) To enable an IAM user to create roles or instance profiles using the IAM console, you must
also add the following actions to the policy:

• iam:AddRoleToInstanceProfile

564
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

• iam:AttachRolePolicy
• iam:CreateInstanceProfile
• iam:CreateRole
• iam:GetRole
• iam:ListPolicies
5. On the Review policy page, enter a policy name and description, and choose Create policy.
6. In the navigation pane, choose Users and select the user.
7. On the Permissions tab, choose Add permissions.
8. Choose Attach existing policies directly. Select the policy that you created earlier and choose Next:
Review.
9. Choose Add permissions.

EC2 Fleet health checks

EC2 Fleet checks the health status of the instances in the fleet every two minutes. The health status
of an instance is either healthy or unhealthy. The fleet determines the health status of an instance
using the status checks provided by Amazon EC2. If the status of either the instance status check or the
system status check is impaired for three consecutive health checks, the health status of the instance is
unhealthy. Otherwise, the health status is healthy. For more information, see Status checks for your
instances (p. 743).

You can configure your EC2 Fleet to replace unhealthy instances. After enabling health check
replacement, an instance is replaced after its health status is reported as unhealthy. The fleet could go
below its target capacity for up to a few minutes while an unhealthy instance is being replaced.

Requirements

• Health check replacement is supported only with EC2 Fleets that maintain a target capacity (fleets of
type maintain), not with one-time fleets (fleets of type request or instant).
• You can configure your EC2 Fleet to replace unhealthy instances only when you create it.
• IAM users can use health check replacement only if they have permission to call the
ec2:DescribeInstanceStatus action.

Generate an EC2 Fleet JSON configuration file

To create an EC2 Fleet, you need only specify the launch template, total target capacity, and whether
the default purchasing option is On-Demand or Spot. If you do not specify a parameter, the fleet uses
the default value. To view the full list of fleet configuration parameters, you can generate a JSON file as
follows.

To generate a JSON file with all possible EC2 Fleet parameters using the command line

• Use the create-fleet (AWS CLI) command and the --generate-cli-skeleton parameter to
generate an EC2 Fleet JSON file:

aws ec2 create-fleet \

--generate-cli-skeleton

The following EC2 Fleet parameters are available:

{
"DryRun": true,
"ClientToken": "",

565
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

"SpotOptions": {
"AllocationStrategy": "lowest-price",
"InstanceInterruptionBehavior": "hibernate",
"InstancePoolsToUseCount": 0,
"SingleInstanceType": true,
"SingleAvailabilityZone": true,
"MaxTotalPrice": 0,
"MinTargetCapacity": 0
},
"OnDemandOptions": {
"AllocationStrategy": "prioritized",
"SingleInstanceType": true,
"SingleAvailabilityZone": true,
"MaxTotalPrice": 0,
"MinTargetCapacity": 0
},
"ExcessCapacityTerminationPolicy": "termination",
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "",
"LaunchTemplateName": "",
"Version": ""
},
"Overrides": [
{
"InstanceType": "t2.micro",
"MaxPrice": "",
"SubnetId": "",
"AvailabilityZone": "",
"WeightedCapacity": null,
"Priority": null,
"Placement": {
"AvailabilityZone": "",
"Affinity": "",
"GroupName": "",
"PartitionNumber": 0,
"HostId": "",
"Tenancy": "dedicated",
"SpreadDomain": ""
}
]
}
],
"TargetCapacitySpecification": {
"TotalTargetCapacity": 0,
"OnDemandTargetCapacity": 0,
"SpotTargetCapacity": 0,
"DefaultTargetCapacityType": "spot"
},
"TerminateInstancesWithExpiration": true,
"Type": "maintain",
"ValidFrom": "1970-01-01T00:00:00",
"ValidUntil": "1970-01-01T00:00:00",
"ReplaceUnhealthyInstances": true,
"TagSpecifications": [
{
"ResourceType": "fleet",
"Tags": [
{
"Key": "",
"Value": ""
}
]
}
]

566
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

EC2 Fleet JSON configuration file reference

Note
Use lowercase for all parameter values; otherwise, you get an error when Amazon EC2 uses the
JSON file to launch the EC2 Fleet.

AllocationStrategy (for SpotOptions)

(Optional) Indicates how to allocate the Spot Instance target capacity across the Spot Instance
pools specified by the EC2 Fleet. Valid values are lowest-price, diversified, and capacity-
optimized. The default is lowest-price. Specify the allocation strategy that meets your needs.
For more information, see Allocation strategies for Spot Instances (p. 551).
InstanceInterruptionBehavior

(Optional) The behavior when a Spot Instance is interrupted. Valid values are hibernate, stop, and
terminate. By default, the Spot service terminates Spot Instances when they are interrupted. If
the fleet type is maintain, you can specify that the Spot service hibernates or stops Spot Instances
when they are interrupted.
InstancePoolsToUseCount

The number of Spot pools across which to allocate your target Spot capacity. Valid only when Spot
AllocationStrategy is set to lowest-price. EC2 Fleet selects the cheapest Spot pools and evenly
allocates your target Spot capacity across the number of Spot pools that you specify.
SingleInstanceType

Indicates that the fleet uses a single instance type to launch all Spot Instances in the fleet.
SingleAvailabilityZone

Indicates that the fleet launches all Spot Instances into a single Availability Zone.
MaxTotalPrice

The maximum amount per hour for Spot Instances that you're willing to pay.
MinTargetCapacity

The minimum target capacity for Spot Instances in the fleet. If the minimum target capacity is not
reached, the fleet launches no instances.
AllocationStrategy (for OnDemandOptions)

The order of the launch template overrides to use in fulfilling On-Demand capacity. If you specify
lowest-price, EC2 Fleet uses price to determine the order, launching the lowest price first. If you
specify prioritized, EC2 Fleet uses the priority that you assigned to each launch template override,
launching the highest priority first. If you do not specify a value, EC2 Fleet defaults to lowest-
price.
SingleInstanceType

Indicates that the fleet uses a single instance type to launch all On-Demand Instances in the fleet.
SingleAvailabilityZone

Indicates that the fleet launches all On-Demand Instances into a single Availability Zone.
MaxTotalPrice

The maximum amount per hour for On-Demand Instances that you're willing to pay.
MinTargetCapacity

The minimum target capacity for On-Demand Instances in the fleet. If the minimum target capacity
is not reached, the fleet launches no instances.

567
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

ExcessCapacityTerminationPolicy

(Optional) Indicates whether running instances should be terminated if the total target capacity
of the EC2 Fleet is decreased below the current size of the EC2 Fleet. Valid values are no-
termination and termination.
LaunchTemplateId

The ID of the launch template to use. You must specify either the launch template ID or launch
template name. The launch template must specify an Amazon Machine Image (AMI). For information
about creating launch templates, see Launch an instance from a launch template (p. 532).
LaunchTemplateName

The name of the launch template to use. You must specify either the launch template ID or launch
template name. The launch template must specify an Amazon Machine Image (AMI). For more
information, see Launch an instance from a launch template (p. 532).
Version

The launch template version number, $Latest, or $Default. You must specify a value, otherwise
the request fails. If the value is $Latest, Amazon EC2 uses the latest version of the launch
template. If the value is $Default, Amazon EC2 uses the default version of the launch template. For
more information, see Manage launch template versions (p. 539).
InstanceType

(Optional) The instance type. If entered, this value overrides the launch template. The instance types
must have the minimum hardware specifications that you need (vCPUs, memory, or storage).
MaxPrice

(Optional) The maximum price per unit hour that you are willing to pay for a Spot Instance. If
entered, this value overrides the launch template. You can use the default maximum price (the On-
Demand price) or specify the maximum price that you are willing to pay. Your Spot Instances are
not launched if your maximum price is lower than the Spot price for the instance types that you
specified.
SubnetId

(Optional) The ID of the subnet in which to launch the instances. If entered, this value overrides the
launch template.

To create a new VPC, go the Amazon VPC console. When you are done, return to the JSON file and
enter the new subnet ID.
AvailabilityZone

(Optional) The Availability Zone in which to launch the instances. The default is to let AWS choose
the zones for your instances. If you prefer, you can specify specific zones. If entered, this value
overrides the launch template.

Specify one or more Availability Zones. If you have more than one subnet in a zone, specify the
appropriate subnet. To add subnets, go to the Amazon VPC console. When you are done, return to
the JSON file and enter the new subnet ID.
WeightedCapacity

(Optional) The number of units provided by the specified instance type. If entered, this value
overrides the launch template.
Priority

The priority for the launch template override. If AllocationStrategy is set to prioritized, EC2
Fleet uses priority to determine which launch template override to use first in fulfilling On-Demand

568
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

capacity. The highest priority is launched first. Valid values are whole numbers starting at 0. The
lower the number, the higher the priority. If no number is set, the override has the lowest priority.
TotalTargetCapacity

The number of instances to launch. You can choose instances or performance characteristics that are
important to your application workload, such as vCPUs, memory, or storage. If the request type is
maintain, you can specify a target capacity of 0 and add capacity later.
OnDemandTargetCapacity

(Optional) The number of On-Demand Instances to launch. This number must be less than the
TotalTargetCapacity.
SpotTargetCapacity

(Optional) The number of Spot Instances to launch. This number must be less than the
TotalTargetCapacity.
DefaultTargetCapacityType

If the value for TotalTargetCapacity is higher than the combined values for
OnDemandTargetCapacity and SpotTargetCapacity, the difference is launched as the instance
purchasing option specified here. Valid values are on-demand or spot.
TerminateInstancesWithExpiration

(Optional) By default, Amazon EC2 terminates your instances when the EC2 Fleet request expires.
The default value is true. To keep them running after your request expires, do not enter a value for
this parameter.
Type

(Optional) The type of request. Valid values are instant, request, and maintain. The default
value is maintain.
• instant – The EC2 Fleet submits a synchronous one-time request for your desired capacity, and
returns errors for any instances that could not be launched.
• request – The EC2 Fleet submits an asynchronous one-time request for your desired capacity, but
does submit Spot requests in alternative capacity pools if Spot capacity is unavailable, and does
not maintain Spot capacity if Spot Instances are interrupted.
• maintain – The EC2 Fleet submits an asynchronous request for your desired capacity, and
continues to maintain your desired Spot capacity by replenishing interrupted Spot Instances.

For more information, see EC2 Fleet request types (p. 551).
ValidFrom

(Optional) To create a request that is valid only during a specific time period, enter a start date.
ValidUntil

(Optional) To create a request that is valid only during a specific time period, enter an end date.
ReplaceUnhealthyInstances

(Optional) To replace unhealthy instances in an EC2 Fleet that is configured to maintain the fleet,
enter true. Otherwise, leave this parameter empty.
TagSpecifications

(Optional) The key-value pair for tagging the EC2 Fleet request on creation. The value for
ResourceType must be fleet, otherwise the fleet request fails. To tag instances at launch, specify
the tags in the launch template (p. 534). For information about tagging after launch, see Tag your
resources (p. 1313).

569
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Create an EC2 Fleet

When you create an EC2 Fleet, you must specify a launch template that includes information about the
instances to launch, such as the instance type, Availability Zone, and the maximum price you are willing
to pay.

You can create an EC2 Fleet that includes multiple launch specifications that override the launch
template. The launch specifications can vary by instance type, Availability Zone, subnet, and maximum
price, and can include a different weighted capacity.

When you create an EC2 Fleet, use a JSON file to specify information about the instances to launch. For
more information, see EC2 Fleet JSON configuration file reference (p. 567).

EC2 Fleets can only be created using the AWS CLI.

To create an EC2 Fleet (AWS CLI)

• Use the create-fleet (AWS CLI) command to create an EC2 Fleet.

aws ec2 create-fleet \

--cli-input-json file://file_name.json

For example configuration files, see EC2 Fleet example configurations (p. 579).

The following is example output for a fleet of type request or maintain.

{
"FleetId": "fleet-12a34b55-67cd-8ef9-ba9b-9208dEXAMPLE"
}

The following is example output for a fleet of type instant that launched the target capacity.

{
"FleetId": "fleet-12a34b55-67cd-8ef9-ba9b-9208dEXAMPLE",
"Errors": [],
"Instances": [
{
"LaunchTemplateAndOverrides": {
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-01234a567b8910abcEXAMPLE",
"Version": "1"
},
"Overrides": {
"InstanceType": "c5.large",
"AvailabilityZone": "us-east-1a"
}
},
"Lifecycle": "on-demand",
"InstanceIds": [
"i-1234567890abcdef0",
"i-9876543210abcdef9"
],
"InstanceType": "c5.large",
"Platform": null
},
{
"LaunchTemplateAndOverrides": {
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-01234a567b8910abcEXAMPLE",

570
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

"Version": "1"
},
"Overrides": {
"InstanceType": "c4.large",
"AvailabilityZone": "us-east-1a"
}
},
"Lifecycle": "on-demand",
"InstanceIds": [
"i-5678901234abcdef0",
"i-5432109876abcdef9"
],
"InstanceType": "c4.large",
"Platform": null
},
]
}

The following is example output for a fleet of type instant that launched part of the target capacity
with errors for instances that were not launched.

{
"FleetId": "fleet-12a34b55-67cd-8ef9-ba9b-9208dEXAMPLE",
"Errors": [
{
"LaunchTemplateAndOverrides": {
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-01234a567b8910abcEXAMPLE",
"Version": "1"
},
"Overrides": {
"InstanceType": "c4.xlarge",
"AvailabilityZone": "us-east-1a",
}
},
"Lifecycle": "on-demand",
"ErrorCode": "InsufficientInstanceCapacity",
"ErrorMessage": "",
"InstanceType": "c4.xlarge",
"Platform": null
},
],
"Instances": [
{
"LaunchTemplateAndOverrides": {
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-01234a567b8910abcEXAMPLE",
"Version": "1"
},
"Overrides": {
"InstanceType": "c5.large",
"AvailabilityZone": "us-east-1a"
}
},
"Lifecycle": "on-demand",
"InstanceIds": [
"i-1234567890abcdef0",
"i-9876543210abcdef9"
],
"InstanceType": "c5.large",
"Platform": null
},
]
}

571
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

The following is example output for a fleet of type instant that launched no instances.

{
"FleetId": "fleet-12a34b55-67cd-8ef9-ba9b-9208dEXAMPLE",
"Errors": [
{
"LaunchTemplateAndOverrides": {
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-01234a567b8910abcEXAMPLE",
"Version": "1"
},
"Overrides": {
"InstanceType": "c4.xlarge",
"AvailabilityZone": "us-east-1a",
}
},
"Lifecycle": "on-demand",
"ErrorCode": "InsufficientCapacity",
"ErrorMessage": "",
"InstanceType": "c4.xlarge",
"Platform": null
},
{
"LaunchTemplateAndOverrides": {
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-01234a567b8910abcEXAMPLE",
"Version": "1"
},
"Overrides": {
"InstanceType": "c5.large",
"AvailabilityZone": "us-east-1a",
}
},
"Lifecycle": "on-demand",
"ErrorCode": "InsufficientCapacity",
"ErrorMessage": "",
"InstanceType": "c5.large",
"Platform": null
},
],
"Instances": []
}

Tag an EC2 Fleet

To help categorize and manage your EC2 Fleet requests, you can tag them with custom metadata. You
can assign a tag to an EC2 Fleet request when you create it, or afterward.

When you tag a fleet request, the instances and volumes that are launched by the fleet are not
automatically tagged. You need to explicitly tag the instances and volumes launched by the fleet. You
can choose to assign tags to only the fleet request, or to only the instances launched by the fleet, or to
only the volumes attached to the instances launched by the fleet, or to all three.
Note
For instant fleet types, you can tag volumes that are attached to On-Demand Instances
and Spot Instances. For request or maintain fleet types, you can only tag volumes that are
attached to On-Demand Instances.

For more information about how tags work, see Tag your Amazon EC2 resources (p. 1311).

Prerequisite

Grant the IAM user the permission to tag resources. For more information, see Example: Tag
resources (p. 1016).

572
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

To grant an IAM user the permission to tag resources

Create a IAM policy that includes the following:

• The ec2:CreateTags action. This grants the IAM user permission to create tags.
• The ec2:CreateFleet action. This grants the IAM user permission to create an EC2 Fleet request.
• For Resource, we recommend that you specify "*". This allows users to tag all resource types.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "TagEC2FleetRequest",
"Effect": "Allow",
"Action": [
"ec2:CreateTags",
"ec2:CreateFleet"
],
"Resource": "*"
}

Important
We currently do not support resource-level permissions for the create-fleet resource. If you
specify create-fleet as a resource, you will get an unauthorized exception when you try to
tag the fleet. The following example illustrates how not to set the policy.

{
"Effect": "Allow",
"Action": [
"ec2:CreateTags",
"ec2:CreateFleet"
],
"Resource": "arn:aws:ec2:us-east-1:111122223333:create-fleet/*"
}

To tag a new EC2 Fleet request

To tag an EC2 Fleet request when you create it, specify the key-value pair in the JSON file (p. 565) used
to create the fleet. The value for ResourceType must be fleet. If you specify another value, the fleet
request fails.

To tag instances and volumes launched by an EC2 Fleet

To tag instances and volumes when they are launched by the fleet, specify the tags in the launch
template (p. 534) that is referenced in the EC2 Fleet request.
Note
You can't tag volumes attached to Spot Instances that are launched by a request or maintain
fleet type.

To tag an existing EC2 Fleet request, instance, and volume (AWS CLI)

Use the create-tags command to tag existing resources.

aws ec2 create-tags \

--resources fleet-12a34b55-67cd-8ef9-
ba9b-9208dEXAMPLE i-1234567890abcdef0 vol-1234567890EXAMPLE \
--tags Key=purpose,Value=test

573
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Monitor your EC2 Fleet

The EC2 Fleet launches On-Demand Instances when there is available capacity, and launches Spot
Instances when your maximum price exceeds the Spot price and capacity is available. The On-Demand
Instances run until you terminate them, and the Spot Instances run until they are interrupted or you
terminate them.

The returned list of running instances is refreshed periodically and might be out of date.

To monitor your EC2 Fleet (AWS CLI)

Use the describe-fleets command to describe your EC2 Fleets.

aws ec2 describe-fleets

The following is example output.

{
"Fleets": [
{
"Type": "maintain",
"FulfilledCapacity": 2.0,
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"Version": "2",
"LaunchTemplateId": "lt-07b3bc7625cdab851"
}
}
],
"TerminateInstancesWithExpiration": false,
"TargetCapacitySpecification": {
"OnDemandTargetCapacity": 0,
"SpotTargetCapacity": 2,
"TotalTargetCapacity": 2,
"DefaultTargetCapacityType": "spot"
},
"FulfilledOnDemandCapacity": 0.0,
"ActivityStatus": "fulfilled",
"FleetId": "fleet-76e13e99-01ef-4bd6-ba9b-9208de883e7f",
"ReplaceUnhealthyInstances": false,
"SpotOptions": {
"InstanceInterruptionBehavior": "terminate",
"InstancePoolsToUseCount": 1,
"AllocationStrategy": "lowest-price"
},
"FleetState": "active",
"ExcessCapacityTerminationPolicy": "termination",
"CreateTime": "2018-04-10T16:46:03.000Z"
}
]
}

Use the describe-fleet-instances command to describe the instances for the specified EC2 Fleet.

aws ec2 describe-fleet-instances \

--fleet-id fleet-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE

{
"ActiveInstances": [

574
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

{
"InstanceId": "i-09cd595998cb3765e",
"InstanceHealth": "healthy",
"InstanceType": "m4.large",
"SpotInstanceRequestId": "sir-86k84j6p"
},
{
"InstanceId": "i-09cf95167ca219f17",
"InstanceHealth": "healthy",
"InstanceType": "m4.large",
"SpotInstanceRequestId": "sir-dvxi7fsm"
}
],
"FleetId": "fleet-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE"
}

Use the describe-fleet-history command to describe the history for the specified EC2 Fleet for the
specified time.

aws ec2 describe-fleet-history --fleet-request-id fleet-73fbd2ce-

aa30-494c-8788-1cee4EXAMPLE --start-time 2018-04-10T00:00:00Z

{
"HistoryRecords": [],
"FleetId": "fleet-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE",
"LastEvaluatedTime": "1970-01-01T00:00:00.000Z",
"StartTime": "2018-04-09T23:53:20.000Z"
}

Modify an EC2 Fleet

You can modify an EC2 Fleet that is in the submitted or active state. When you modify a fleet, it
enters the modifying state.

You can only modify an EC2 Fleet that is of type maintain. You cannot modify an EC2 Fleet of type
request or instant.

You can modify the following parameters of an EC2 Fleet:

• target-capacity-specification – Increase or decrease the target capacity for

TotalTargetCapacity, OnDemandTargetCapacity, and SpotTargetCapacity.
• excess-capacity-termination-policy – Whether running instances should be terminated if the
total target capacity of the EC2 Fleet is decreased below the current size of the fleet. Valid values are
no-termination and termination.

When you increase the target capacity, the EC2 Fleet launches the additional instances according to the
instance purchasing option specified for DefaultTargetCapacityType, which are either On-Demand
Instances or Spot Instances.

If the DefaultTargetCapacityType is spot, the EC2 Fleet launches the additional Spot Instances
according to its allocation strategy. If the allocation strategy is lowest-price, the fleet launches
the instances from the lowest-priced Spot Instance pool in the request. If the allocation strategy is
diversified, the fleet distributes the instances across the pools in the request.

When you decrease the target capacity, the EC2 Fleet deletes any open requests that exceed the new
target capacity. You can request that the fleet terminate instances until the size of the fleet reaches
the new target capacity. If the allocation strategy is lowest-price, the fleet terminates the instances
with the highest price per unit. If the allocation strategy is diversified, the fleet terminates instances

575
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

across the pools. Alternatively, you can request that EC2 Fleet keep the fleet at its current size, but not
replace any Spot Instances that are interrupted or any instances that you terminate manually.

When an EC2 Fleet terminates a Spot Instance because the target capacity was decreased, the instance
receives a Spot Instance interruption notice.

To modify an EC2 Fleet (AWS CLI)

Use the modify-fleet command to update the target capacity of the specified EC2 Fleet.

aws ec2 modify-fleet \

--fleet-id fleet-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE \
--target-capacity-specification TotalTargetCapacity=20

If you are decreasing the target capacity but want to keep the fleet at its current size, you can modify the
previous command as follows.

aws ec2 modify-fleet \

--fleet-id fleet-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE \
--target-capacity-specification TotalTargetCapacity=10 \
--excess-capacity-termination-policy no-termination

Delete an EC2 Fleet

If you no longer require an EC2 Fleet, you can delete it. After you delete a fleet, it launches no new
instances.

When you delete an EC2 Fleet, you must specify if you want to also terminate its instances. If you specify
that the instances must be terminated when the fleet is deleted, it enters the deleted_terminating
state. Otherwise, it enters the deleted_running state, and the instances continue to run until they are
interrupted or you terminate them manually.

Restrictions

• You can delete up to 25 instant fleets in a single request. If you exceed this number, no instant
fleets are deleted and an error is returned. There is no restriction on the number of fleets of type
maintain or request that can be deleted in a single request.
• Up to 1000 instances can be terminated in a single request to delete instant fleets.

To delete an EC2 Fleet and terminate its instances (AWS CLI)

Use the delete-fleets command and the --terminate-instances parameter to delete the specified
EC2 Fleet and terminate the instances.

aws ec2 delete-fleets \

--fleet-ids fleet-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE \
--terminate-instances

The following is example output.

{
"UnsuccessfulFleetDeletions": [],
"SuccessfulFleetDeletions": [
{
"CurrentFleetState": "deleted_terminating",
"PreviousFleetState": "active",

576
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

"FleetId": "fleet-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE"
}
]
}

To delete an EC2 Fleet without terminating the instances (AWS CLI)

You can modify the previous command using the --no-terminate-instances parameter to delete
the specified EC2 Fleet without terminating the instances.
Note
--no-terminate-instances is not supported for instant fleets.

aws ec2 delete-fleets \

--fleet-ids fleet-73fbd2ce-aa30-494c-8788-1cee4EXAMPLE \
--no-terminate-instances

The following is example output.

{
"UnsuccessfulFleetDeletions": [],
"SuccessfulFleetDeletions": [
{
"CurrentFleetState": "deleted_running",
"PreviousFleetState": "active",
"FleetId": "fleet-4b8aaae8-dfb5-436d-a4c6-3dafa4c6b7dcEXAMPLE"
}
]
}

Troubleshoot when a fleet fails to delete

If an EC2 Fleet fails to delete, UnsuccessfulFleetDeletions in the output returns the ID of the EC2
Fleet, an error code, and an error message.

The error codes are:

• ExceededInstantFleetNumForDeletion
• fleetIdDoesNotExist
• fleetIdMalformed
• fleetNotInDeletableState
• NoTerminateInstancesNotSupported
• UnauthorizedOperation
• unexpectedError

Troubleshooting ExceededInstantFleetNumForDeletion

If you try to delete more than 25 instant fleets in a single request, the
ExceededInstantFleetNumForDeletion error is returned. The following is example output for this
error.

{
"UnsuccessfulFleetDeletions": [
{
"FleetId": " fleet-5d130460-0c26-bfd9-2c32-0100a098f625",

577
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

"Error": {
"Message": "Can’t delete more than 25 instant fleets in a single
request.",
"Code": "ExceededInstantFleetNumForDeletion"
}
},
{
"FleetId": "fleet-9a941b23-0286-5bf4-2430-03a029a07e31",
"Error": {
"Message": "Can’t delete more than 25 instant fleets in a single
request.",
"Code": "ExceededInstantFleetNumForDeletion"
}
}
.
.
.
],
"SuccessfulFleetDeletions": []
}

Troubleshoot NoTerminateInstancesNotSupported

If you specify that the instances in an instant fleet must not be terminated when you delete the fleet,
the NoTerminateInstancesNotSupported error is returned. --no-terminate-instances is not
supported for instant fleets. The following is example output for this error.

{
"UnsuccessfulFleetDeletions": [
{
"FleetId": "fleet-5d130460-0c26-bfd9-2c32-0100a098f625",
"Error": {
"Message": "NoTerminateInstances option is not supported for
instant fleet",
"Code": "NoTerminateInstancesNotSupported"
}
}
],
"SuccessfulFleetDeletions": []

Troubleshoot UnauthorizedOperation

If you do not have permission to terminate instances, you get the UnauthorizedOperation error when
deleting a fleet that must terminate its instances. The following is the error response.

<Response><Errors><Error><Code>UnauthorizedOperation</Code><Message>You are not authorized

to perform this
operation. Encoded authorization failure message: VvuncIxj7Z_CPGNYXWqnuFV-
YjByeAU66Q9752NtQ-I3-qnDLWs6JLFd
KnSMMiq5s6cGqjjPtEDpsnGHzzyHasFHOaRYJpaDVravoW25azn6KNkUQQlFwhJyujt2dtNCdduJfrqcFYAjlEiRMkfDHt7N63SKlwe
BHturzDK6A560Y2nDSUiMmAB1y9UNtqaZJ9SNe5sNxKMqZaqKtjRbk02RZu5V2vn9VMk6fm2aMVHbY9JhLvGypLcMUjtJ76H9ytg2zR
VPiU5v2s-
UgZ7h0p2yth6ysUdhlONg6dBYu8_y_HtEI54invCj4CoK0qawqzMNe6rcmCQHvtCxtXsbkgyaEbcwmrm2m01-
EMhekLFZeJLr
DtYOpYcEl4_nWFX1wtQDCnNNCmxnJZAoJvb3VMDYpDTsxjQv1PxODZuqWHs23YXWVywzgnLtHeRf2o4lUhGBw17mXsS07k7XAfdPMP_
PT9vrHtQiILor5VVTsjSPWg7edj__1rsnXhwPSu8gI48ZLRGrPQqFq0RmKO_QIE8N8s6NWzCK4yoX-9gDcheurOGpkprPIC9YPGMLK9
</Message></Error></Errors><RequestID>89b1215c-7814-40ae-a8db-41761f43f2b0</RequestID></
Response>

To resolve the error, you must add the ec2:TerminateInstances action to the IAM policy, as shown in
the following example.

578
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "DeleteFleetsAndTerminateInstances",
"Effect": "Allow",
"Action": [
"ec2:DeleteFleets"
"ec2:TerminateInstances"
],
"Resource": "*"
}
]
}

EC2 Fleet example configurations

The following examples show launch configurations that you can use with the create-fleet command to
create an EC2 Fleet. For more information about the create-fleet parameters, see the EC2 Fleet JSON
configuration file reference (p. 567).

Examples
• Example 1: Launch Spot Instances as the default purchasing option (p. 579)
• Example 2: Launch On-Demand Instances as the default purchasing option (p. 580)
• Example 3: Launch On-Demand Instances as the primary capacity (p. 580)
• Example 4: Launch Spot Instances using the lowest-price allocation strategy (p. 580)
• Example 5: Launch On-Demand Instances using Capacity Reservations and the prioritized allocation
strategy (p. 581)
• Example 6: Launch On-Demand Instances using Capacity Reservations and the prioritized
allocation strategy when the total target capacity is more than the number of unused Capacity
Reservations (p. 583)
• Example 7: Launch On-Demand Instances using Capacity Reservations and the lowest-price
allocation strategy (p. 585)
• Example 8: Launch On-Demand Instances using Capacity Reservations and the lowest-price
allocation strategy when the total target capacity is more than the number of unused Capacity
Reservations (p. 588)
• Example 9: Configure Capacity Rebalancing to launch replacement Spot Instances (p. 590)

Example 1: Launch Spot Instances as the default purchasing option

The following example specifies the minimum parameters required in an EC2 Fleet: a launch template,
target capacity, and default purchasing option. The launch template is identified by its launch template
ID and version number. The target capacity for the fleet is 2 instances, and the default purchasing option
is spot, which results in the fleet launching 2 Spot Instances.

{
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-0e8c754449b27161c",
"Version": "1"
}

}
],
"TargetCapacitySpecification": {

579
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

"TotalTargetCapacity": 2,
"DefaultTargetCapacityType": "spot"
}
}

Example 2: Launch On-Demand Instances as the default purchasing option

The following example specifies the minimum parameters required in an EC2 Fleet: a launch template,
target capacity, and default purchasing option. The launch template is identified by its launch template
ID and version number. The target capacity for the fleet is 2 instances, and the default purchasing option
is on-demand, which results in the fleet launching 2 On-Demand Instances.

{
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-0e8c754449b27161c",
"Version": "1"
}

}
],
"TargetCapacitySpecification": {
"TotalTargetCapacity": 2,
"DefaultTargetCapacityType": "on-demand"
}
}

Example 3: Launch On-Demand Instances as the primary capacity

The following example specifies the total target capacity of 2 instances for the fleet, and a target
capacity of 1 On-Demand Instance. The default purchasing option is spot. The fleet launches 1
On-Demand Instance as specified, but needs to launch one more instance to fulfill the total target
capacity. The purchasing option for the difference is calculated as TotalTargetCapacity –
OnDemandTargetCapacity = DefaultTargetCapacityType, which results in the fleet launching 1
Spot Instance.

{
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-0e8c754449b27161c",
"Version": "1"
}

}
],
"TargetCapacitySpecification": {
"TotalTargetCapacity": 2,
"OnDemandTargetCapacity": 1,
"DefaultTargetCapacityType": "spot"
}
}

Example 4: Launch Spot Instances using the lowest-price allocation strategy

If the allocation strategy for Spot Instances is not specified, the default allocation strategy, which is
lowest-price, is used. The following example uses the lowest-price allocation strategy. The three
launch specifications, which override the launch template, have different instance types but the same
weighted capacity and subnet. The total target capacity is 2 instances and the default purchasing option

580
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

is spot. The EC2 Fleet launches 2 Spot Instances using the instance type of the launch specification with
the lowest price.

{
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-0e8c754449b27161c",
"Version": "1"
}
"Overrides": [
{
"InstanceType": "c4.large",
"WeightedCapacity": 1,
"SubnetId": "subnet-a4f6c5d3"
},
{
"InstanceType": "c3.large",
"WeightedCapacity": 1,
"SubnetId": "subnet-a4f6c5d3"
},
{
"InstanceType": "c5.large",
"WeightedCapacity": 1,
"SubnetId": "subnet-a4f6c5d3"
}
]

}
],
"TargetCapacitySpecification": {
"TotalTargetCapacity": 2,
"DefaultTargetCapacityType": "spot"
}
}

Example 5: Launch On-Demand Instances using Capacity Reservations and the prioritized
allocation strategy

You can configure a fleet to use On-Demand Capacity Reservations first when launching On-Demand
Instances by setting the usage strategy for Capacity Reservations to use-capacity-reservations-
first. And if multiple instance pools have unused Capacity Reservations, the chosen On-Demand
allocation strategy is applied. In this example, the On-Demand allocation strategy is prioritized.

In this example, there are 15 available unused Capacity Reservations. This is more than the fleet's target
On-Demand capacity of 12 On-Demand Instances.

The account has the following 15 unused Capacity Reservations in 3 different pools. The number of
Capacity Reservations in each pool is indicated by AvailableInstanceCount.

{
"CapacityReservationId": "cr-111",
"InstanceType": "c4.large",
"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount": 5,
"InstanceMatchCriteria": "open",
"State": "active"
}

{
"CapacityReservationId": "cr-222",

581
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

"InstanceType": "c3.large",
"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount": 5,
"InstanceMatchCriteria": "open",
"State": "active"
}

{
"CapacityReservationId": "cr-333",
"InstanceType": "c5.large",
"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount":5,
"InstanceMatchCriteria": "open",
"State": "active"
}

The following fleet configuration shows only the pertinent configurations for this example. The On-
Demand allocation strategy is prioritized, and the usage strategy for Capacity Reservations is use-
capacity-reservations-first. The total target capacity is 12, and the default target capacity type
is on-demand.
Note
The fleet type must be instant. use-capacity-reservations-first is not supported for
other fleet types.

{
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-1234567890abcdefg",
"Version": "1"
}
"Overrides": [
{
"InstanceType": "c4.large",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1,
"Priority": 1.0
},
{
"InstanceType": "c3.large",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1,
"Priority": 2.0
},
{
"InstanceType": "c5.large",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1,
"Priority": 3.0
}
]
}
],
"TargetCapacitySpecification": {
"TotalTargetCapacity": 12,
"DefaultTargetCapacityType": "on-demand"
},
"OnDemandOptions": {
"AllocationStrategy": "prioritized"
"CapacityReservationOptions": {
"UsageStrategy": "use-capacity-reservations-first"

582
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

}
},
"Type": "instant",
}

After you create the instant fleet using the preceding configuration, the following 12 instances are
launched to meet the target capacity:

• 5 c4.large On-Demand Instances in us-east-1a – c4.large in us-east-1a is prioritized first, and there are
5 available unused c4.large Capacity Reservations
• 5 c3.large On-Demand Instances in us-east-1a – c3.large in us-east-1a is prioritized second, and there
are 5 available unused c3.large Capacity Reservations
• 2 c5.large On-Demand Instances in us-east-1a – c5.large in us-east-1a is prioritized third, and there
are 5 available unused c5.large Capacity Reservations of which only 2 are needed to meet the target
capacity

After the fleet is launched, you can run describe-capacity-reservations to see how many unused Capacity
Reservations are remaining. In this example, you should see the following response, which shows that
all of the c4.large and c3.large Capacity Reservations were used, with 3 c5.large Capacity Reservations
remaining unused.

{
"CapacityReservationId": "cr-111",
"InstanceType": "c4.large",
"AvailableInstanceCount": 0
}

{
"CapacityReservationId": "cr-222",
"InstanceType": "c3.large",
"AvailableInstanceCount": 0
}

{
"CapacityReservationId": "cr-333",
"InstanceType": "c5.large",
"AvailableInstanceCount": 3
}

Example 6: Launch On-Demand Instances using Capacity Reservations and the prioritized
allocation strategy when the total target capacity is more than the number of unused Capacity
Reservations

You can configure a fleet to use On-Demand Capacity Reservations first when launching On-Demand
Instances by setting the usage strategy for Capacity Reservations to use-capacity-reservations-
first. And if the number of unused Capacity Reservations is less than the On-Demand target capacity,
the remaining On-Demand target capacity is launched according to the chosen On-Demand allocation
strategy. In this example, the On-Demand allocation strategy is prioritized.

In this example, there are 15 available unused Capacity Reservations. This is less than the fleet's On-
Demand target capacity of 16 On-Demand Instances.

The account has the following 15 unused Capacity Reservations in 3 different pools. The number of
Capacity Reservations in each pool is indicated by AvailableInstanceCount.

{
"CapacityReservationId": "cr-111",
"InstanceType": "c4.large",

583
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount": 5,
"InstanceMatchCriteria": "open",
"State": "active"
}

{
"CapacityReservationId": "cr-111",
"InstanceType": "c3.large",
"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount": 5,
"InstanceMatchCriteria": "open",
"State": "active"
}

{
"CapacityReservationId": "cr-111",
"InstanceType": "c5.large",
"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount":5,
"InstanceMatchCriteria": "open",
"State": "active"
}

The following fleet configuration shows only the pertinent configurations for this example. The On-
Demand allocation strategy is prioritized, and the usage strategy for Capacity Reservations is use-
capacity-reservations-first. The total target capacity is 16, and the default target capacity type
is on-demand.
Note
The fleet type must be instant. use-capacity-reservations-first is not supported for
other fleet types.

{
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-0e8c754449b27161c",
"Version": "1"
}
"Overrides": [
{
"InstanceType": "c4.large",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1,
"Priority": 1.0
},
{
"InstanceType": "c3.large",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1,
"Priority": 2.0
},
{
"InstanceType": "c5.large",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1,
"Priority": 3.0
}
]

584
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

}
],
"TargetCapacitySpecification": {
"TotalTargetCapacity": 16,
"DefaultTargetCapacityType": "on-demand"
},
"OnDemandOptions": {
"AllocationStrategy": "prioritized"
"CapacityReservationOptions": {
"UsageStrategy": "use-capacity-reservations-first"
}
},
"Type": "instant",
}

After you create the instant fleet using the preceding configuration, the following 16 instances are
launched to meet the target capacity:

• 6 c4.large On-Demand Instances in us-east-1a – c4.large in us-east-1a is prioritized first, and there are
5 available unused c4.large Capacity Reservations. The Capacity Reservations are used first to launch
5 On-Demand Instances plus an additional On-Demand Instance is launched according to the On-
Demand allocation strategy, which is prioritized in this example.
• 5 c3.large On-Demand Instances in us-east-1a – c3.large in us-east-1a is prioritized second, and there
are 5 available unused c3.large Capacity Reservations
• 5 c5.large On-Demand Instances in us-east-1a – c5.large in us-east-1a is prioritized third, and there are
5 available unused c5.large Capacity Reservations

After the fleet is launched, you can run describe-capacity-reservations to see how many unused Capacity
Reservations are remaining. In this example, you should see the following response, which shows that all
of the Capacity Reservations in all of the pools were used.

{
"CapacityReservationId": "cr-111",
"InstanceType": "c4.large",
"AvailableInstanceCount": 0
}

{
"CapacityReservationId": "cr-222",
"InstanceType": "c3.large",
"AvailableInstanceCount": 0
}

{
"CapacityReservationId": "cr-333",
"InstanceType": "c5.large",
"AvailableInstanceCount": 0
}

Example 7: Launch On-Demand Instances using Capacity Reservations and the lowest-price
allocation strategy

You can configure a fleet to use On-Demand Capacity Reservations first when launching On-Demand
Instances by setting the usage strategy for Capacity Reservations to use-capacity-reservations-
first. And if multiple instance pools have unused Capacity Reservations, the chosen On-Demand
allocation strategy is applied. In this example, the On-Demand allocation strategy is lowest-price.

In this example, there are 15 available unused Capacity Reservations. This is more than the fleet's target
On-Demand capacity of 12 On-Demand Instances.

585
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

The account has the following 15 unused Capacity Reservations in 3 different pools. The number of
Capacity Reservations in each pool is indicated by AvailableInstanceCount.

{
"CapacityReservationId": "cr-111",
"InstanceType": "m5.large",
"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount": 5,
"InstanceMatchCriteria": "open",
"State": "active"
}

{
"CapacityReservationId": "cr-222",
"InstanceType": "m4.xlarge",
"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount": 5,
"InstanceMatchCriteria": "open",
"State": "active"
}

{
"CapacityReservationId": "cr-333",
"InstanceType": "m4.2xlarge",
"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount":5,
"InstanceMatchCriteria": "open",
"State": "active"
}

The following fleet configuration shows only the pertinent configurations for this example. The On-
Demand allocation strategy is lowest-price, and the usage strategy for Capacity Reservations is use-
capacity-reservations-first. The total target capacity is 12, and the default target capacity type
is on-demand.

In this example, the On-Demand Instance price is:

• m5.large – $0.096 per hour

• m4.xlarge – $0.20 per hour
• m4.2xlarge – $0.40 per hour

Note
The fleet type must be instant. use-capacity-reservations-first is not supported for
other fleet types.

{
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-0e8c754449b27161c",
"Version": "1"
}
"Overrides": [
{
"InstanceType": "m5.large",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1

586
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

},
{
"InstanceType": "m4.xlarge",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1
},
{
"InstanceType": "m4.2xlarge",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1
}
]

}
],
"TargetCapacitySpecification": {
"TotalTargetCapacity": 12,
"DefaultTargetCapacityType": "on-demand"
},
"OnDemandOptions": {
"AllocationStrategy": "lowest-price"
"CapacityReservationOptions": {
"UsageStrategy": "use-capacity-reservations-first"
}
},
"Type": "instant",
}

After you create the instant fleet using the preceding configuration, the following 12 instances are
launched to meet the target capacity:

• 5 m5.large On-Demand Instances in us-east-1a – m5.large in us-east-1a is the lowest price, and there
are 5 available unused m5.large Capacity Reservations
• 5 m4.xlarge On-Demand Instances in us-east-1a – m4.xlarge in us-east-1a is the next lowest price, and
there are 5 available unused m4.xlarge Capacity Reservations
• 2 m4.2xlarge On-Demand Instances in us-east-1a – m4.2xlarge in us-east-1a is the third lowest price,
and there are 5 available unused m4.2xlarge Capacity Reservations of which only 2 are needed to meet
the target capacity

After the fleet is launched, you can run describe-capacity-reservations to see how many unused Capacity
Reservations are remaining. In this example, you should see the following response, which shows that all
of the m5.large and m4.xlarge Capacity Reservations were used, with 3 m4.2xlarge Capacity Reservations
remaining unused.

{
"CapacityReservationId": "cr-111",
"InstanceType": "m5.large",
"AvailableInstanceCount": 0
}

{
"CapacityReservationId": "cr-222",
"InstanceType": "m4.xlarge",
"AvailableInstanceCount": 0
}

{
"CapacityReservationId": "cr-333",
"InstanceType": "m4.2xlarge",
"AvailableInstanceCount": 3
}

587
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Example 8: Launch On-Demand Instances using Capacity Reservations and the lowest-price
allocation strategy when the total target capacity is more than the number of unused Capacity
Reservations

You can configure a fleet to use On-Demand Capacity Reservations first when launching On-Demand
Instances by setting the usage strategy for Capacity Reservations to use-capacity-reservations-
first. And if the number of unused Capacity Reservations is less than the On-Demand target capacity,
the remaining On-Demand target capacity is launched according to the chosen On-Demand allocation
strategy. In this example, the On-Demand allocation strategy is lowest-price.

In this example, there are 15 available unused Capacity Reservations. This is less than the fleet's On-
Demand target capacity of 16 On-Demand Instances.

The account has the following 15 unused Capacity Reservations in 3 different pools. The number of
Capacity Reservations in each pool is indicated by AvailableInstanceCount.

{
"CapacityReservationId": "cr-111",
"InstanceType": "m5.large",
"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount": 5,
"InstanceMatchCriteria": "open",
"State": "active"
}

{
"CapacityReservationId": "cr-222",
"InstanceType": "m4.xlarge",
"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount": 5,
"InstanceMatchCriteria": "open",
"State": "active"
}

{
"CapacityReservationId": "cr-333",
"InstanceType": "m4.2xlarge",
"InstancePlatform": "Linux/UNIX",
"AvailabilityZone": "us-east-1a",
"AvailableInstanceCount":5,
"InstanceMatchCriteria": "open",
"State": "active"
}

The following fleet configuration shows only the pertinent configurations for this example. The On-
Demand allocation strategy is lowest-price, and the usage strategy for Capacity Reservations is use-
capacity-reservations-first. The total target capacity is 16, and the default target capacity type
is on-demand.

In this example, the On-Demand Instance price is:

• m5.large – $0.096 per hour

• m4.xlarge – $0.20 per hour
• m4.2xlarge – $0.40 per hour

Note
The fleet type must be instant. use-capacity-reservations-first is not supported for
other fleet types.

588
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

{
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateId": "lt-0e8c754449b27161c",
"Version": "1"
}
"Overrides": [
{
"InstanceType": "m5.large",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1
},
{
"InstanceType": "m4.xlarge",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1
},
{
"InstanceType": "m4.2xlarge",
"AvailabilityZone": "us-east-1a",
"WeightedCapacity": 1
}
]

}
],
"TargetCapacitySpecification": {
"TotalTargetCapacity": 16,
"DefaultTargetCapacityType": "on-demand"
},
"OnDemandOptions": {
"AllocationStrategy": "lowest-price"
"CapacityReservationOptions": {
"UsageStrategy": "use-capacity-reservations-first"
}
},
"Type": "instant",
}

After you create the instant fleet using the preceding configuration, the following 16 instances are
launched to meet the target capacity:

• 6 m5.large On-Demand Instances in us-east-1a – m5.large in us-east-1a is the lowest price, and there
are 5 available unused m5.large Capacity Reservations. The Capacity Reservations are used first to
launch 5 On-Demand Instances plus an additional On-Demand Instance is launched according to the
On-Demand allocation strategy, which is lowest-price in this example.
• 5 m4.xlarge On-Demand Instances in us-east-1a – m4.xlarge in us-east-1a is the next lowest price, and
there are 5 available unused m4.xlarge Capacity Reservations
• 5 m4.2xlarge On-Demand Instances in us-east-1a – m4.2xlarge in us-east-1a is the third lowest price,
and there are 5 available unused m4.2xlarge Capacity Reservations

After the fleet is launched, you can run describe-capacity-reservations to see how many unused Capacity
Reservations are remaining. In this example, you should see the following response, which shows that all
of the Capacity Reservations in all of the pools were used.

{
"CapacityReservationId": "cr-111",
"InstanceType": "m5.large",
"AvailableInstanceCount": 0
}

589
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

{
"CapacityReservationId": "cr-222",
"InstanceType": "m4.xlarge",
"AvailableInstanceCount": 0
}

{
"CapacityReservationId": "cr-333",
"InstanceType": "m4.2xlarge",
"AvailableInstanceCount": 0
}

Example 9: Configure Capacity Rebalancing to launch replacement Spot Instances

The following example configures the EC2 Fleet to launch a replacement Spot Instance when Amazon
EC2 emits a rebalance recommendation for a Spot Instance in the fleet. To configure the automatic
replacement of Spot Instances, for ReplacementStrategy, specify launch.
Note
When a replacement instance is launched, the instance marked for rebalance is not
automatically terminated. You can terminate it, or you can leave it running. You are charged for
both instances while they are running.

The effectiveness of the Capacity Rebalancing strategy depends on the number of Spot Instance pools
specified in the EC2 Fleet request. We recommend that you configure the fleet with a diversified set of
instance types and Availability Zones, and for AllocationStrategy, specify capacity-optimized.
For more information about what you should consider when configuring an EC2 Fleet for Capacity
Rebalancing, see Capacity Rebalancing (p. 554).

{
"ExcessCapacityTerminationPolicy": "termination",
"LaunchTemplateConfigs": [
{
"LaunchTemplateSpecification": {
"LaunchTemplateName": "LaunchTemplate",
"Version": "1"
},
"Overrides": [
{
"InstanceType": "c3.large",
"WeightedCapacity": 1,
"Placement": {
"AvailabilityZone": "us-east-1a"
}
},
{
"InstanceType": "c4.large",
"WeightedCapacity": 1,
"Placement": {
"AvailabilityZone": "us-east-1a"
}
},
{
"InstanceType": "c5.large",
"WeightedCapacity": 1,
"Placement": {
"AvailabilityZone": "us-east-1a"
}
}
]
}
],

590
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

"TargetCapacitySpecification": {
"TotalTargetCapacity": 5,
"DefaultTargetCapacityType": "spot"
},
"SpotOptions": {
"AllocationStrategy": "capacity-optimized",
"MaintenanceStrategies": {
"CapacityRebalance": {
"ReplacementStrategy": "launch"
}
}
}
}

Use Amazon EventBridge to monitor EC2 Fleet events

When the state of an EC2 Fleet changes, the EC2 Fleet emits a notification. The notification is made
available as an event that is sent to Amazon EventBridge (formerly known as Amazon CloudWatch
Events). Events are emitted on a best effort basis.

With Amazon EventBridge, you can create rules that trigger programmatic actions in response to
an event. For example, you can create two EventBridge rules, one that's triggered when a fleet state
changes, and one that's triggered when an instance in the fleet is terminated. If the fleet state changes,
the first EventBridge rule invokes an SNS topic to send an email notification to you. If an instance is
terminated, the second EventBridge rule invokes a Lambda function to launch a new instance.
Note
Only fleets of type maintain and request emit events. Fleets of type instant do not emit
events because they submit synchronous one-time requests, and the state of the fleet is known
immediately in the response.

Topics
• EC2 Fleet event types (p. 591)
• Use Amazon EventBridge (p. 596)

EC2 Fleet event types

There are five EC2 Fleet event types. For each event type, there are several sub-types.

The events are sent to EventBridge in JSON format. The following fields in the event form the event
pattern that is defined in the rule, and which trigger an action:

"source": "aws.ec2fleet"

Identifies that the event is from EC2 Fleet.

"detail-type": "EC2 Fleet State Change"

Identifies the event type.

"sub-type": "submitted"

Identifies the event sub-type.

Event types
• EC2 Fleet State Change (p. 592)
• EC2 Fleet Spot Instance Request Change (p. 593)
• EC2 Fleet Instance Change (p. 593)

591
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

• EC2 Fleet Information (p. 594)

• EC2 Fleet Error (p. 595)

EC2 Fleet State Change

EC2 Fleet sends an EC2 Fleet State Change event to Amazon EventBridge when an EC2 Fleet
changes state.

The following is example data for this event.

{
"version": "0",
"id": "715ed6b3-b8fc-27fe-fad6-528c7b8bf8a2",
"detail-type": "EC2 Fleet State Change",
"source": "aws.ec2fleet",
"account": "123456789012",
"time": "2020-11-09T09:00:20Z",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:123456789012:fleet/fleet-598fb973-87b7-422d-
be4d-6b0809bfff0a"
],
"detail": {
"sub-type": "active"
}
}

The possible values for sub-type are:

submitted

The EC2 Fleet request is being evaluated and Amazon EC2 is preparing to launch the target number
of instances.
active

The EC2 Fleet request has been validated and Amazon EC2 is attempting to maintain the target
number of running Spot Instances.
progress

The EC2 Fleet request is in the process of being fulfilled.

cancelled_terminating

The EC2 Fleet request is deleted and its instances are terminating. The request remains in this state
until all instances are terminated.
cancelled_running

The EC2 Fleet request is deleted and does not launch additional instances. Its existing instances
continue to run until they are interrupted or terminated. The request remains in this state until all
instances are interrupted or terminated.
cancelled

The EC2 Fleet request is deleted and has no running instances. The EC2 Fleet will be deleted two
days after its instances are terminated.
modify_in_progress

The EC2 Fleet request is being modified. The request remains in this state until the modification is
fully processed or the EC2 Fleet request is deleted.

592
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

modify_succeeded

The EC2 Fleet request was modified. This state does not apply to instant fleets because instant
fleets cannot be modified.
expired

The EC2 Fleet request has expired. If the request was created with
TerminateInstancesWithExpiration set, a subsequent event indicates that the instances are
terminated.

EC2 Fleet Spot Instance Request Change

EC2 Fleet sends an EC2 Fleet Spot Instance Request Change event to Amazon EventBridge
when a Spot Instance request in the fleet changes state.

The following is example data for this event.

{
"version": "0",
"id": "19331f74-bf4b-a3dd-0f1b-ddb1422032b9",
"detail-type": "EC2 Fleet Spot Instance Request Change",
"source": "aws.ec2fleet",
"account": "123456789012",
"time": "2020-11-09T09:00:05Z",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:123456789012:fleet/
fleet-83fd4e48-552a-40ef-9532-82a3acca5f10"
],
"detail": {
"spot-instance-request-id": "sir-rmqske6h",
"description": "SpotInstanceRequestId sir-rmqske6h, PreviousState:
cancelled_running",
"sub-type": "cancelled"
}
}

The possible values for sub-type are:

submitted

The request is submitted.

disabled

You stopped the Spot Instance.

active

The request is fulfilled and has an associated Spot Instance.

cancelled

You cancelled the request, or the request expired.

EC2 Fleet Instance Change

EC2 Fleet sends an EC2 Fleet Instance Change event to Amazon EventBridge when an instance in
the fleet changes state.

593
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

The following is example data for this event.

{
"version": "0",
"id": "542ce428-c8f1-0608-c015-e8ed6522c5bc",
"detail-type": "EC2 Fleet Instance Change",
"source": "aws.ec2fleet",
"account": "123456789012",
"time": "2020-11-09T09:00:23Z",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:123456789012:fleet/fleet-598fb973-87b7-422d-
be4d-6b0809bfff0a"
],
"detail": {
"instance-id": "i-0c594155dd5ff1829",
"description": "{\"instanceType\":\"c5.large\",\"image\":\"ami-6057e21a\",
\"productDescription\":\"Linux/UNIX\",\"availabilityZone\":\"us-east-1d\"}",
"sub-type": "launched"
}
}

The possible values for sub-type are:

launched

A new instance was launched.

terminated

The instance was terminated.

termination_notified

An instance termination notification was sent.

EC2 Fleet Information

EC2 Fleet sends an EC2 Fleet Information event to Amazon EventBridge when there is an error
during fulfillment. The information event does not block the fleet from attempting to fulfil its target
capacity.

The following is example data for this event.

{
"version": "0",
"id": "76529817-d605-4571-7224-d36cc1b2c0c4",
"detail-type": "EC2 Fleet Information",
"source": "aws.ec2fleet",
"account": "123456789012",
"time": "2020-11-09T08:17:07Z",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:123456789012:fleet/fleet-8becf5fe-
bb9e-415d-8f54-3fa5a8628b91"
],
"detail": {
"description": "r3.8xlarge, ami-032930428bf1abbff, Linux/UNIX, us-east-1a, Spot bid
price is less than Spot market price $0.5291",
"sub-type": "launchSpecUnusable"
}
}

594
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

The possible values for sub-type are:

launchSpecUnusable

The price in a launch specification is not valid because it is below the Spot price or the Spot price is
above the On-Demand price.
fleetProgressHalted

The price in every launch specification is not valid. A launch specification might become valid if the
Spot price changes.
registerWithLoadBalancersFailed

An attempt to register instances with load balancers failed. For more information, see the
description of the event.
launchSpecTemporarilyBlacklisted

The configuration is not valid and several attempts to launch instances have failed. For more
information, see the description of the event.

EC2 Fleet Error

EC2 Fleet sends an EC2 Fleet Error event to Amazon EventBridge when there is an error during
fulfillment. The error event blocks the fleet from attempting to fulfil its target capacity.

The following is example data for this event.

{
"version": "0",
"id": "69849a22-6d0f-d4ce-602b-b47c1c98240e",
"detail-type": "EC2 Fleet Error",
"source": "aws.ec2fleet",
"account": "123456789012",
"time": "2020-10-07T01:44:24Z",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:123456789012:fleet/fleet-9bb19bc6-60d3-4fd2-ae47-
d33e68eafa08"
],
"detail": {
"description": "m3.large, ami-00068cd7555f543d5, Linux/UNIX: IPv6 is not supported
for the instance type 'm3.large'. ",
"sub-type": "spotFleetRequestConfigurationInvalid"
}
}

The possible values for sub-type are:

allLaunchSpecsTemporarilyBlacklisted

None of the configurations are valid, and several attempts to launch instances have failed. For more
information, see the description of the event.
spotFleetRequestConfigurationInvalid

The configuration is not valid. For more information, see the description of the event.
spotInstanceCountLimitExceeded

You’ve reached the limit on the number of Spot Instances that you can launch.

595
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

Use Amazon EventBridge

When a notification of a state change is emitted for an EC2 Fleet, the event for the notification is sent to
Amazon EventBridge. If EventBridge detects an event pattern that matches a pattern defined in a rule,
EventBridge invokes a target (or targets) specified in the rule.

The preceding sections provide example events for the EC2 Fleet notifications. The event JSON files
contain the fields that form the event pattern.

The following fields form the event pattern that is defined in the rule:

"source": "aws.ec2fleet"

Identifies that the event is from EC2 Fleet.

"detail-type": "EC2 Fleet State Change"

Identifies the event type.

"sub-type": "submitted"

Identifies the event sub-type.

You can write an EventBridge rule and automate what actions to take when the event pattern matches
the rule.

Examples
• Create an EventBridge rule to send a notification (p. 436)
• Create an EventBridge rule to trigger a Lambda function (p. 436)

Create an EventBridge rule to send a notification

The following example creates an EventBridge rule to send an email, text message, or mobile push
notification every time that Amazon EC2 emits an EC2 Fleet notification. The signal in this example is
emitted as an EC2 Fleet State Change event, which triggers the action defined by the rule.

To create an EventBridge rule to send a notification when an EC2 Fleet state changes

1. Open the Amazon EventBridge console at https://console.aws.amazon.com/events/.

2. Choose Create rule.
3. Enter a Name for the rule, and, optionally, a description.

A rule can't have the same name as another rule in the same Region and on the same event bus.
4. For Define pattern, choose Event pattern.
5. Under Event matching pattern, choose Custom pattern.
6. In the Event pattern box, add the following pattern to match the EC2 Fleet State Change
event, and then choose Save.

{
"source": [ "aws.ec2fleet" ],
"detail-type": [ "EC2 Fleet State Change" ],
"sub-type": [ "submitted" ]
}

7. For Select event bus, choose AWS default event bus. When an AWS service in your account emits an
event, it always goes to your account's default event bus.
8. Confirm that Enable the rule on the selected event bus is toggled on.

596
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch

9. For Target, choose SNS topic to send an email, text message, or mobile push notification when the
event occurs.
10. For Topic, choose an existing topic. You first need to create an Amazon SNS topic using the Amazon
SNS console. For more information, see Using Amazon SNS for application-to-person (A2P)
messaging in the Amazon Simple Notification Service Developer Guide.
11. For Configure input, choose the input for the email, text message, or mobile push notification.
12. Choose Create.

For more information, see Creating a rule for an AWS service and Event Patterns in the Amazon
EventBridge User Guide

Create an EventBridge rule to trigger a Lambda function

The following example creates an EventBridge rule to trigger a Lambda function every time that Amazon
EC2 emits an EC2 Fleet notification. The signal in this example is emitted as an EC2 Fleet Instance
Change event, which triggers the action defined by the rule. Before creating the EventBridge rule, you
must create the Lambda function.

To create an EventBridge rule to trigger a Lambda function when an instance in an EC2 Fleet
changes state

1. Open the AWS Lambda console at https://console.aws.amazon.com/lambda/.

2. Choose Create function.
3. Enter a name for your function, configure the code, and then choose Create function.

For more information about using Lambda, see Create a Lambda function with the console in the
AWS Lambda Developer Guide.
4. Open the Amazon EventBridge console at https://console.aws.amazon.com/events/.
5. Choose Create rule.
6. Enter a Name for the rule, and, optionally, a description.

A rule can't have the same name as another rule in the same Region and on the same event bus.
7. For Define pattern, choose Event pattern.
8. Under Event matching pattern, you can choose Pre-defined pattern by service or Custom pattern.
The Custom pattern allows you to create a more detailed rule.

a. If you choose Pre-defined pattern by service, do the following:

i. For Service provider, choose AWS.

ii. For Service name, choose EC2 Fleet.
iii. For Event type, select the required event type. For this example, choose EC2 Fleet Instance
Change.
b. If you choose Custom pattern, do the following:

• In the Event pattern box, add the following pattern to match the EC2 Fleet Instance
Change event and launched sub-type for this example, and then choose Save.

{
"source": [ "aws.ec2fleet" ],
"detail-type": [ "EC2 Fleet Instance Change" ],
"sub-type": [ "launched" ]
}

9. For Target, choose Lambda function, and for Function, choose the function that you created to
respond when the event occurs.

597
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

10. Choose Create.

In this example, the Lambda function will be triggered when the EC2 Fleet Instance Change
event with the sub-type launched occurs.

For a tutorial on how to create a Lambda function and an EventBridge rule that runs the Lambda
function, see Tutorial: Log the State of an Amazon EC2 Instance Using EventBridge in the AWS Lambda
Developer Guide.

Connect to your Linux instance

Connect to the Linux instances that you launched and transfer files between your local computer and
your instance.

To connect to a Windows instance, see Connecting to Your Windows Instance in the Amazon EC2 User
Guide for Windows Instances.

Connection options
The operating system of your local computer determines the options that you have to connect from your
local computer to your Linux instance.

If your local computer operating system is Linux or macOS X

• SSH client (p. 601)
• EC2 Instance Connect (p. 604)
• AWS Systems Manager Session Manager
If your local computer operating system is Windows
• PuTTY (p. 614)
• SSH client (p. 601)
• AWS Systems Manager Session Manager
• Windows Subsystem for Linux (p. 619)

General prerequisites for connecting to your instance

Before you connect to your Linux instance, verify the following general prerequisites:
• Get information about your instance (p. 598)
• Enable inbound traffic to your instance (p. 600)
• Locate the private key (p. 600)
• (Optional) Get the instance fingerprint (p. 600)

Get information about your instance

• Get the ID of the instance.

You can get the ID of your instance using the Amazon EC2 console (from the Instance ID column). If
you prefer, you can use the describe-instances (AWS CLI) or Get-EC2Instance (AWS Tools for Windows
PowerShell) command.
• Get the public DNS name of the instance.

You can get the public DNS for your instance using the Amazon EC2 console. Check
the Public DNS (IPv4) column. If this column is hidden, choose the settings icon (

598
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

) in the top-right
corner of the screen and select Public DNS (IPv4). If you prefer, you can use the describe-instances
(AWS CLI) or Get-EC2Instance (AWS Tools for Windows PowerShell) command.
• (IPv6 only) Get the IPv6 address of the instance.

If you've assigned an IPv6 address to your instance, you can optionally connect to the instance using
its IPv6 address instead of a public IPv4 address or public IPv4 DNS hostname. Your local computer
must have an IPv6 address and must be configured to use IPv6. You can get the IPv6 address of
your instance using the Amazon EC2 console. Check the IPv6 IPs field. If you prefer, you can use the
describe-instances (AWS CLI) or Get-EC2Instance (AWS Tools for Windows PowerShell) command. For
more information about IPv6, see IPv6 addresses (p. 811).
• Get the user name for your instance.

You can connect to your instance using the user name for your user account or the default user name
for the AMI that you used to launch your instance.
• Get the user name for your user account.

For more information about how to create a user account, see Manage user accounts on your
Amazon Linux instance (p. 656).
• Get the default user name for the AMI that you used to launch your instance:
• For Amazon Linux 2 or the Amazon Linux AMI, the user name is ec2-user.
• For a CentOS AMI, the user name is centos.
• For a Debian AMI, the user name is admin.
• For a Fedora AMI, the user name is ec2-user or fedora.
• For a RHEL AMI, the user name is ec2-user or root.

599
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

• For a SUSE AMI, the user name is ec2-user or root.

• For an Ubuntu AMI, the user name is ubuntu.
• Otherwise, if ec2-user and root don't work, check with the AMI provider.

Enable inbound traffic to your instance

• Enable inbound SSH traffic from your IP address to your instance.

Ensure that the security group associated with your instance allows incoming SSH traffic from your
IP address. The default security group for the VPC does not allow incoming SSH traffic by default.
The security group created by the launch instance wizard enables SSH traffic by default. For more
information, see Authorize inbound traffic for your Linux instances (p. 1041).

Locate the private key

• Locate the private key

Get the fully-qualified path to the location on your computer of the .pem file for the key pair that you
specified when you launched the instance. For more information about how you created your key pair,
see Creating a Key Pair Using Amazon EC2.
• Set the permissions of your private key

If you will use an SSH client on a macOS or Linux computer to connect to your Linux instance, use the
following command to set the permissions of your private key file so that only you can read it.

chmod 400 my-key-pair.pem

If you do not set these permissions, then you cannot connect to your instance using this key pair. For
more information, see Error: Unprotected private key file (p. 1336).

(Optional) Get the instance fingerprint

To protect yourself from man-in-the-middle attacks, you can verify the RSA key fingerprint when you
connect to your instance. Verifying the fingerprint is useful if you've launched your instance from a
public AMI from a third party.

First you get the instance fingerprint. Then, when you connect to the instance, you are prompted to
verify the fingerprint. You can compare the fingerprint you obtained with the fingerprint displayed for
verification. If these fingerprints don't match, someone might be attempting a "man-in-the-middle"
attack. If they match, you can confidently connect to your instance.

Prerequisites for getting the instance fingerprint:

• To get the instance fingerprint, you must use the AWS CLI. For information about installing the AWS
CLI, see Installing the AWS Command Line Interface in the AWS Command Line Interface User Guide.
• The instance must not be in the pending state. The fingerprint is available only after the first boot of
the instance is complete.

To get the instance fingerprint

1. On your local computer (not on the instance), use the get-console-output (AWS CLI) command as
follows to obtain the fingerprint:

600
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

aws ec2 get-console-output --instance-id instance_id --output text

2. Here is an example of what you should look for in the output. The exact output can vary by the
operating system, AMI version, and whether you had AWS create the key.

ec2: #############################################################
ec2: -----BEGIN SSH HOST KEY FINGERPRINTS-----
ec2: 1024 SHA256:7HItIgTONZ/b0CH9c5Dq1ijgqQ6kFn86uQhQ5E/F9pU root@ip-10-0-2-182 (DSA)
ec2: 256 SHA256:l4UB/neBad9tvkgJf1QZWxheQmR59WgrgzEimCG6kZY root@ip-10-0-2-182 (ECDSA)
ec2: 256 SHA256:kpEa+rw/Uq3zxaYZN8KT501iBtJOIdHG52dFi66EEfQ no comment (ED25519)
ec2: 2048 SHA256:L8l6pepcA7iqW/jBecQjVZClUrKY+o2cHLI0iHerbVc root@ip-10-0-2-182 (RSA)
ec2: -----END SSH HOST KEY FINGERPRINTS-----
ec2: #############################################################

Connect to your Linux instance using SSH

After you launch your instance, you can connect to it and use it the way that you'd use a computer sitting
in front of you.

The following instructions explain how to connect to your instance using an SSH client. If you
receive an error while attempting to connect to your instance, see Troubleshoot connecting to your
instance (p. 1329). For more connection options, see Connect to your Linux instance (p. 598).

Prerequisites
Before you connect to your Linux instance, complete the following prerequisites.

Check your instance status

After you launch an instance, it can take a few minutes for the instance to be ready so that you can
connect to it. Check that your instance has passed its status checks. You can view this information in
the Status check column on the Instances page.
Get the public DNS name and user name to connect to your instance

To find the public DNS name or IP address of your instance and the user name that you should use
to connect to your instance, see Prerequisites for connecting to your instance (p. 598).
Install an SSH client on your local computer as needed

Your local computer might have an SSH client installed by default. You can verify this by typing
ssh at the command line. If your computer doesn't recognize the command, you can install an SSH
client.
• Recent versions of Windows Server 2019 and Windows 10 - OpenSSH is included as an installable
component. For more information, see OpenSSH in Windows.
• Earlier versions of Windows - Download and install OpenSSH. For more information, see Win32-
OpenSSH.
• Linux and macOS X - Download and install OpenSSH. For more information, see http://
www.openssh.com.

Connect to your Linux instance using an SSH client

Use the following procedure to connect to your Linux instance using an SSH client. If you receive an error
while attempting to connect to your instance, see Troubleshoot connecting to your instance (p. 1329).

601
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

To connect to your instance using SSH

1. In a terminal window, use the ssh command to connect to the instance. You specify the path and
file name of the private key (.pem), the user name for your instance, and the public DNS name or
IPv6 address for your instance. For more information about how to find the private key, the user
name for your instance, and the DNS name or IPv6 address for an instance, see Locate the private
key (p. 600) and Get information about your instance (p. 598). To connect to your instance, use
one of the following commands.
• (Public DNS) To connect using your instance's public DNS name, enter the following command.

ssh -i /path/my-key-pair.pem my-instance-user-name@my-instance-public-dns-name

• (IPv6) Alternatively, if your instance has an IPv6 address, to connect using your instance's IPv6
address, enter the following command.

ssh -i /path/my-key-pair.pem my-instance-user-name@my-instance-IPv6-address

You see a response like the following:

The authenticity of host 'ec2-198-51-100-1.compute-1.amazonaws.com (198-51-100-1)'

can't be established.
ECDSA key fingerprint is l4UB/neBad9tvkgJf1QZWxheQmR59WgrgzEimCG6kZY.
Are you sure you want to continue connecting (yes/no)?

2. (Optional) Verify that the fingerprint in the security alert matches the fingerprint that you previously
obtained in (Optional) Get the instance fingerprint (p. 600). If these fingerprints don't match,
someone might be attempting a "man-in-the-middle" attack. If they match, continue to the next
step.
3. Enter yes.

You see a response like the following:

Warning: Permanently added 'ec2-198-51-100-1.compute-1.amazonaws.com' (ECDSA) to the

list of known hosts.

Transfer files to Linux instances from Linux using SCP

One way to transfer files between your local computer and a Linux instance is to use the secure copy
protocol (SCP). This section describes how to transfer files with SCP. The procedure is similar to the
procedure for connecting to an instance with SSH.

Prerequisites

• Verify the general prerequisites for transferring files to your instance.

The general prerequisites for transferring files to an instance are the same as the general prerequisites
for connecting to an instance. For more information, see General prerequisites for connecting to your
instance (p. 598).
• Install an SCP client

Most Linux, Unix, and Apple computers include an SCP client by default. If yours doesn't, the OpenSSH
project provides a free implementation of the full suite of SSH tools, including an SCP client. For more
information, see http://www.openssh.org.

602
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

The following procedure steps you through using SCP to transfer a file. If you've already connected to
the instance with SSH and have verified its fingerprints, you can start with the step that contains the SCP
command (step 4).

To use SCP to transfer a file

1. Transfer a file to your instance using the instance's public DNS name, or the IPv6 address if your
instance has one. For example, if the name of your private key file is my-key-pair, the file to
transfer is SampleFile.txt, the user name for your instance is my-instance-user-name, and
the public DNS name of the instance is my-instance-public-dns-name, or my-instance-
IPv6-address if your instance has an IPv6 address, use one of the following commands to copy
the file to the my-instance-user-name home directory.
• (Public DNS) To transfer a file to your instance using your instance's public DNS name, enter the
following command.

scp -i /path/my-key-pair.pem /path/SampleFile.txt my-instance-user-name@my-

instance-public-dns-name:~

• (IPv6) Alternatively, if your instance has an IPv6 address, to transfer a file using the instance's
IPv6 address, enter the following command. The IPv6 address must be enclosed in square
brackets ([ ]), which must be escaped (\).

scp -i /path/my-key-pair.pem /path/SampleFile.txt my-instance-user-name@\[my-

instance-IPv6-address\]:~

You see a response like the following:

The authenticity of host 'ec2-198-51-100-1.compute-1.amazonaws.com (10.254.142.33)'

can't be established.
RSA key fingerprint is 1f:51:ae:28:bf:89:e9:d8:1f:25:5d:37:2d:7d:b8:ca:9f:f5:f1:6f.
Are you sure you want to continue connecting (yes/no)?

2. (Optional) Verify that the fingerprint in the security alert matches the fingerprint that you previously
obtained in (Optional) Get the instance fingerprint (p. 600). If these fingerprints don't match,
someone might be attempting a "man-in-the-middle" attack. If they match, continue to the next
step.
3. Enter yes.

You see a response like the following:

Warning: Permanently added 'ec2-198-51-100-1.compute-1.amazonaws.com' (RSA)

to the list of known hosts.
Sending file modes: C0644 20 SampleFile.txt
Sink: C0644 20 SampleFile.txt
SampleFile.txt 100% 20 0.0KB/s 00:00

If you receive a "bash: scp: command not found" error, you must first install scp on your Linux
instance. For some operating systems, this is located in the openssh-clients package. For
Amazon Linux variants, such as the Amazon ECS-optimized AMI, use the following command to
install scp:

[ec2-user ~]$ sudo yum install -y openssh-clients

4. To transfer files in the other direction (from your Amazon EC2 instance to your local computer),
reverse the order of the host parameters. For example, to transfer the SampleFile.txt file from
your EC2 instance back to the home directory on your local computer as SampleFile2.txt, use of
the following commands on your local computer.

603
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

• (Public DNS) To transfer a file from your instance using your instance's public DNS name, enter
the following command.

scp -i /path/my-key-pair.pem my-instance-user-name@my-instance-public-dns-name:~/

SampleFile.txt ~/SampleFile2.txt

• (IPv6) Alternatively, if your instance has an IPv6 address, to transfer a file using the instance's
IPv6 address, enter the following command. The IPv6 address must be enclosed in square
brackets ([ ]), which must be escaped (\).

scp -i /path/my-key-pair.pem my-instance-user-name@\[my-instance-IPv6-address\]:~/

SampleFile.txt ~/SampleFile2.txt

Connect to your Linux instance using EC2 Instance Connect

Amazon EC2 Instance Connect provides a simple and secure way to connect to your Linux instances using
Secure Shell (SSH). With EC2 Instance Connect, you use AWS Identity and Access Management (IAM)
policies and principals to control SSH access to your instances, removing the need to share and manage
SSH keys. All connection requests using EC2 Instance Connect are logged to AWS CloudTrail so that you
can audit connection requests (p. 807).

You can use EC2 Instance Connect to connect to your instances using the Amazon EC2 console (browser-
based client), the Amazon EC2 Instance Connect CLI, or the SSH client of your choice.

When you connect to an instance using EC2 Instance Connect, the Instance Connect API pushes a
one-time-use SSH public key to the instance metadata (p. 704) where it remains for 60 seconds. An
IAM policy attached to your IAM user authorizes your IAM user to push the public key to the instance
metadata. The SSH daemon uses AuthorizedKeysCommand and AuthorizedKeysCommandUser,
which are configured when Instance Connect is installed, to look up the public key from the instance
metadata for authentication, and connects you to the instance.

You can use EC2 Instance Connect to connect to instances that have public or private IP addresses. For
more information, see Connect using EC2 Instance Connect (p. 610).

Tip

If you are connecting to a Linux instance from a local computer running Windows, see the following
documentation instead:

• Connect to your Linux instance from Windows using PuTTY (p. 614)
• Connect to your Linux instance using SSH (p. 601)
• Connect to your Linux instance from Windows using Windows Subsystem for Linux (p. 619)

Contents
• Set up EC2 Instance Connect (p. 604)
• Connect using EC2 Instance Connect (p. 610)
• Uninstall EC2 Instance Connect (p. 613)

Set up EC2 Instance Connect

To use EC2 Instance Connect to connect to an instance, you need to configure every instance that will
support using Instance Connect (this is a one-time requirement for each instance), and you need to grant
permission to every IAM principal that will use Instance Connect.

604
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

Tasks to set up Instance Connect

• Task 1: Configure network access to an instance (p. 605)
• Task 2: (Conditional) Install EC2 Instance Connect on an instance (p. 606)
• Task 3: (Optional) Install the EC2 Instance Connect CLI (p. 608)
• Task 4: Configure IAM permissions for EC2 Instance Connect (p. 608)

For more information about setting up EC2 Instance Connect, see Securing your bastion hosts with
Amazon EC2 Instance Connect.

Limitations

• The following Linux distributions are supported:

• Amazon Linux 2 (any version)
• Ubuntu 16.04 or later
• If you configured the AuthorizedKeysCommand and AuthorizedKeysCommandUser settings
for SSH authentication, the EC2 Instance Connect installation will not update them. As a result, you
cannot use Instance Connect.

Prerequisites for installing EC2 Instance Connect

• Verify the general prerequisites for connecting to your instance using SSH.

For more information, see General prerequisites for connecting to your instance (p. 598).
• Install an SSH client on your local computer.

Your local computer most likely has an SSH client installed by default. You can check for an SSH client
by typing ssh at the command line. If your local computer doesn't recognize the command, you can
install an SSH client. For information about installing an SSH client on Linux or macOS X, see http://
www.openssh.com. For information about installing an SSH client on Windows 10, see OpenSSH in
Windows.
• Install the AWS CLI on your local computer.

To configure the IAM permissions, you must use the AWS CLI. For more information about installing
the AWS CLI, see Installing the AWS CLI in the AWS Command Line Interface User Guide.
• [Ubuntu] Install the AWS CLI on your instance.

To install EC2 Instance Connect on an Ubuntu instance, you must use the AWS CLI on the instance. For
more information about installing the AWS CLI, see Installing the AWS CLI in the AWS Command Line
Interface User Guide.

Task 1: Configure network access to an instance

You must configure the following network access so that your users can connect to your instance using
EC2 Instance Connect:

• If your users will access your instance over the internet, then your instance must have a public IP
address. For more information, see Public IPv4 addresses and external DNS hostnames (p. 810).
• If your users will access your instance via the instance's private IP address, then you must establish
private network connectivity to your VPC, such as by using AWS Direct Connect, AWS Site-to-Site VPN,
or VPC peering, so that your users can reach the instance's private IP address.
• Ensure that the security group associated with your instance allows inbound SSH traffic (p. 1041)
on port 22 from your IP address or from your network. The default security group for the VPC does

605
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

not allow incoming SSH traffic by default. The security group created by the launch wizard allows
incoming SSH traffic by default. For more information, see Authorize inbound traffic for your Linux
instances (p. 1041).
• (Amazon EC2 console browser-based client) We recommend that your instance allows inbound SSH
traffic from the recommended IP block published for the service. Use the EC2_INSTANCE_CONNECT
filter for the service parameter to get the IP address ranges in the EC2 Instance Connect subset. For
more information, see AWS IP address ranges in the Amazon Web Services General Reference.

Task 2: (Conditional) Install EC2 Instance Connect on an instance

Amazon Linux 2 2.0.20190618 or later and Ubuntu 20.04 or later are preconfigured with EC2 Instance
Connect. If you launched your instance using one of these AMIs, you can skip this task. For other
supported Linux distributions, you must install Instance Connect on every instance that will support
connecting using Instance Connect.

Installing Instance Connect configures the SSH daemon on the instance. The procedure for installing
Instance Connect is different for instances launched using Amazon Linux 2 and Ubuntu.

Amazon Linux 2

To install EC2 Instance Connect on an instance launched with Amazon Linux 2

1. Connect to your instance using SSH.

Use the SSH key pair that was assigned to your instance when you launched it and the default
user name of the AMI that you used to launch your instance. For Amazon Linux 2, the default
user name is ec2-user.

For example, if your instance was launched using Amazon Linux 2, your instance's public
DNS name is ec2-a-b-c-d.us-west-2.compute.amazonaws.com, and the key pair is
my_ec2_private_key.pem, use the following command to SSH into your instance:

$ ssh -i my_ec2_private_key.pem [email protected]

west-2.compute.amazonaws.com

For more information about connecting to your instance, see Connect to your Linux instance
using SSH (p. 601).
2. Install the EC2 Instance Connect package on your instance.

For Amazon Linux 2, use the yum install command.

[ec2-user ~]$ sudo yum install ec2-instance-connect

You should see four new scripts in the /opt/aws/bin/ folder:

eic_curl_authorized_keys
eic_harvest_hostkeys
eic_parse_authorized_keys
eic_run_authorized_keys

3. (Optional) Verify that Instance Connect was successfully installed on your instance.

Use the sudo less command to check that the /etc/ssh/sshd_config file was correctly
updated as follows:

[ec2-user ~]$ sudo less /etc/ssh/sshd_config

606
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

Instance Connect was successfully installed if the AuthorizedKeysCommand and

AuthorizedKeysCommandUser lines in the /etc/ssh/sshd_config file contain the
following values:

AuthorizedKeysCommand /opt/aws/bin/eic_run_authorized_keys %u %f
AuthorizedKeysCommandUser ec2-instance-connect

• AuthorizedKeysCommand sets the eic_run_authorized_keys script to look up the keys

from the instance metadata
• AuthorizedKeysCommandUser sets the system user as ec2-instance-connect

Note
If you previously configured AuthorizedKeysCommand and
AuthorizedKeysCommandUser, the Instance Connect installation will not change the
values and you will not be able to use Instance Connect.

Ubuntu

To install EC2 Instance Connect on an instance launched with Ubuntu 16.04 or later

1. Connect to your instance using SSH.

Use the SSH key pair that was assigned to your instance when you launched it and use the
default user name of the AMI that you used to launch your instance. For an Ubuntu AMI, the
user name is ubuntu.

If your instance was launched using Ubuntu, your instance's public DNS name is ec2-a-b-c-
d.us-west-2.compute.amazonaws.com, and the key pair is my_ec2_private_key.pem,
use the following command to SSH into your instance:

$ ssh -i my_ec2_private_key.pem [email protected]

For more information about connecting to your instance, see Connect to your Linux instance
using SSH (p. 601).
2. (Optional) Ensure your instance has the latest Ubuntu AMI.

For Ubuntu, use the following commands to update all the packages on your instance.

ubuntu:~$ sudo apt-get update

ubuntu:~$ sudo apt-get upgrade

3. Install the Instance Connect package on your instance.

For Ubuntu, use the sudo apt-get command to install the .deb package.

ubuntu:~$ sudo apt-get install ec2-instance-connect

You should see four new scripts in the /usr/share/ec2-instance-connect/ folder:

eic_curl_authorized_keys
eic_harvest_hostkeys
eic_parse_authorized_keys

607
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

eic_run_authorized_keys

4. (Optional) Verify that Instance Connect was successfully installed on your instance.

Use the sudo less command to check that the /lib/systemd/system/ssh.service.d/

ec2-instance-connect.conf was correctly updated as follows:

ubuntu:~$ sudo less /lib/systemd/system/ssh.service.d/ec2-instance-connect.conf

Instance Connect was successfully installed if the AuthorizedKeysCommand and

AuthorizedKeysCommandUser lines in the /lib/systemd/system/ssh.service.d/ec2-
instance-connect.conf file contain the following values:

AuthorizedKeysCommand /usr/share/ec2-instance-connect/eic_run_authorized_keys %u %f
AuthorizedKeysCommandUser ec2-instance-connect

• AuthorizedKeysCommand sets the eic_run_authorized_keys script to look up the keys

from the instance metadata
• AuthorizedKeysCommandUser sets the system user as ec2-instance-connect

Note
If you previously configured AuthorizedKeysCommand and
AuthorizedKeysCommandUser, the Instance Connect installation will not change the
values and you will not be able to use Instance Connect.

For more information about the EC2 Instance Connect package, see aws/aws-ec2-instance-connect-
config on the GitHub website.

Task 3: (Optional) Install the EC2 Instance Connect CLI

The EC2 Instance Connect CLI provides a simplified experience to connect to EC2 instances through
a single command, mssh instance_id. For more information, see Connect using the EC2 Instance
Connect CLI (p. 611).
Note
There is no need to install the EC2 Instance Connect CLI if users will only use the Amazon EC2
console (browser-based client) or an SSH client to connect to an instance.

To install the EC2 Instance Connect CLI package

Use pip to install the ec2instanceconnectcli package. For more information, see aws/aws-ec2-
instance-connect-cli on the GitHub website, and https://pypi.org/project/ec2instanceconnectcli/ on the
Python Package Index (PyPI) website.

$ pip install ec2instanceconnectcli

Task 4: Configure IAM permissions for EC2 Instance Connect

For your IAM principals to connect to an instance using EC2 Instance Connect, you must grant them
permission to push the public key to the instance. You grant them the permission by creating an IAM
policy and attaching the policy to the IAM principals that require the permission. For more information,
see Actions, resources, and condition keys for Amazon EC2 Instance Connect.

The following instructions explain how to create the policy and attach it to an IAM user using the AWS
CLI. The same policy could be applied to other IAM principals, such as IAM roles. For instructions that

608
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

use the AWS Management Console, see Creating IAM policies (console), Adding permissions by attaching
policies directly to the user, and Creating IAM roles in the IAM User Guide.

To grant an IAM principal permission for EC2 Instance Connect (AWS CLI)

1. Create a JSON policy document that includes the following:

• The ec2-instance-connect:SendSSHPublicKey action. This grants an IAM

principal permission to push the public key to an instance. With ec2-instance-
connect:SendSSHPublicKey, consider restricting access to specific EC2 instances. Otherwise,
all IAM principals with this permission can connect to all EC2 instances. You can also restrict access
by specifying resource ARNs or by using resource tags as condition keys.
• The ec2:osuser condition. This specifies the name of the OS user that can push the public key
to an instance. Use the default user name for the AMI that you used to launch the instance. The
default user name for Amazon Linux 2 is ec2-user, and for Ubuntu it's ubuntu.
• The ec2:DescribeInstances action. This is required when using the EC2 Instance Connect CLI
because the wrapper calls this action. IAM principals might already have permission to call this
action from another policy.

The following is an example policy document. You can omit the statement for the
ec2:DescribeInstances action if your users will only use an SSH client to connect to your
instances. You can replace the specified instances in Resource with the wildcard * to grant users
access to all EC2 instances using EC2 Instance Connect.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2-instance-connect:SendSSHPublicKey",
"Resource": [
"arn:aws:ec2:region:account-id:instance/i-1234567890abcdef0",
"arn:aws:ec2:region:account-id:instance/i-0598c7d356eba48d7"
],
"Condition": {
"StringEquals": {
"ec2:osuser": "ami-username"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:DescribeInstances",
"Resource": "*"
}
]
}

The preceding policy allows access to specific instances, identified by their instance ID. Alternatively,
you can use resource tags to control access to an instance. Attribute-based access control is an
authorization strategy that defines permissions based on tags that can be attached to users and
AWS resources. For example, the following policy allows an IAM user to access an instance only if
that instance has a resource tag with key=tag-key and value=tag-value. For more information
about using tags to control access to your AWS resources, see Controlling access to AWS resources in
the IAM User Guide.

{
"Version":"2012-10-17",
"Statement":[

609
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

{
"Effect":"Allow",
"Action":"ec2-instance-connect:SendSSHPublicKey",
"Resource": "arn:aws:ec2:region:account-id:instance/*",
"Condition":{
"StringEquals":{
"aws:ResourceTag/tag-key":"tag-value"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:DescribeInstances",
"Resource": "*"
}
]
}

2. Use the create-policy command to create a new managed policy, and specify the JSON document
that you created to use as the content for the new policy.

$ aws iam create-policy --policy-name my-policy --policy-document file://JSON-file-name

3. Use the attach-user-policy command to attach the managed policy to the specified IAM user. For the
--user-name parameter, specify the friendly name (not the ARN) of the IAM user.

$ aws iam attach-user-policy --policy-arn arn:aws:iam::account-id:policy/my-policy --

user-name IAM-friendly-name

Connect using EC2 Instance Connect

The following instructions explain how to connect to your Linux instance using EC2 Instance Connect. If
you receive an error while attempting to connect to your instance, see Troubleshoot connecting to your
instance (p. 1329).

Options to connect using EC2 Instance Connect

• Connect using the Amazon EC2 console (browser-based client) (p. 611)
• Connect using the EC2 Instance Connect CLI (p. 611)
• Connect using your own key and SSH client (p. 612)

Limitations

• The following Linux distributions are supported:

• Amazon Linux 2 (any version)
• Ubuntu 16.04 or later
• To connect using the Amazon EC2 console (browser-based client), the instance must have a public IPv4
address.
• If the instance does not have a public IP address, you can connect to the instance using an SSH client
or the EC2 Instance Connect CLI, and over a private network, such as from a machine within the same
VPC or from a machine connected to the VPC over VPN or AWS Direct Connect.
• EC2 Instance Connect does not support connecting using an IPv6 address.

Prerequisites

• Install Instance Connect on your instance.

610
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

For more information, see Set up EC2 Instance Connect (p. 604).
• (Optional) Install an SSH client on your local computer.

There is no need to install an SSH client if users only use the Amazon EC2 console (browser-based
client) or the EC2 Instance Connect CLI to connect to an instance. Your local computer most likely has
an SSH client installed by default. You can check for an SSH client by typing ssh at the command line.
If your local computer doesn't recognize the command, you can install an SSH client. For information
about installing an SSH client on Linux or macOS X, see http://www.openssh.com. For information
about installing an SSH client on Windows 10, see OpenSSH in Windows.
• (Optional) Install the EC2 Instance Connect CLI on your local computer.

There is no need to install the EC2 Instance Connect CLI if users only use the Amazon EC2 console
(browser-based client) or an SSH client to connect to an instance. For more information, see Task 3:
(Optional) Install the EC2 Instance Connect CLI (p. 608). This connection method works for instances
with public IP addresses.

Connect using the Amazon EC2 console (browser-based client)

You can connect to an instance using the Amazon EC2 console (browser-based client) by selecting
the instance from the console and choosing to connect using EC2 Instance Connect. Instance Connect
handles the permissions and provides a successful connection.

To connect to your instance using the browser-based client from the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Connect.
4. Choose EC2 Instance Connect.
5. Verify the user name and choose Connect to open a terminal window.

Connect using the EC2 Instance Connect CLI

You can connect to an instance using the EC2 Instance Connect CLI by providing only the instance ID,
while the Instance Connect CLI performs the following three actions in one call: it generates a one-time-
use SSH public key, pushes the key to the instance where it remains for 60 seconds, and connects the
user to the instance. You can use basic SSH/SFTP commands with the Instance Connect CLI.

This connection method works for instances with public and private IP addresses. When connecting to an
instance that only has private IP addresses, the local computer from which you are initiating the session
must have connectivity to the EC2 Instance Connect service endpoint (to push your SSH public key to the
instance) as well as network connectivity to the instance's private IP address. The EC2 Instance Connect
service endpoint is reachable over the internet or over an AWS Direct Connect public virtual interface. To
connect to the instance's private IP address, you can leverage services such as AWS Direct Connect, AWS
Site-to-Site VPN, or VPC peering.
Note
-i is not supported when using mssh. When using the mssh command to connect to your
instance, you do not need to specify any kind of identity file because Instance Connect manages
the key pair.

Amazon Linux 2

To connect to an instance using the EC2 Instance Connect CLI

Use the mssh command with the instance ID as follows. You do not need to specify the user name
for the AMI.

611
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

$ mssh i-001234a4bf70dec41EXAMPLE

Ubuntu

To connect to an instance using the EC2 Instance Connect CLI

Use the mssh command with the instance ID and the default user name for the Ubuntu AMI as
follows. You must specify the user name for the AMI or you get the following error: Authentication
failed.

$ mssh ubuntu@i-001234a4bf70dec41EXAMPLE

Connect using your own key and SSH client

You can use your own SSH key and connect to your instance from the SSH client of your choice while
using the EC2 Instance Connect API. This enables you to benefit from the Instance Connect capability to
push a public key to the instance. This connection method works for instances with public and private IP
addresses.

Requirements

• The supported RSA key types are OpenSSH and SSH2. The supported lengths are 2048 and 4096. For
more information, see Option 2: Import your own public key to Amazon EC2 (p. 1046).
• When connecting to an instance that only has private IP addresses, the local computer from which you
are initiating the SSH session must have connectivity to the EC2 Instance Connect service endpoint (to
push your SSH public key to the instance) as well as network connectivity to the instance's private IP
address to establish the SSH session. The EC2 Instance Connect service endpoint is reachable over the
internet or over an AWS Direct Connect public virtual interface. To connect to the instance's private IP
address, you can leverage services such as AWS Direct Connect, AWS Site-to-Site VPN, or VPC peering.

To connect to your instance using your own key and any SSH client

1. (Optional) Generate new SSH private and public keys

You can generate new SSH private and public keys, my_rsa_key and my_rsa_key.pub, using the
following command:

$ ssh-keygen -t rsa -f my_rsa_key

2. Push your SSH public key to the instance

Use the send-ssh-public-key command to push your SSH public key to the instance. If you launched
your instance using Amazon Linux 2, the default user name for the AMI is ec2-user. If you launched
your instance using Ubuntu, the default user name for the AMI is ubuntu.

The following example pushes the public key to the specified instance in the specified Availability
Zone, to authenticate ec2-user:

$ aws ec2-instance-connect send-ssh-public-key \

--instance-id i-001234a4bf70dec41EXAMPLE \
--availability-zone us-west-2b \
--instance-os-user ec2-user \
--ssh-public-key file://my_rsa_key.pub

612
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

3. Connect to the instance using your private key

Use the ssh command to connect to the instance using the private key before the public key is
removed from the instance metadata (you have 60 seconds before it is removed). Specify the private
key that corresponds to the public key, the default user name for the AMI that you used to launch
your instance, and the instance's public DNS name (if connecting over a private network, specify the
private DNS name or IP address). Add the IdentitiesOnly=yes option to ensure that only the
files in the ssh config and the specified key are used for the connection.

$ ssh -o "IdentitiesOnly=yes" -i my_rsa_key ec2-

[email protected]

Uninstall EC2 Instance Connect

To disable EC2 Instance Connect, connect to your instance and uninstall the ec2-instance-connect
package that you installed on the OS. If the sshd configuration matches what it was set to when
you installed EC2 Instance Connect, uninstalling ec2-instance-connect also removes the sshd
configuration. If you modified the sshd configuration after installing EC2 Instance Connect, you must
update it manually.

Amazon Linux 2

You can uninstall EC2 Instance Connect on Amazon Linux 2 2.0.20190618 or later, where EC2
Instance Connect is preconfigured.

To uninstall EC2 Instance Connect on an instance launched with Amazon Linux 2

1. Connect to your instance using SSH. Specify the SSH key pair you used for your instance when
you launched it and the default user name for the Amazon Linux 2 AMI, which is ec2-user.

For example, the following ssh command connects to the instance with the public DNS
name ec2-a-b-c-d.us-west-2.compute.amazonaws.com, using the key pair
my_ec2_private_key.pem.

$ ssh -i my_ec2_private_key.pem [email protected]

west-2.compute.amazonaws.com

2. Uninstall the ec2-instance-connect package using the yum command.

[ec2-user ~]$ sudo yum remove ec2-instance-connect

Ubuntu

To uninstall EC2 Instance Connect on an instance launched with an Ubuntu AMI

1. Connect to your instance using SSH. Specify the SSH key pair you used for your instance when
you launched it and the default user name for the Ubuntu AMI, which is ubuntu.

For example, the following ssh command connects to the instance with the public DNS
name ec2-a-b-c-d.us-west-2.compute.amazonaws.com, using the key pair
my_ec2_private_key.pem.

$ ssh -i my_ec2_private_key.pem [email protected]

2. Uninstall the ec2-instance-connect package using the apt-get command.

613
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

ubuntu:~$ sudo apt-get remove ec2-instance-connect

Connect to your Linux instance from Windows using PuTTY

After you launch your instance, you can connect to it and use it the way that you'd use a computer sitting
in front of you.

The following instructions explain how to connect to your instance using PuTTY, a free SSH client for
Windows. If you receive an error while attempting to connect to your instance, see Troubleshooting
Connecting to Your Instance.

Prerequisites
Before you connect to your Linux instance using PuTTY, complete the following prerequisites.

Verify that the instance is ready

After you launch an instance, it can take a few minutes for the instance to be ready so that you can
connect to it. Check that your instance has passed its status checks. You can view this information in
the Status check column on the Instances page.
Verify the general prerequisites for connecting to your instance

To find the public DNS name or IP address of your instance and the user name that you should use
to connect to your instance, see General prerequisites for connecting to your instance (p. 598).
Install PuTTY on your local computer

Download and install PuTTY from the PuTTY download page. If you already have an older version of
PuTTY installed, we recommend that you download the latest version. Be sure to install the entire
suite.
Convert your private key using PuTTYgen

Locate the private key (.pem file) for the key pair that you specified when you launched the instance.
Convert the .pem file to a .ppk file for use with PuTTY. For more information, follow the steps in the
next section.

Convert your private key using PuTTYgen

PuTTY does not natively support the private key format for SSH keys. PuTTY provides a tool named
PuTTYgen, which converts keys to the required format for PuTTY. You must convert your private key
(.pem file) into this format (.ppk file) as follows in order to connect to your instance using PuTTY.

To convert your private key

1. From the Start menu, choose All Programs, PuTTY, PuTTYgen.

2. Under Type of key to generate, choose RSA. If your version of PuTTYgen does not include this
option, choose SSH-2 RSA.

3. Choose Load. By default, PuTTYgen displays only files with the extension .ppk. To locate your .pem
file, choose the option to display files of all types.

614
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

4. Select your .pem file for the key pair that you specified when you launched your instance and
choose Open. PuTTYgen displays a notice that the .pem file was successfully imported. Choose OK.
5. To save the key in the format that PuTTY can use, choose Save private key. PuTTYgen displays a
warning about saving the key without a passphrase. Choose Yes.
Note
A passphrase on a private key is an extra layer of protection. Even if your private key is
discovered, it can't be used without the passphrase. The downside to using a passphrase
is that it makes automation harder because human intervention is needed to log on to an
instance, or to copy files to an instance.
6. Specify the same name for the key that you used for the key pair (for example, my-key-pair) and
choose Save. PuTTY automatically adds the .ppk file extension.

Your private key is now in the correct format for use with PuTTY. You can now connect to your instance
using PuTTY's SSH client.

Connect to your Linux instance

Use the following procedure to connect to your Linux instance using PuTTY. You need the .ppk
file that you created for your private key. For more information, see Convert your private key using
PuTTYgen (p. 614) in the preceding section. If you receive an error while attempting to connect to your
instance, see Troubleshooting Connecting to Your Instance.

To connect to your instance using PuTTY

1. Start PuTTY (from the Start menu, choose All Programs, PuTTY, PuTTY).
2. In the Category pane, choose Session and complete the following fields:

a. In the Host Name box, do one of the following:

• (Public DNS) To connect using your instance's public DNS name, enter my-instance-
user-name@my-instance-public-dns-name.
• (IPv6) Alternatively, if your instance has an IPv6 address, to connect using your instance's
IPv6 address, enter my-instance-user-name@my-instance-IPv6-address.

For information about how to get the user name for your instance, and the public DNS name or
IPv6 address of your instance, see Get information about your instance (p. 598).
b. Ensure that the Port value is 22.
c. Under Connection type, select SSH.

615
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

3. (Optional) You can configure PuTTY to automatically send 'keepalive' data at regular intervals to
keep the session active. This is useful to avoid disconnecting from your instance due to session
inactivity. In the Category pane, choose Connection, and then enter the required interval in the
Seconds between keepalives field. For example, if your session disconnects after 10 minutes of
inactivity, enter 180 to configure PuTTY to send keepalive data every 3 minutes.
4. In the Category pane, expand Connection, expand SSH, and then choose Auth. Complete the
following:

a. Choose Browse.
b. Select the .ppk file that you generated for your key pair and choose Open.
c. (Optional) If you plan to start this session again later, you can save the session information for
future use. Under Category, choose Session, enter a name for the session in Saved Sessions,
and then choose Save.
d. Choose Open.
5. If this is the first time you have connected to this instance, PuTTY displays a security alert dialog box
that asks whether you trust the host to which you are connecting.

a. (Optional) Verify that the fingerprint in the security alert dialog box matches the fingerprint
that you previously obtained in (Optional) Get the instance fingerprint (p. 600). If these
fingerprints don't match, someone might be attempting a "man-in-the-middle" attack. If they
match, continue to the next step.
b. Choose Yes. A window opens and you are connected to your instance.
Note
If you specified a passphrase when you converted your private key to PuTTY's format,
you must provide that passphrase when you log in to the instance.

If you receive an error while attempting to connect to your instance, see Troubleshooting Connecting to
Your Instance.

616
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

Transfer files to your Linux instance using the PuTTY Secure Copy client
The PuTTY Secure Copy client (PSCP) is a command line tool that you can use to transfer files between
your Windows computer and your Linux instance. If you prefer a graphical user interface (GUI), you
can use an open source GUI tool named WinSCP. For more information, see Transfer files to your Linux
instance using WinSCP (p. 617).

To use PSCP, you need the private key you generated in Convert your private key using
PuTTYgen (p. 614). You also need the public DNS name of your Linux instance, or the IPv6 address if
your instance has one.

The following example transfers the file Sample_file.txt from the C:\ drive on a Windows computer
to the my-instance-user-name home directory on an Amazon Linux instance. To transfer a file, use
one of the following commands.

• (Public DNS) To transfer a file using your instance's public DNS name, enter the following command.

pscp -i C:\path\my-key-pair.ppk C:\path\Sample_file.txt my-instance-user-name@my-

instance-public-dns-name:/home/my-instance-user-name/Sample_file.txt

• (IPv6) Alternatively, if your instance has an IPv6 address, to transfer a file using your instance's IPv6
address, enter the following command. The IPv6 address must be enclosed in square brackets ([ ]).

pscp -i C:\path\my-key-pair.ppk C:\path\Sample_file.txt my-instance-user-name@[my-

instance-IPv6-address]:/home/my-instance-user-name/Sample_file.txt

Transfer files to your Linux instance using WinSCP

WinSCP is a GUI-based file manager for Windows that allows you to upload and transfer files to a remote
computer using the SFTP, SCP, FTP, and FTPS protocols. WinSCP allows you to drag and drop files from
your Windows computer to your Linux instance or synchronize entire directory structures between the
two systems.

Requirements

• You must have the private key that you generated in Convert your private key using
PuTTYgen (p. 614).
• You must have the public DNS name of your Linux instance.
• Your Linux instance must have scp installed. For some operating systems, you install the openssh-
clients package. For others, such as the Amazon ECS-optimized AMI, you install the scp package.
Check the documentation for your Linux distribution.

To connect to your instance using WinSCP

1. Download and install WinSCP from http://winscp.net/eng/download.php. For most users, the
default installation options are OK.
2. Start WinSCP.
3. At the WinSCP login screen, for Host name, enter one of the following:
• (Public DNS or IPv4 address) To log in using your instance's public DNS name or public IPv4
address, enter the public DNS name or public IPv4 address for your instance.
• (IPv6) Alternatively, if your instance has an IPv6 address, to log in using your instance's IPv6
address, enter the IPv6 address for your instance.
4. For User name, enter the default user name for your AMI.

• For Amazon Linux 2 or the Amazon Linux AMI, the user name is ec2-user.

617
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

• For a CentOS AMI, the user name is centos.

• For a Debian AMI, the user name is admin.
• For a Fedora AMI, the user name is ec2-user or fedora.
• For a RHEL AMI, the user name is ec2-user or root.
• For a SUSE AMI, the user name is ec2-user or root.
• For an Ubuntu AMI, the user name is ubuntu.
• Otherwise, if ec2-user and root don't work, check with the AMI provider.
5. Specify the private key for your instance. For Private key, enter the path to your private key,
or choose the "..." button to browse for the file. To open the advanced site settings, for newer
versions of WinSCP, choose Advanced. To find the Private key file setting, under SSH, choose
Authentication.

Here is a screenshot from WinSCP version 5.9.4:

WinSCP requires a PuTTY private key file (.ppk). You can convert a .pem security key file to
the .ppk format using PuTTYgen. For more information, see Convert your private key using
PuTTYgen (p. 614).
6. (Optional) In the left panel, choose Directories. For Remote directory, enter the path for the
directory to which to add files. To open the advanced site settings for newer versions of WinSCP,
choose Advanced. To find the Remote directory setting, under Environment, choose Directories.

618
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

7. Choose Login. To add the host fingerprint to the host cache, choose Yes.

8. After the connection is established, in the connection window your Linux instance is on the right
and your local machine is on the left. You can drag and drop files between the remote file system
and your local machine. For more information on WinSCP, see the project documentation at http://
winscp.net/eng/docs/start.

If you receive an error that you cannot run SCP to start the transfer, verify that you installed scp on
the Linux instance.

Connect to your Linux instance from Windows using Windows

Subsystem for Linux
After you launch your instance, you can connect to it and use it the way that you'd use a computer sitting
in front of you.

The following instructions explain how to connect to your instance using a Linux distribution on the
Windows Subsystem for Linux (WSL). WSL is a free download and enables you to run native Linux
command line tools directly on Windows, alongside your traditional Windows desktop, without the
overhead of a virtual machine.

By installing WSL, you can use a native Linux environment to connect to your Linux EC2 instances instead
of using PuTTY or PuTTYgen. The Linux environment makes it easier to connect to your Linux instances
because it comes with a native SSH client that you can use to connect to your Linux instances and change
the permissions of the .pem key file. The Amazon EC2 console provides the SSH command for connecting

619
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

to the Linux instance, and you can get verbose output from the SSH command for troubleshooting. For
more information, see the Windows Subsystem for Linux Documentation.
Note
After you've installed the WSL, all the prerequisites and steps are the same as those described in
Connect to your Linux instance using SSH (p. 601), and the experience is just like using native
Linux.

If you receive an error while attempting to connect to your instance, see Troubleshooting Connecting to
Your Instance.

Contents
• Prerequisites (p. 601)
• Connect to your Linux instance using WSL (p. 620)
• Transfer files to Linux instances from Linux using SCP (p. 621)
• Uninstall WSL (p. 623)

Prerequisites
Before you connect to your Linux instance, complete the following prerequisites.

Verify that the instance is ready

After you launch an instance, it can take a few minutes for the instance to be ready so that you can
connect to it. Check that your instance has passed its status checks. You can view this information in
the Status check column on the Instances page.
Verify the general prerequisites for connecting to your instance

To find the public DNS name or IP address of your instance and the user name that you should use
to connect to your instance, see General prerequisites for connecting to your instance (p. 598).
Install the Windows Subsystem for Linux (WSL) and a Linux distribution on your local computer

Install the WSL and a Linux distribution using the instructions in the Windows 10 Installation Guide.
The example in the instructions installs the Ubuntu distribution of Linux, but you can install any
distribution. You are prompted to restart your computer for the changes to take effect.
Copy the private key from Windows to WSL

In a WSL terminal window, copy the .pem file (for the key pair that you specified when you launched
the instance) from Windows to WSL. Note the fully-qualified path to the .pem file on WSL to use
when connecting to your instance. For information about how to specify the path to your Windows
hard drive, see How do I access my C drive?. For more information about key pairs and Windows
instances, see Amazon EC2 key pairs and Windows instances.

cp /mnt/<Windows drive letter>/path/my-key-pair.pem ~/WSL-path/my-key-pair.pem

Connect to your Linux instance using WSL

Use the following procedure to connect to your Linux instance using the Windows Subsystem for
Linux (WSL). If you receive an error while attempting to connect to your instance, see Troubleshooting
Connecting to Your Instance.

To connect to your instance using SSH

1. In a terminal window, use the ssh command to connect to the instance. You specify the path and
file name of the private key (.pem), the user name for your instance, and the public DNS name or

620
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

IPv6 address for your instance. For more information about how to find the private key, the user
name for your instance, and the DNS name or IPv6 address for an instance, see Locate the private
key (p. 600) and Get information about your instance (p. 598). To connect to your instance, use
one of the following commands.
• (Public DNS) To connect using your instance's public DNS name, enter the following command.

ssh -i /path/my-key-pair.pem my-instance-user-name@my-instance-public-dns-name

• (IPv6) Alternatively, if your instance has an IPv6 address, you can connect to the instance using
its IPv6 address. Specify the ssh command with the path to the private key (.pem) file, the
appropriate user name, and the IPv6 address.

ssh -i /path/my-key-pair.pem my-instance-user-name@my-instance-IPv6-address

You see a response like the following:

The authenticity of host 'ec2-198-51-100-1.compute-1.amazonaws.com (10.254.142.33)'

can't be established.
RSA key fingerprint is 1f:51:ae:28:bf:89:e9:d8:1f:25:5d:37:2d:7d:b8:ca:9f:f5:f1:6f.
Are you sure you want to continue connecting (yes/no)?

2. (Optional) Verify that the fingerprint in the security alert matches the fingerprint that you previously
obtained in (Optional) Get the instance fingerprint (p. 600). If these fingerprints don't match,
someone might be attempting a "man-in-the-middle" attack. If they match, continue to the next
step.
3. Enter yes.

You see a response like the following:

Warning: Permanently added 'ec2-198-51-100-1.compute-1.amazonaws.com' (RSA)

to the list of known hosts.

Transfer files to Linux instances from Linux using SCP

One way to transfer files between your local computer and a Linux instance is to use the secure copy
protocol (SCP). This section describes how to transfer files with SCP. The procedure is similar to the
procedure for connecting to an instance with SSH.

Prerequisites

• Verify the general prerequisites for transferring files to your instance.

The general prerequisites for transferring files to an instance are the same as the general prerequisites
for connecting to an instance. For more information, see General prerequisites for connecting to your
instance (p. 598).
• Install an SCP client

Most Linux, Unix, and Apple computers include an SCP client by default. If yours doesn't, the OpenSSH
project provides a free implementation of the full suite of SSH tools, including an SCP client. For more
information, see http://www.openssh.org.

The following procedure steps you through using SCP to transfer a file. If you've already connected to
the instance with SSH and have verified its fingerprints, you can start with the step that contains the SCP
command (step 4).

621
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect

To use SCP to transfer a file

1. Transfer a file to your instance using the instance's public DNS name. For example, if the name of
the private key file is my-key-pair, the file to transfer is SampleFile.txt, the user name is my-
instance-user-name, and the public DNS name of the instance is my-instance-public-dns-
name or the IPv6 address is my-instance-IPv6-address, use one the following commands to
copy the file to the my-instance-user-name home directory.
• (Public DNS) To transfer a file using your instance's public DNS name, enter the following
command.

scp -i /path/my-key-pair.pem /path/SampleFile.txt my-instance-user-name@my-

instance-public-dns-name:~

• (IPv6) Alternatively, if your instance has an IPv6 address, you can transfer a file using the
instance's IPv6 address. The IPv6 address must be enclosed in square brackets ([ ]), which must
be escaped (\).

scp -i /path/my-key-pair.pem /path/SampleFile.txt my-instance-user-name@\[my-

instance-IPv6-address\]:~

You see a response like the following:

The authenticity of host 'ec2-198-51-100-1.compute-1.amazonaws.com (10.254.142.33)'

can't be established.
RSA key fingerprint is 1f:51:ae:28:bf:89:e9:d8:1f:25:5d:37:2d:7d:b8:ca:9f:f5:f1:6f.
Are you sure you want to continue connecting (yes/no)?

2. (Optional) Verify that the fingerprint in the security alert matches the fingerprint that you previously
obtained in (Optional) Get the instance fingerprint (p. 600). If these fingerprints don't match,
someone might be attempting a "man-in-the-middle" attack. If they match, continue to the next
step.
3. Enter yes.

You see a response like the following:

Warning: Permanently added 'ec2-198-51-100-1.compute-1.amazonaws.com' (RSA)

to the list of known hosts.
Sending file modes: C0644 20 SampleFile.txt
Sink: C0644 20 SampleFile.txt
SampleFile.txt 100% 20 0.0KB/s 00:00

If you receive a "bash: scp: command not found" error, you must first install scp on your Linux
instance. For some operating systems, this is located in the openssh-clients package. For
Amazon Linux variants, such as the Amazon ECS-optimized AMI, use the following command to
install scp:

[ec2-user ~]$ sudo yum install -y openssh-clients

4. To transfer files in the other direction (from your Amazon EC2 instance to your local computer),
reverse the order of the host parameters. For example, to transfer the SampleFile.txt file from
your EC2 instance back to the home directory on your local computer as SampleFile2.txt, use
one of the following commands on your local computer.
• (Public DNS) To transfer a file using your instance's public DNS name, enter the following
command.

622
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Stop and start

scp -i /path/my-key-pair.pem my-instance-user-

[email protected]:~/SampleFile.txt ~/SampleFile2.txt

• (IPv6) Alternatively, if your instance has an IPv6 address, to transfer files in the other direction
using the instance's IPv6 address, enter the following command.

scp -i /path/my-key-pair.pem my-instance-user-name@

\[2001:db8:1234:1a00:9691:9503:25ad:1761\]:~/SampleFile.txt ~/SampleFile2.txt

Uninstall WSL
For information about uninstalling Windows Subsystem for Linux, see How do I uninstall a WSL
Distribution?.

Connect to your Linux instance using Session Manager

Session Manager is a fully managed AWS Systems Manager capability that lets you manage your Amazon
EC2 instances through an interactive one-click browser-based shell or through the AWS CLI. You can
use Session Manager to start a session with an instance in your account. After the session is started, you
can run bash commands as you would through any other connection type. For more information about
Session Manager, see AWS Systems Manager Session Manager in the AWS Systems Manager User Guide.

Before attempting to connect to an instance using Session Manager, ensure that the necessary setup
steps have been completed. For more information and instructions, see Getting Started with Session
Manager.

To connect to a Linux instance using Session Manager using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Connect.
4. For Connection method, choose Session Manager.
5. Choose Connect.

Troubleshooting

If you receive an error that you’re not authorized to perform one or more Systems Manager actions
(ssm:command-name), then you must update your policies to allow you to start sessions from the
Amazon EC2 console. For more information, see Quickstart Default IAM Policies for Session Manager in
the AWS Systems Manager User Guide.

Stop and start your instance

You can stop and start your instance if it has an Amazon EBS volume as its root device. The instance
retains its instance ID, but can change as described in the Overview (p. 624) section.

When you stop an instance, we shut it down. We don't charge usage for a stopped instance, or data
transfer fees, but we do charge for the storage for any Amazon EBS volumes. Each time you start a
stopped instance we charge a minimum of one minute for usage. After one minute, we charge only for
the seconds you use. For example, if you run an instance for 20 seconds and then stop it, we charge for
a full one minute. If you run an instance for 3 minutes and 40 seconds, we charge for exactly 3 minutes
and 40 seconds of usage.

623
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Stop and start

While the instance is stopped, you can treat its root volume like any other volume, and modify it (for
example, repair file system problems or update software). You just detach the volume from the stopped
instance, attach it to a running instance, make your changes, detach it from the running instance, and
then reattach it to the stopped instance. Make sure that you reattach it using the storage device name
that's specified as the root device in the block device mapping for the instance.

If you decide that you no longer need an instance, you can terminate it. As soon as the state of an
instance changes to shutting-down or terminated, we stop charging for that instance. For more
information, see Terminate your instance (p. 643). If you'd rather hibernate the instance, see Hibernate
your Linux instance (p. 627). For more information, see Differences between reboot, stop, hibernate,
and terminate (p. 523).

Contents
• Overview (p. 624)
• What happens when you stop an instance (p. 625)
• Stop and start your instances (p. 625)
• Modify a stopped instance (p. 626)
• Troubleshoot stopping your instance (p. 627)

Overview
You can only stop an Amazon EBS-backed instance. To verify the root device type of your instance,
describe the instance and check whether the device type of its root volume is ebs (Amazon EBS-backed
instance) or instance store (instance store-backed instance). For more information, see Determin the
root device type of your AMI (p. 102).

When you stop a running instance, the following happens:

• The instance performs a normal shutdown and stops running; its status changes to stopping and
then stopped.
• Any Amazon EBS volumes remain attached to the instance, and their data persists.
• Any data stored in the RAM of the host computer or the instance store volumes of the host computer
is gone.
• In most cases, the instance is migrated to a new underlying host computer when it's started (though in
some cases, it remains on the current host).
• The instance retains its private IPv4 addresses and any IPv6 addresses when stopped and started. We
release the public IPv4 address and assign a new one when you start it.
• The instance retains its associated Elastic IP addresses. You're charged for any Elastic IP addresses
associated with a stopped instance. With EC2-Classic, an Elastic IP address is dissociated from your
instance when you stop it. For more information, see EC2-Classic (p. 943).
• When you stop and start a Windows instance, the EC2Config service performs tasks on the instance,
such as changing the drive letters for any attached Amazon EBS volumes. For more information
about these defaults and how you can change them, see Configuring a Windows instance using the
EC2Config service in the Amazon EC2 User Guide for Windows Instances.
• If your instance is in an Auto Scaling group, the Amazon EC2 Auto Scaling service marks the stopped
instance as unhealthy, and may terminate it and launch a replacement instance. For more information,
see Health Checks for Auto Scaling Instances in the Amazon EC2 Auto Scaling User Guide.
• When you stop a ClassicLink instance, it's unlinked from the VPC to which it was linked. You must
link the instance to the VPC again after starting it. For more information about ClassicLink, see
ClassicLink (p. 950).

For more information, see Differences between reboot, stop, hibernate, and terminate (p. 523).

624
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Stop and start

You can modify the following attributes of an instance only when it is stopped:

• Instance type
• User data
• Kernel
• RAM disk

If you try to modify these attributes while the instance is running, Amazon EC2 returns the
IncorrectInstanceState error.

What happens when you stop an instance

When an EC2 instance is stopped using the stop-instances command, the following is registered at
the OS level:

• The API request sends a button press event to the guest.

• Various system services are stopped as a result of the button press event. Graceful shutdown is
triggered by the ACPI shutdown button press event from the hypervisor.
• ACPI shutdown is initiated.
• The instance shuts down when the graceful shutdown process exits. There is no configurable OS
shutdown time.
• If the instance OS does not shut down cleanly within a few minutes, a hard shutdown is performed.

By default, when you initiate a shutdown from an Amazon EBS-backed instance (for example, using the
shutdown or poweroff command), the instance stops. You can change this behavior so that it terminates
instead. For more information, see Change the instance initiated shutdown behavior (p. 646).

Using the halt command from an instance does not initiate a shutdown. If used, the instance does not
terminate; instead, it places the CPU into HLT and the instance remains running.

Stop and start your instances

You can stop and start your Amazon EBS-backed instance using the console or the command line.

New console

To stop and start an Amazon EBS-backed instance using the console

1. When you stop an instance, the data on any instance store volumes is erased. Before you stop an
instance, verify that you've copied any data that you need from your instance store volumes to
persistent storage, such as Amazon EBS or Amazon S3.
2. In the navigation pane, choose Instances and select the instance.
3. Choose Instance state, Stop instance. If this option is disabled, either the instance is already
stopped or its root device is an instance store volume.
4. When prompted for confirmation, choose Stop. It can take a few minutes for the instance to
stop.
5. (Optional) While your instance is stopped, you can modify certain instance attributes. For more
information, see Modify a stopped instance (p. 626).
6. To start the stopped instance, select the instance, and choose Instance state, Start instance.
7. It can take a few minutes for the instance to enter the running state.

625
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Stop and start

Old console

To stop and start an Amazon EBS-backed instance using the console

1. When you stop an instance, the data on any instance store volumes is erased. Before you stop an
instance, verify that you've copied any data that you need from your instance store volumes to
persistent storage, such as Amazon EBS or Amazon S3.
2. In the navigation pane, choose Instances and select the instance.
3. Choose Actions, Instance State, Stop. If this option is disabled, either the instance is already
stopped or its root device is an instance store volume.
4. When prompted for confirmation, choose Yes, Stop. It can take a few minutes for the instance
to stop.
5. (Optional) While your instance is stopped, you can modify certain instance attributes. For more
information, see Modify a stopped instance (p. 626).
6. To start the stopped instance, select the instance, and choose Actions, Instance State, Start.
7. In the confirmation dialog box, choose Yes, Start. It can take a few minutes for the instance to
enter the running state.

To stop and start an Amazon EBS-backed instance using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• stop-instances and start-instances (AWS CLI)

• Stop-EC2Instance and Start-EC2Instance (AWS Tools for Windows PowerShell)

Modify a stopped instance

You can change the instance type, user data, and EBS-optimization attributes of a stopped instance
using the AWS Management Console or the command line interface. You can't use the AWS Management
Console to modify the DeleteOnTermination, kernel, or RAM disk attributes.

To modify an instance attribute

• To change the instance type, see Change the instance type (p. 311).
• To change the user data for your instance, see Work with instance user data (p. 718).
• To enable or disable EBS–optimization for your instance, see Modifying EBS–Optimization (p. 1236).
• To change the DeleteOnTermination attribute of the root volume for your instance, see Update
the block device mapping of a running instance (p. 1301). You are not required to stop the instance to
change this attribute.

To modify an instance attribute using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• modify-instance-attribute (AWS CLI)

• Edit-EC2InstanceAttribute (AWS Tools for Windows PowerShell)

626
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

Troubleshoot stopping your instance

If you have stopped your Amazon EBS-backed instance and it appears "stuck" in the stopping state, you
can forcibly stop it. For more information, see Troubleshoot stopping your instance (p. 1339).

Hibernate your Linux instance

When you hibernate an instance, Amazon EC2 signals the operating system to perform hibernation
(suspend-to-disk). Hibernation saves the contents from the instance memory (RAM) to your Amazon
Elastic Block Store (Amazon EBS) root volume. Amazon EC2 persists the instance's EBS root volume and
any attached EBS data volumes. When you start your instance:

• The EBS root volume is restored to its previous state

• The RAM contents are reloaded
• The processes that were previously running on the instance are resumed
• Previously attached data volumes are reattached and the instance retains its instance ID

You can hibernate an instance only if it's enabled for hibernation (p. 632) and it meets the hibernation
prerequisites (p. 628).

If an instance or application takes a long time to bootstrap and build a memory footprint to become
fully productive, you can use hibernation to pre-warm the instance. To pre-warm the instance, you:

1. Launch it with hibernation enabled.

2. Bring it to a desired state.
3. Hibernate it, ready to be resumed to the same state as needed.

You're not charged for instance usage for a hibernated instance when it is in the stopped state. You
are charged for instance usage while the instance is in the stopping state, when the contents of the
RAM are transferred to the EBS root volume. (This is different from when you stop an instance (p. 623)
without hibernating it.) You're not charged for data transfer. However, you are charged for storage of any
EBS volumes, including storage for the RAM contents.

If you no longer need an instance, you can terminate it at any time, including when it is in a stopped
(hibernated) state. For more information, see Terminate your instance (p. 643).
Note
For information about using hibernation on Windows instances, see Hibernate Your Windows
Instance in the Amazon EC2 User Guide for Windows Instances.

Contents
• Overview of hibernation (p. 628)
• Hibernation prerequisites (p. 628)
• Limitations (p. 629)
• Configure an existing AMI to support hibernation (p. 630)
• Enable hibernation for an instance (p. 632)
• Disable KASLR on an instance (Ubuntu only) (p. 635)
• Hibernate an instance (p. 635)
• Start a hibernated instance (p. 637)
• Troubleshoot hibernation (p. 638)

627
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

Overview of hibernation
The following diagram shows a basic overview of the hibernation process.

When you hibernate a running instance, the following happens:

• When you initiate hibernation, the instance moves to the stopping state. Amazon EC2 signals
the operating system to perform hibernation (suspend-to-disk). The hibernation freezes all of the
processes, saves the contents of the RAM to the EBS root volume, and then performs a regular
shutdown.
• After the shutdown is complete, the instance moves to the stopped state.
• Any EBS volumes remain attached to the instance, and their data persists, including the saved contents
of the RAM.
• Any Amazon EC2 instance store volumes remain attached to the instance, but the data on the instance
store volumes is lost.
• In most cases, the instance is migrated to a new underlying host computer when it's started. This is
also what happens when you stop and start an instance.
• When you start the instance, the instance boots up and the operating system reads in the contents of
the RAM from the EBS root volume, before unfreezing processes to resume its state.
• The instance retains its private IPv4 addresses and any IPv6 addresses. When you start the instance,
the instance continues to retain its private IPv4 addresses and any IPv6 addresses.
• Amazon EC2 releases the public IPv4 address. When you start the instance, Amazon EC2 assigns a new
public IPv4 address to the instance.
• The instance retains its associated Elastic IP addresses. You're charged for any Elastic IP addresses
associated with a hibernated instance. With EC2-Classic, an Elastic IP address is disassociated from
your instance when you hibernate it. For more information, see EC2-Classic (p. 943).
• When you hibernate a ClassicLink instance, it's unlinked from the VPC to which it was linked. You must
link the instance to the VPC again after starting it. For more information, see ClassicLink (p. 950).

For information about how hibernation differs from reboot, stop, and terminate, see Differences between
reboot, stop, hibernate, and terminate (p. 523).

Hibernation prerequisites
To hibernate an instance, the following prerequisites must be in place:

• Supported instance families

• C3, C4, C5
• I3
• M3, M4, M5, M5a, M5ad
• R3, R4, R5, R5a, R5ad
• T2, T3, T3a

628
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

• Instance RAM size - must be less than 150 GB.

• Instance size - not supported for bare metal instances.
• Supported AMIs (must be an HVM AMI that supports hibernation):
• Amazon Linux 2 AMI released 2019.08.29 or later.
• Amazon Linux AMI 2018.03 released 2018.11.16 or later.
• Ubuntu 18.04 LTS - Bionic AMI released with serial number 20190722.1 or later.*
• Ubuntu 16.04 LTS - Xenial AMI.* (Additional configuration is required.)

*We recommend disabling KASLR on instances with Ubuntu 18.04 LTS - Bionic and Ubuntu 16.04
LTS - Xenial. For more information, see Disable KASLR on an instance (Ubuntu only) (p. 635).

To configure your own AMI to support hibernation, see Configure an existing AMI to support
hibernation (p. 630).

Support for other versions of Ubuntu and other operating systems is coming soon.

For information about the supported AMIs for Windows, see Hibernation prerequisites in the Amazon
EC2 User Guide for Windows Instances.
• Root volume type - must be an EBS volume, not an instance store volume.
• Supported EBS volume types - General Purpose SSD (gp2 and gp3) or Provisioned IOPS SSD (gp2
and gp3). If you choose a Provisioned IOPS SSD volume type, to achieve optimum performance for
hibernation, you must provision the EBS volume with the appropriate IOPS. For more information, see
Amazon EBS volume types (p. 1084).
• EBS root volume size - must be large enough to store the RAM contents and accommodate your
expected usage, for example, OS or applications. If you enable hibernation, space is allocated on the
root volume at launch to store the RAM.
• EBS root volume encryption - To use hibernation, the root volume must be encrypted to ensure the
protection of sensitive content that is in memory at the time of hibernation. When RAM data is moved
to the EBS root volume, it is always encrypted. Encryption of the root volume is enforced at instance
launch. Use one of the following three options to ensure that the root volume is an encrypted EBS
volume:
• EBS "single-step" encryption: You can launch encrypted EBS-backed EC2 instances from an
unencrypted AMI and also enable hibernation at the same time. For more information, see Use
encryption with EBS-backed AMIs (p. 161).
• EBS encryption by default: You can enable EBS encryption by default to ensure all new EBS volumes
created in your AWS account are encrypted. This way, you can enable hibernation for your instances
without specifying encryption intent at instance launch. For more information, see Encryption by
default (p. 1178).
• Encrypted AMI: You can enable EBS encryption by using an encrypted AMI to launch your instance.
If your AMI does not have an encrypted root snapshot, you can copy it to a new AMI and request
encryption. For more information, see Encrypt an unencrypted image during copy (p. 165) and Copy
an AMI (p. 170).
• Enable hibernation at launch - You cannot enable hibernation on an existing instance (running or
stopped). For more information, see Enable hibernation for an instance (p. 632).
• Purchasing options - This feature is available for On-Demand Instances and Reserved Instances.
It is not available for Spot Instances. For more information, see Hibernate interrupted Spot
Instances (p. 454).

Limitations
• When you hibernate an instance, the data on any instance store volumes is lost.
• You can't hibernate an instance that has more than 150 GB of RAM.

629
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

• If you create a snapshot or AMI from an instance that is hibernated or has hibernation enabled, you
might not be able to connect to the instance.
• You can't change the instance type or size of an instance with hibernation enabled.
• You cannot hibernate an instance that is in an Auto Scaling group or used by Amazon ECS. If your
instance is in an Auto Scaling group and you try to hibernate it, the Amazon EC2 Auto Scaling service
marks the stopped instance as unhealthy, and might terminate it and launch a replacement instance.
For more information, see Health Checks for Auto Scaling Instances in the Amazon EC2 Auto Scaling
User Guide.
• We do not support keeping an instance hibernated for more than 60 days. To keep the instance for
longer than 60 days, you must start the hibernated instance, stop the instance, and start it.
• We constantly update our platform with upgrades and security patches, which can conflict with
existing hibernated instances. We notify you about critical updates that require a start for hibernated
instances so that we can perform a shutdown or a reboot to apply the necessary upgrades and security
patches.

Configure an existing AMI to support hibernation

To hibernate an instance that was launched using your own AMI, you must first configure your AMI
to support hibernation. For more information, see Update instance software on your Amazon Linux
instance (p. 651).

If you use one of the supported AMIs (p. 628) (except Ubuntu 16.04 LTS), or if you create an AMI based
on one of the supported AMIs, you do not need to configure it to support hibernation. These AMIs are
preconfigured to support hibernation. To configure Ubuntu 16.04 LTS to support hibernation, you need
to install the linux-aws-hwe kernel package version 4.15.0-1058-aws or later and the ec2-hibinit-agent.
For the configuration steps, choose the Ubuntu 16.04 - Xenial tab below.

Amazon Linux 2

To configure an Amazon Linux 2 AMI to support hibernation

1. Update to the latest kernel to 4.14.138-114.102 or later using the following command.

[ec2-user ~]$ sudo yum update kernel

2. Install the ec2-hibinit-agent package from the repositories using the following command.

[ec2-user ~]$ sudo yum install ec2-hibinit-agent

3. Reboot the instance using the following command.

[ec2-user ~]$ sudo reboot

4. Confirm that the kernel version is updated to 4.14.138-114.102 or later using the following
command.

[ec2-user ~]$ uname -a

5. Stop the instance and create an AMI. For more information, see Create a Linux AMI from an
instance (p. 126).

630
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

Amazon Linux

To configure an Amazon Linux AMI to support hibernation

1. Update to the latest kernel to 4.14.77-70.59 or later using the following command.

[ec2-user ~]$ sudo yum update kernel

2. Install the ec2-hibinit-agent package from the repositories using the following command.

[ec2-user ~]$ sudo yum install ec2-hibinit-agent

3. Reboot the instance using the following command.

[ec2-user ~]$ sudo reboot

4. Confirm that the kernel version is updated to 4.14.77-70.59 or greater using the following
command.

[ec2-user ~]$ uname -a

5. Stop the instance and create an AMI. For more information, see Create a Linux AMI from an
instance (p. 126).

Ubuntu 18.04 - Bionic

To configure an Ubuntu 18.04 LTS AMI to support hibernation

1. Update to the latest kernel to 4.15.0-1044 or later using the following commands.

[ec2-user ~]$ sudo apt update

[ec2-user ~]$ sudo apt dist-upgrade

2. Install the ec2-hibinit-agent package from the repositories using the following command.

[ec2-user ~]$ sudo apt install ec2-hibinit-agent

3. Reboot the instance using the following command.

[ec2-user ~]$ sudo reboot

4. Confirm that the kernel version is updated to 4.15.0-1044 or later using the following
command.

[ec2-user ~]$ uname -a

Ubuntu 16.04 - Xenial

To configure an Ubuntu 16.04 LTS AMI to support hibernation

1. Update to the latest kernel to 4.15.0-1058-aws or later using the following commands.

[ec2-user ~]$ sudo apt update

[ec2-user ~]$ sudo apt install linux-aws-hwe

631
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

Note
The linux-aws-hwe kernel package is fully supported by Canonical. The package will
continue to receive regular updates until standard support for Ubuntu 16.04 LTS ends
in April 2021, and will receive additional security updates until the Extended Security
Maintenance support ends in 2024. For more information, see Amazon EC2 Hibernation
for Ubuntu 16.04 LTS now available on the Canonical Ubuntu Blog.
2. Install the ec2-hibinit-agent package from the repositories using the following command.

[ec2-user ~]$ sudo apt install ec2-hibinit-agent

3. Reboot the instance using the following command.

[ec2-user ~]$ sudo reboot

4. Confirm that the kernel version is updated to 4.15.0-1058-aws or later using the following
command.

[ec2-user ~]$ uname -a

Enable hibernation for an instance

To hibernate an instance, it must first be enabled for hibernation. To enable hibernation, you must do it
while launching the instance.
Important
You can't enable or disable hibernation for an instance after you launch it.

Console

To enable hibernation using the console

1. Follow the Launch an instance using the Launch Instance Wizard (p. 526) procedure.
2. On the Choose an Amazon Machine Image (AMI) page, select an AMI that
supports hibernation. For more information about supported AMIs, see Hibernation
prerequisites (p. 628).
3. On the Choose an Instance Type page, select a supported instance type, and choose Next:
Configure Instance Details. For information about supported instance types, see Hibernation
prerequisites (p. 628).
4. On the Configure Instance Details page, for Stop - Hibernate Behavior, select the Enable
hibernation as an additional stop behavior check box.
5. On the Add Storage page, for the root volume, specify the following information:

• For Size (GiB), enter the EBS root volume size. The volume must be large enough to store the
RAM contents and accommodate your expected usage.
• For Volume Type, select a supported EBS volume type, General Purpose SSD (gp2 and gp3) or
Provisioned IOPS SSD (io1 and io2).
• For Encryption, select the encryption key for the volume. If you enabled encryption by
default in this AWS Region, the default encryption key is selected.

For more information about the prerequisites for the root volume, see Hibernation
prerequisites (p. 628).

632
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

6. Continue as prompted by the wizard. When you've finished reviewing your options on the
Review Instance Launch page, choose Launch. For more information, see Launch an instance
using the Launch Instance Wizard (p. 526).

AWS CLI

To enable hibernation using the AWS CLI

Use the run-instances command to launch an instance. Specify the EBS root volume parameters
using the --block-device-mappings file://mapping.json parameter, and enable
hibernation using the --hibernation-options Configured=true parameter.

aws ec2 run-instances \

--image-id ami-0abcdef1234567890 \
--instance-type m5.large \
--block-device-mappings file://mapping.json \
--hibernation-options Configured=true \
--count 1 \
--key-name MyKeyPair

Specify the following in mapping.json:

[
{
"DeviceName": "/dev/xvda",
"Ebs": {
"VolumeSize": 30,
"VolumeType": "gp2",
"Encrypted": true
}
}
]

Note
The value for DeviceName must match the root device name associated with the AMI. To
find the root device name, use the describe-images command, as follows:

aws ec2 describe-images --image-id ami-0abcdef1234567890

If you enabled encryption by default in this AWS Region, you can omit "Encrypted":
true.
PowerShell

To enable hibernation using the AWS Tools for Windows PowerShell

Use the New-EC2Instance command to launch an instance. Specify the EBS root
volume by first defining the block device mapping, and then adding it to the command
using the -BlockDeviceMappings parameter. Enable hibernation using the -
HibernationOptions_Configured $true parameter.

PS C:\> $ebs_encrypt = New-Object Amazon.EC2.Model.BlockDeviceMapping

PS C:\> $ebs_encrypt.DeviceName = "/dev/xvda"
PS C:\> $ebs_encrypt.Ebs = New-Object Amazon.EC2.Model.EbsBlockDevice
PS C:\> $ebs_encrypt.Ebs.VolumeSize = 30
PS C:\> $ebs_encrypt.Ebs.VolumeType = "gp2"
PS C:\> $ebs_encrypt.Ebs.Encrypted = $true

633
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

PS C:\> New-EC2Instance `
-ImageId ami-0abcdef1234567890 `
-InstanceType m5.large `
-BlockDeviceMappings $ebs_encrypt `
-HibernationOptions_Configured $true `
-MinCount 1 `
-MaxCount 1 `
-KeyName MyKeyPair

Note
The value for DeviceName must match the root device name associated with the AMI. To
find the root device name, use the Get-EC2Image command, as follows:

Get-EC2Image -ImageId ami-0abcdef1234567890

If you enabled encryption by default in this AWS Region, you can omit Encrypted =
$true from the block device mapping.

New console

To view if an instance is enabled for hibernation using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and, on the Details tab, in the Instance details section, inspect Stop-
hibernate behavior. Enabled indicates that the instance is enabled for hibernation.

Old console

To view if an instance is enabled for hibernation using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and, in the details pane, inspect Stop - Hibernation behavior. Enabled
indicates that the instance is enabled for hibernation.

AWS CLI

To view if an instance is enabled for hibernation using the AWS CLI

Use the describe-instances command and specify the --filters "Name=hibernation-

options.configured,Values=true" parameter to filter instances that are enabled for
hibernation.

aws ec2 describe-instances \

--filters "Name=hibernation-options.configured,Values=true"

The following field in the output indicates that the instance is enabled for hibernation.

"HibernationOptions": {
"Configured": true
}

634
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

PowerShell

To view if an instance is enabled for hibernation using the AWS Tools for Windows PowerShell

Use the Get-EC2Instance command and specify the -Filter @{ Name="hibernation-

options.configured"; Value="true"} parameter to filter instances that are enabled for
hibernation.

Get-EC2Instance `
-Filter @{ Name="hibernation-options.configured"; Value="true"}

The output lists the EC2 instances that are enabled for hibernation.

Disable KASLR on an instance (Ubuntu only)

To run hibernation on a newly launched instance with Ubuntu 16.04 LTS - Xenial or Ubuntu 18.04 LTS -
Bionic released with serial 20190722.1 or later, we recommend disabling KASLR (Kernel Address Space
Layout Randomization). On Ubuntu 16.04 LTS or Ubuntu 18.04 LTS, KASLR is enabled by default. KASLR
is a standard Linux kernel security feature that helps to mitigate exposure to and ramifications of yet-
undiscovered memory access vulnerabilities by randomizing the base address value of the kernel. With
KASLR enabled, there is a possibility that the instance might not resume after it has been hibernated.

To learn more about KASLR, see Ubuntu Features.

To disable KASLR on an instance launched with Ubuntu

1. Connect to your instance using SSH. For more information, see Connect to your Linux instance using
SSH (p. 601).
2. Open the /etc/default/grub.d/50-cloudimg-settings.cfg file in your editor of choice. Edit
the GRUB_CMDLINE_LINUX_DEFAULT line to append the nokaslr option to its end, as shown in
the following example.

GRUB_CMDLINE_LINUX_DEFAULT="console=tty1 console=ttyS0 nvme_core.io_timeout=4294967295

nokaslr"

3. Save the file and exit your editor.

4. Run the following command to rebuild the grub configuration.

[ec2-user ~]$ sudo update-grub

5. Reboot the instance.

[ec2-user ~]$ sudo reboot

6. Confirm that nokaslr has been added when running the following command.

[ec2-user ~]$ cat /proc/cmdline

The output of the command should include the nokaslr option.

Hibernate an instance
You can hibernate an instance if the instance is enabled for hibernation (p. 632) and meets the
hibernation prerequisites (p. 628). If an instance cannot hibernate successfully, a normal shutdown
occurs.

635
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

New console

To hibernate an Amazon EBS-backed instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select an instance, and choose Instance state, Hibernate instance. If Hibernate instance is
disabled, the instance is already hibernated or stopped, or it can't be hibernated. For more
information, see Hibernation prerequisites (p. 628).
4. When prompted for confirmation, choose Hibernate. It can take a few minutes for the instance
to hibernate. The instance state changes to Stopping while the instance is hibernating, and then
Stopped when the instance has hibernated.

Old console

To hibernate an Amazon EBS-backed instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select an instance, and choose Actions, Instance State, Stop - Hibernate. If Stop - Hibernate
is disabled, the instance is already hibernated or stopped, or it can't be hibernated. For more
information, see Hibernation prerequisites (p. 628).
4. In the confirmation dialog box, choose Yes, Stop - Hibernate. It can take a few minutes for the
instance to hibernate. The Instance State changes to Stopping while the instance is hibernating,
and then Stopped when the instance has hibernated.

AWS CLI

To hibernate an Amazon EBS-backed instance using the AWS CLI

Use the stop-instances command and specify the --hibernate parameter.

aws ec2 stop-instances \

--instance-ids i-1234567890abcdef0 \
--hibernate

PowerShell

To hibernate an Amazon EBS-backed instance using the AWS Tools for Windows PowerShell

Use the Stop-EC2Instance command and specify the -Hibernate $true parameter.

Stop-EC2Instance `
-InstanceId i-1234567890abcdef0 `
-Hibernate $true

New console

To view if hibernation was initiated on an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

636
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

2. In the navigation pane, choose Instances.

3. Select the instance and, on the Details tab, in the Instance details section, inspect State
transition message. The message Client.UserInitiatedHibernate: User initiated hibernate
indicates that hibernation was initiated on the instance.

Old console

To view if hibernation was initiated on an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and, in the details pane, inspect State transition reason message. The
message Client.UserInitiatedHibernate: User initiated hibernate indicates that hibernation
was initiated on the instance.

AWS CLI

To view if hibernation was initiated on an instance using the AWS CLI

Use the describe-instances command and specify the state-reason-code filter to see the
instances on which hibernation was initiated.

aws ec2 describe-instances \

--filters "Name=state-reason-code,Values=Client.UserInitiatedHibernate"

The following field in the output indicates that hibernation was initiated on the instance.

"StateReason": {
"Code": "Client.UserInitiatedHibernate"
}

PowerShell

To view if hibernation was initiated on an instance using the AWS Tools for Windows PowerShell

Use the Get-EC2Instance command and specify the state-reason-code filter to see the instances
on which hibernation was initiated.

Get-EC2Instance `
-Filter @{Name="state-reason-code";Value="Client.UserInitiatedHibernate"}

The output lists the EC2 instances on which hibernation was initiated.

Start a hibernated instance

Start a hibernated instance by starting it in the same way that you would start a stopped instance.

New console

To start a hibernated instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.

637
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Hibernate

3. Select a hibernated instance, and choose Instance state, Start instance. It can take a few
minutes for the instance to enter the running state. During this time, the instance status
checks (p. 744) show the instance in a failed state until the instance has started.

Old console

To start a hibernated instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select a hibernated instance, and choose Actions, Instance State, Start. It can take a few
minutes for the instance to enter the running state. During this time, the instance status
checks (p. 744) show the instance in a failed state until the instance has started.

AWS CLI

To start a hibernated instance using the AWS CLI

Use the start-instances command.

aws ec2 start-instances \

--instance-ids i-1234567890abcdef0

PowerShell

To start a hibernated instance using the AWS Tools for Windows PowerShell

Use the Start-EC2Instance command.

Start-EC2Instance `
-InstanceId i-1234567890abcdef0

Troubleshoot hibernation
Use this information to help diagnose and fix issues that you might encounter when hibernating an
instance.

Can't hibernate immediately after launch

If you try to hibernate an instance too quickly after you've launched it, you get an error.

You must wait for about two minutes after launch before hibernating.

Takes too long to transition from stopping to stopped, and memory state not
restored after start
If it takes a long time for your hibernating instance to transition from the stopping state to stopped,
and if the memory state is not restored after you start, this could indicate that hibernation was not
properly configured.

Check the instance system log and look for messages that are related to hibernation. To access the
system log, connect (p. 598) to the instance or use the get-console-output command. Find the log lines
from the hibinit-agent. If the log lines indicate a failure or the log lines are missing, there was most
likely a failure configuring hibernation at launch.

638
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Reboot

For example, the following message indicates that the instance root volume is not large enough:
hibinit-agent: Insufficient disk space. Cannot create setup for hibernation.
Please allocate a larger root device.

If the last log line from the hibinit-agent is hibinit-agent: Running: swapoff /swap,
hibernation was successfully configured.

If you do not see any logs from these processes, your AMI might not support hibernation. For
information about supported AMIs, see Hibernation prerequisites (p. 628). If you used your
own AMI, make sure that you followed the instructions for Configure an existing AMI to support
hibernation (p. 630).

Instance "stuck" in the stopping state

If you hibernated your instance and it appears "stuck" in the stopping state, you can forcibly stop it. For
more information, see Troubleshoot stopping your instance (p. 1339).

Reboot your instance

An instance reboot is equivalent to an operating system reboot. In most cases, it takes only a few
minutes to reboot your instance. When you reboot an instance, it keeps its public DNS name (IPv4),
private IPv4 address, IPv6 address (if applicable), and any data on its instance store volumes.

Rebooting an instance doesn't start a new instance billing period (with a minimum one-minute charge),
unlike stopping and starting your instance.

We might schedule your instance for a reboot for necessary maintenance, such as to apply updates that
require a reboot. No action is required on your part; we recommend that you wait for the reboot to occur
within its scheduled window. For more information, see Scheduled events for your instances (p. 750).

We recommend that you use the Amazon EC2 console, a command line tool, or the Amazon EC2 API to
reboot your instance instead of running the operating system reboot command from your instance. If
you use the Amazon EC2 console, a command line tool, or the Amazon EC2 API to reboot your instance,
we perform a hard reboot if the instance does not cleanly shut down within a few minutes. If you use
AWS CloudTrail, then using Amazon EC2 to reboot your instance also creates an API record of when your
instance was rebooted.

New console

To reboot an instance using the console

1. Open the Amazon EC2 console.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, Instance state, Reboot instance.
4. Choose Reboot when prompted for confirmation. The instance remains in the running state.

Old console

To reboot an instance using the console

1. Open the Amazon EC2 console.

2. In the navigation pane, choose Instances.

639
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Retire

3. Select the instance and choose Actions, Instance State, Reboot.

4. Choose Yes, Reboot when prompted for confirmation. The instance remains in the running
state.

To reboot an instance using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• reboot-instances (AWS CLI)

• Restart-EC2Instance (AWS Tools for Windows PowerShell)

Instance retirement
An instance is scheduled to be retired when AWS detects irreparable failure of the underlying hardware
that hosts the instance. When an instance reaches its scheduled retirement date, it is stopped or
terminated by AWS.

• If your instance root device is an Amazon EBS volume, the instance is stopped, and you can start it
again at any time. Starting the stopped instance migrates it to new hardware.
• If your instance root device is an instance store volume, the instance is terminated, and cannot be used
again.

For more information about the types of instance events, see Scheduled events for your
instances (p. 750).

Contents
• Identify instances scheduled for retirement (p. 640)
• Actions to take for EBS-backed instances scheduled for retirement (p. 641)
• Actions to take for instance-store backed instances scheduled for retirement (p. 642)

Identify instances scheduled for retirement

If your instance is scheduled for retirement, you receive an email prior to the event with the instance
ID and retirement date. You can also check for instances that are scheduled for retirement using the
Amazon EC2 console or the command line.
Important
If an instance is scheduled for retirement, we recommend that you take action as soon as
possible because the instance might be unreachable. (The email notification you receive states
the following: "Due to this degradation your instance could already be unreachable.") For more
information about the recommended action you should take, see Check if your instance is
reachable.

Ways to identify instances scheduled for retirement

• Email notification (p. 640)
• Console identification (p. 641)

Email notification
If your instance is scheduled for retirement, you receive an email prior to the event with the instance ID
and retirement date.

640
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Retire

The email is sent to the address that's associated with your account. It's the same email address that you
use to log in to the AWS Management Console. To update the contact information for your account, go
to the Account Settings page.

Console identification
If you use an email account that you do not check regularly for instance retirement notifications, you can
use the Amazon EC2 console or the command line to determine if any of your instances are scheduled for
retirement.

To identify instances scheduled for retirement using the console

1. Open the Amazon EC2 console.

2. In the navigation pane, choose EC2 Dashboard. Under Scheduled events, you can see the events
that are associated with your Amazon EC2 instances and volumes, organized by Region.

3. If you have an instance with a scheduled event listed, select its link below the Region name to go to
the Events page.
4. The Events page lists all resources that have events associated with them. To view instances that are
scheduled for retirement, select Instance resources from the first filter list, and then Instance stop
or retirement from the second filter list.
5. If the filter results show that an instance is scheduled for retirement, select it, and note the date and
time in the Start time field in the details pane. This is your instance retirement date.

To identify instances scheduled for retirement using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• describe-instance-status (AWS CLI)

• Get-EC2InstanceStatus (AWS Tools for Windows PowerShell)

Actions to take for EBS-backed instances scheduled for

retirement
To preserve the data on your retiring instance, you can perform one of the following actions. It's
important that you take this action before the instance retirement date to prevent unforeseen downtime
and data loss.

If you are not sure whether your instance is backed by EBS or instance store, see Determine the root
device type of your instance (p. 23).

Check if your instance is reachable

When you are notified that your instance is scheduled for retirement, we recommend that you take the
following action as soon as possible:

• Check if your instance is reachable by either connecting (p. 598) to or pinging your instance.

641
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Retire

• If your instance is reachable, you should plan to stop/start your instance at an appropriate time before
the scheduled retirement date, when the impact is minimal. For more information about stopping
and starting your instance, and what to expect when your instance is stopped, such as the effect on
public, private, and Elastic IP addresses that are associated with your instance, see Stop and start your
instance (p. 623). Note that data on instance store volumes is lost when you stop and start your
instance.
• If your instance is unreachable, you should take immediate action and perform a stop/start (p. 623)
to recover your instance.
• Alternatively, if you want to terminate (p. 643) your instance, plan to do so as soon as possible so
that you stop incurring charges for the instance.

Create a backup of your instance

Create an EBS-backed AMI from your instance so that you have a backup. To ensure data integrity,
stop the instance before you create the AMI. You can wait for the scheduled retirement date when the
instance is stopped, or stop the instance yourself before the retirement date. You can start the instance
again at any time. For more information, see Create an Amazon EBS-backed Linux AMI (p. 125).

Launch a replacement instance

After you create an AMI from your instance, you can use the AMI to launch a replacement instance. From
the Amazon EC2 console, select your new AMI and then choose Actions, Launch. Follow the wizard to
launch your instance. For more information about each step in the wizard, see Launch an instance using
the Launch Instance Wizard (p. 526).

Actions to take for instance-store backed instances scheduled

for retirement
To preserve the data on your retiring instance, you can perform one of the following actions. It's
important that you take this action before the instance retirement date to prevent unforeseen downtime
and data loss.
Warning
If your instance store-backed instance passes its retirement date, it is terminated and you cannot
recover the instance or any data that was stored on it. Regardless of the root device of your
instance, the data on instance store volumes is lost when the instance is retired, even if the
volumes are attached to an EBS-backed instance.

Check if your instance is reachable

When you are notified that your instance is scheduled for retirement, we recommend that you take the
following action as soon as possible:

• Check if your instance is reachable by either connecting (p. 598) to or pinging your instance.
• If your instance is unreachable, there is likely very little that can be done to recover your instance.
For more information, see Troubleshoot an unreachable instance (p. 1363). AWS will terminate your
instance on the scheduled retirement date, so, for an unreachable instance, you can immediately
terminate (p. 643) the instance yourself.

Launch a replacement instance

Create an instance store-backed AMI from your instance using the AMI tools, as described in Create an
instance store-backed Linux AMI (p. 130). From the Amazon EC2 console, select your new AMI and then
choose Actions, Launch. Follow the wizard to launch your instance. For more information about each
step in the wizard, see Launch an instance using the Launch Instance Wizard (p. 526).

Convert your instance to an EBS-backed instance

642
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Terminate

Transfer your data to an EBS volume, take a snapshot of the volume, and then create AMI from the
snapshot. You can launch a replacement instance from your new AMI. For more information, see Convert
your instance store-backed AMI to an Amazon EBS-backed AMI (p. 142).

Terminate your instance

You can delete your instance when you no longer need it. This is referred to as terminating your instance.
As soon as the state of an instance changes to shutting-down or terminated, you stop incurring
charges for that instance.

You can't connect to or start an instance after you've terminated it. However, you can launch additional
instances using the same AMI. If you'd rather stop and start your instance, or hibernate it, see Stop and
start your instance (p. 623) or Hibernate your Linux instance (p. 627). For more information, see
Differences between reboot, stop, hibernate, and terminate (p. 523).

Contents
• Instance termination (p. 643)
• What happens when you terminate an instance (p. 644)
• Terminate an instance (p. 644)
• Enable termination protection (p. 645)
• Change the instance initiated shutdown behavior (p. 646)
• Preserve Amazon EBS volumes on instance termination (p. 646)
• Troubleshoot instance termination (p. 648)

Instance termination
After you terminate an instance, it remains visible in the console for a short while, and then the entry
is automatically deleted. You cannot delete the terminated instance entry yourself. After an instance is
terminated, resources such as tags and volumes are gradually disassociated from the instance and may
no longer be visible on the terminated instance after a short while.

When an instance terminates, the data on any instance store volumes associated with that instance is
deleted.

By default, Amazon EBS root device volumes are automatically deleted when the instance terminates.
However, by default, any additional EBS volumes that you attach at launch, or any EBS volumes that
you attach to an existing instance persist even after the instance terminates. This behavior is controlled
by the volume's DeleteOnTermination attribute, which you can modify. For more information, see
Preserve Amazon EBS volumes on instance termination (p. 646).

You can prevent an instance from being terminated accidentally by someone using the AWS
Management Console, the CLI, and the API. This feature is available for both Amazon EC2 instance store-
backed and Amazon EBS-backed instances. Each instance has a DisableApiTermination attribute
with the default value of false (the instance can be terminated through Amazon EC2). You can modify
this instance attribute while the instance is running or stopped (in the case of Amazon EBS-backed
instances). For more information, see Enable termination protection (p. 645).

You can control whether an instance should stop or terminate when shutdown is initiated from the
instance using an operating system command for system shutdown. For more information, see Change
the instance initiated shutdown behavior (p. 646).

If you run a script on instance termination, your instance might have an abnormal termination, because
we have no way to ensure that shutdown scripts run. Amazon EC2 attempts to shut an instance down
cleanly and run any system shutdown scripts; however, certain events (such as hardware failure) may
prevent these system shutdown scripts from running.

643
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Terminate

What happens when you terminate an instance

When an EC2 instance is terminated using the terminate-instances command, the following is
registered at the OS level:

• The API request will send a button press event to the guest.
• Various system services will be stopped as a result of the button press event. systemd handles a
graceful shutdown of the system. Graceful shutdown is triggered by the ACPI shutdown button press
event from the hypervisor.
• ACPI shutdown will be initiated.
• The instance will shut down when the graceful shutdown process exits. There is no configurable OS
shutdown time.

Terminate an instance
You can terminate an instance using the AWS Management Console or the command line.

By default, when you initiate a shutdown from an Amazon EBS-backed instance (using the shutdown or
poweroff commands), the instance stops. The halt command does not initiate a shutdown. If used, the
instance does not terminate; instead, it places the CPU into HLT and the instance remains running.

New console

To terminate an instance using the console

1. Before you terminate an instance, verify that you won't lose any data by checking that your
Amazon EBS volumes won't be deleted on termination and that you've copied any data that you
need from your instance store volumes to persistent storage, such as Amazon EBS or Amazon
S3.
2. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.
3. In the navigation pane, choose Instances.
4. Select the instance, and choose Actions, Instance state, Terminate instance.
5. Choose Terminate when prompted for confirmation.

Old console

To terminate an instance using the console

1. Before you terminate an instance, verify that you won't lose any data by checking that your
Amazon EBS volumes won't be deleted on termination and that you've copied any data that you
need from your instance store volumes to persistent storage, such as Amazon EBS or Amazon
S3.
2. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.
3. In the navigation pane, choose Instances.
4. Select the instance, and choose Actions, Instance State, Terminate.
5. Choose Yes, Terminate when prompted for confirmation.

To terminate an instance using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

644
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Terminate

• terminate-instances (AWS CLI)

• Stop-EC2Instance (AWS Tools for Windows PowerShell)

Enable termination protection

By default, you can terminate your instance using the Amazon EC2 console, command line interface, or
API. To prevent your instance from being accidentally terminated using Amazon EC2, you can enable
termination protection for the instance. The DisableApiTermination attribute controls whether the
instance can be terminated using the console, CLI, or API. By default, termination protection is disabled
for your instance. You can set the value of this attribute when you launch the instance, while the instance
is running, or while the instance is stopped (for Amazon EBS-backed instances).

The DisableApiTermination attribute does not prevent you from terminating an instance by
initiating shutdown from the instance (using an operating system command for system shutdown) when
the InstanceInitiatedShutdownBehavior attribute is set. For more information, see Change the
instance initiated shutdown behavior (p. 646).

Limitations

You can't enable termination protection for Spot Instances—a Spot Instance is terminated when the
Spot price exceeds the amount you're willing to pay for Spot Instances. However, you can prepare
your application to handle Spot Instance interruptions. For more information, see Spot Instance
interruptions (p. 452).

The DisableApiTermination attribute does not prevent Amazon EC2 Auto Scaling from terminating
an instance. For instances in an Auto Scaling group, use the following Amazon EC2 Auto Scaling features
instead of Amazon EC2 termination protection:

• To prevent instances that are part of an Auto Scaling group from terminating on scale in, use instance
protection. For more information, see Instance Protection in the Amazon EC2 Auto Scaling User Guide.
• To prevent Amazon EC2 Auto Scaling from terminating unhealthy instances, suspend the
ReplaceUnhealthy process. For more information, see Suspending and Resuming Scaling Processes
in the Amazon EC2 Auto Scaling User Guide.
• To specify which instances Amazon EC2 Auto Scaling should terminate first, choose a termination
policy. For more information, see Customizing the Termination Policy in the Amazon EC2 Auto Scaling
User Guide.

To enable termination protection for an instance at launch time

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. On the dashboard, choose Launch Instance and follow the directions in the wizard.
3. On the Configure Instance Details page, select the Enable termination protection check box.

To enable termination protection for a running or stopped instance

1. Select the instance, and choose Actions, Instance Settings, Change Termination Protection.
2. Choose Yes, Enable.

To disable termination protection for a running or stopped instance

1. Select the instance, and choose Actions, Instance Settings, Change Termination Protection.
2. Choose Yes, Disable.

645
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Terminate

To enable or disable termination protection using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• modify-instance-attribute (AWS CLI)

• Edit-EC2InstanceAttribute (AWS Tools for Windows PowerShell)

Change the instance initiated shutdown behavior

By default, when you initiate a shutdown from an Amazon EBS-backed instance (using a command such
as shutdown or poweroff), the instance stops (Note that halt does not issue a poweroff command and, if
used, the instance will not terminate; instead, it will place the CPU into HLT and the instance will remain
running). You can change this behavior using the InstanceInitiatedShutdownBehavior attribute
for the instance so that it terminates instead. You can update this attribute while the instance is running
or stopped.

You can update the InstanceInitiatedShutdownBehavior attribute using the Amazon EC2 console
or the command line. The InstanceInitiatedShutdownBehavior attribute only applies when you
perform a shutdown from the operating system of the instance itself; it does not apply when you stop an
instance using the StopInstances API or the Amazon EC2 console.

To change the shutdown behavior of an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance.
4. Choose Actions, Instance settings, Change shutdown behavior. The current behavior is selected.
5. To change the behavior, select Stop or Terminate from Shutdown behavior and then choose Apply.

To change the shutdown behavior of an instance using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• modify-instance-attribute (AWS CLI)

• Edit-EC2InstanceAttribute (AWS Tools for Windows PowerShell)

Preserve Amazon EBS volumes on instance termination

When an instance terminates, Amazon EC2 uses the value of the DeleteOnTermination attribute for
each attached Amazon EBS volume to determine whether to preserve or delete the volume.

The default value for the DeleteOnTermination attribute differs depending on whether the volume is
the root volume of the instance or a non-root volume attached to the instance.

Root volume

By default, the DeleteOnTermination attribute for the root volume of an instance is set to true.
Therefore, the default is to delete the root volume of the instance when the instance terminates. The
DeleteOnTermination attribute can be set by the creator of an AMI as well as by the person who
launches an instance. When the attribute is changed by the creator of an AMI or by the person who
launches an instance, the new setting overrides the original AMI default setting. We recommend that
you verify the default setting for the DeleteOnTermination attribute after you launch an instance
with an AMI.

646
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Terminate

Non-root volume

By default, when you attach a non-root EBS volume to an instance (p. 1106), its
DeleteOnTermination attribute is set to false. Therefore, the default is to preserve these
volumes. After the instance terminates, you can take a snapshot of the preserved volume or attach
it to another instance. You must delete a volume to avoid incurring further charges. For more
information, see Delete an Amazon EBS volume (p. 1126).

To verify the value of the DeleteOnTermination attribute for an EBS volume that is in use, look at the
instance's block device mapping. For more information, see View the EBS volumes in an instance block
device mapping (p. 1301).

You can change the value of the DeleteOnTermination attribute for a volume when you launch the
instance or while the instance is running.

Examples
• Change the root volume to persist at launch using the console (p. 647)
• Change the root volume to persist at launch using the command line (p. 647)
• Change the root volume of a running instance to persist using the command line (p. 648)

Change the root volume to persist at launch using the console

Using the console, you can change the DeleteOnTermination attribute when you launch an instance.
To change this attribute for a running instance, you must use the command line.

To change the root volume of an instance to persist at launch using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the console dashboard, select Launch Instance.
3. On the Choose an Amazon Machine Image (AMI) page, choose an AMI and choose Select.
4. Follow the wizard to complete the Choose an Instance Type and Configure Instance Details pages.
5. On the Add Storage page, deselect the Delete On Termination check box for the root volume.
6. Complete the remaining wizard pages, and then choose Launch.

You can verify the setting by viewing details for the root device volume on the instance's details pane.
Next to Block devices, choose the entry for the root device volume. By default, Delete on termination is
True. If you change the default behavior, Delete on termination is False.

Change the root volume to persist at launch using the command line
When you launch an EBS-backed instance, you can use one of the following commands to change the
root device volume to persist. For more information about these command line interfaces, see Access
Amazon EC2 (p. 3).

• run-instances (AWS CLI)

• New-EC2Instance (AWS Tools for Windows PowerShell)

For example, add the following option to your run-instances command:

--block-device-mappings file://mapping.json

Specify the following in mapping.json:

647
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Recover

[
{
"DeviceName": "/dev/sda1",
"Ebs": {
"DeleteOnTermination": false,
"SnapshotId": "snap-1234567890abcdef0",
"VolumeType": "gp2"
}
}
]

Change the root volume of a running instance to persist using the command line
You can use one of the following commands to change the root device volume of a running EBS-backed
instance to persist. For more information about these command line interfaces, see Access Amazon
EC2 (p. 3).

• modify-instance-attribute (AWS CLI)

• Edit-EC2InstanceAttribute (AWS Tools for Windows PowerShell)

For example, use the following command:

aws ec2 modify-instance-attribute --instance-id i-1234567890abcdef0 --block-device-mappings

file://mapping.json

Specify the following in mapping.json:

[
{
"DeviceName": "/dev/sda1",
"Ebs": {
"DeleteOnTermination": false
}
}
]

Troubleshoot instance termination

If you terminate your instance and another instance starts, most likely you have configured automatic
scaling through a feature like EC2 Fleet or Amazon EC2 Auto Scaling.

If your instance is in the shutting-down state for longer than usual, it should be cleaned up
(terminated) by automated processes within the Amazon EC2 service. For more information, see Delayed
instance termination (p. 1342).

Recover your instance

You can create an Amazon CloudWatch alarm that monitors an Amazon EC2 instance and automatically
recovers the instance if it becomes impaired due to an underlying hardware failure or a problem that
requires AWS involvement to repair. Terminated instances cannot be recovered. A recovered instance is
identical to the original instance, including the instance ID, private IP addresses, Elastic IP addresses, and
all instance metadata. If the impaired instance is in a placement group, the recovered instance runs in the
placement group. For more information about using Amazon CloudWatch alarms to recover an instance,
see Add recover actions to Amazon CloudWatch alarms (p. 789). To troubleshoot issues with instance
recovery failures, see Troubleshoot instance recovery failures (p. 649).

648
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Configure instances

When the StatusCheckFailed_System alarm is triggered, and the recover action is initiated, you
will be notified by the Amazon SNS topic that you selected when you created the alarm and associated
the recover action. During instance recovery, the instance is migrated during an instance reboot, and
any data that is in-memory is lost. When the process is complete, information is published to the SNS
topic you've configured for the alarm. Anyone who is subscribed to this SNS topic will receive an email
notification that includes the status of the recovery attempt and any further instructions. You will notice
an instance reboot on the recovered instance.

Examples of problems that cause system status checks to fail include:

• Loss of network connectivity

• Loss of system power
• Software issues on the physical host
• Hardware issues on the physical host that impact network reachability

If your instance has a public IPv4 address, it retains the public IPv4 address after recovery.

Requirements
The recover action is supported only on instances with the following characteristics:

• Uses one of the following instance types: A1, C3, C4, C5, C5a, C5n, C6g, C6gn, Inf1, M3, M4, M5, M5a,
M5n, M5zn, M6g, P3, P4, R3, R4, R5, R5a, R5b, R5n, R6g, T2, T3, T3a, T4g, X1, or X1e
• Runs in a virtual private cloud (VPC)
• Uses default or dedicated instance tenancy
• Has only EBS volumes (do not configure instance store volumes)

Troubleshoot instance recovery failures

The following issues can cause automatic recovery of your instance to fail:

• Temporary, insufficient capacity of replacement hardware.

• The instance has an attached instance store storage, which is an unsupported configuration for
automatic instance recovery.
• There is an ongoing Service Health Dashboard event that prevented the recovery process from
successfully executing. Refer to http://status.aws.amazon.com/ for the latest service availability
information.
• The instance has reached the maximum daily allowance of three recovery attempts.

The automatic recovery process attempts to recover your instance for up to three separate failures per
day. If the instance system status check failure persists, we recommend that you manually stop and start
the instance. For more information, see Stop and start your instance (p. 623).

Your instance may subsequently be retired if automatic recovery fails and a hardware degradation is
determined to be the root cause for the original system status check failure.

Configure your Amazon Linux instance

After you have successfully launched and logged into your Amazon Linux instance, you can make
changes to it. There are many different ways you can configure an instance to meet the needs of a
specific application. The following are some common tasks to help get you started.

649
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Common configuration scenarios

Contents
• Common configuration scenarios (p. 650)
• Manage software on your Amazon Linux instance (p. 650)
• Manage user accounts on your Amazon Linux instance (p. 656)
• Processor state control for your EC2 instance (p. 658)
• Set the time for your Linux instance (p. 664)
• Optimize CPU options (p. 668)
• Change the hostname of your Amazon Linux instance (p. 694)
• Set up dynamic DNS on Your Amazon Linux instance (p. 697)
• Run commands on your Linux instance at launch (p. 698)
• Instance metadata and user data (p. 704)

Common configuration scenarios

The base distribution of Amazon Linux contains many software packages and utilities that are required
for basic server operations. However, many more software packages are available in various software
repositories, and even more packages are available for you to build from source code. For more
information on installing and building software from these locations, see Manage software on your
Amazon Linux instance (p. 650).

Amazon Linux instances come pre-configured with an ec2-user account, but you may want to add
other user accounts that do not have super-user privileges. For more information on adding and
removing user accounts, see Manage user accounts on your Amazon Linux instance (p. 656).

The default time configuration for Amazon Linux instances uses Amazon Time Sync Service to set the
system time on an instance. The default time zone is UTC. For more information on setting the time zone
for an instance or using your own time server, see Set the time for your Linux instance (p. 664).

If you have your own network with a domain name registered to it, you can change the hostname
of an instance to identify itself as part of that domain. You can also change the system prompt to
show a more meaningful name without changing the hostname settings. For more information, see
Change the hostname of your Amazon Linux instance (p. 694). You can configure an instance to use
a dynamic DNS service provider. For more information, see Set up dynamic DNS on Your Amazon Linux
instance (p. 697).

When you launch an instance in Amazon EC2, you have the option of passing user data to the instance
that can be used to perform common configuration tasks and even run scripts after the instance starts.
You can pass two types of user data to Amazon EC2: cloud-init directives and shell scripts. For more
information, see Run commands on your Linux instance at launch (p. 698).

Manage software on your Amazon Linux instance

The base distribution of Amazon Linux contains many software packages and utilities that are required
for basic server operations. However, many more software packages are available in various software
repositories, and even more packages are available for you to build from source code.

Contents
• Update instance software on your Amazon Linux instance (p. 651)
• Add repositories on an Amazon Linux instance (p. 652)
• Find software packages on an Amazon Linux instance (p. 653)
• Install software packages on an Amazon Linux instance (p. 654)

650
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Manage software

• Prepare to compile software on an Amazon Linux instance (p. 655)

It is important to keep software up-to-date. Many packages in a Linux distribution are updated
frequently to fix bugs, add features, and protect against security exploits. For more information, see
Update instance software on your Amazon Linux instance (p. 651).

By default, Amazon Linux instances launch with the following repositories enabled:

• Amazon Linux 2: amzn2-core and amzn2extra-docker

• Amazon Linux AMI: amzn-main and amzn-updates

While there are many packages available in these repositories that are updated by Amazon Web Services,
there may be a package that you wish to install that is contained in another repository. For more
information, see Add repositories on an Amazon Linux instance (p. 652). For help finding packages
in enabled repositories, see Find software packages on an Amazon Linux instance (p. 653). For
information about installing software on an Amazon Linux instance, see Install software packages on an
Amazon Linux instance (p. 654).

Not all software is available in software packages stored in repositories; some software must be
compiled on an instance from its source code. For more information, see Prepare to compile software on
an Amazon Linux instance (p. 655).

Amazon Linux instances manage their software using the yum package manager. The yum package
manager can install, remove, and update software, as well as manage all of the dependencies for each
package. Debian-based Linux distributions, like Ubuntu, use the apt-get command and dpkg package
manager, so the yum examples in the following sections do not work for those distributions.

Update instance software on your Amazon Linux instance

It is important to keep software up-to-date. Many packages in a Linux distribution are updated
frequently to fix bugs, add features, and protect against security exploits. When you first launch and
connect to an Amazon Linux instance, you may see a message asking you to update software packages
for security purposes. This section shows how to update an entire system, or just a single package.
Important
This information applies to Amazon Linux. For information about other distributions, see their
specific documentation.

To update all packages on an Amazon Linux instance

1. (Optional) Start a screen session in your shell window. Sometimes you may experience a network
interruption that can disconnect the SSH connection to your instance. If this happens during a
long software update, it can leave the instance in a recoverable, although confused state. A screen
session allows you to continue running the update even if your connection is interrupted, and you
can reconnect to the session later without problems.

a. Execute the screen command to begin the session.

[ec2-user ~]$ screen

b. If your session is disconnected, log back into your instance and list the available screens.

[ec2-user ~]$ screen -ls

There is a screen on:
17793.pts-0.ip-12-34-56-78 (Detached)
1 Socket in /var/run/screen/S-ec2-user.

651
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Manage software

c. Reconnect to the screen using the screen -r command and the process ID from the previous
command.

[ec2-user ~]$ screen -r 17793

d. When you are finished using screen, use the exit command to close the session.

[ec2-user ~]$ exit

[screen is terminating]

2. Run the yum update command. Optionally, you can add the --security flag to apply only security
updates.

[ec2-user ~]$ sudo yum update

3. Review the packages listed, type y, and press Enter to accept the updates. Updating all of the
packages on a system can take several minutes. The yum output shows the status of the update
while it is running.
4. (Optional) Reboot your instance to ensure that you are using the latest packages and libraries from
your update; kernel updates are not loaded until a reboot occurs. Updates to any glibc libraries
should also be followed by a reboot. For updates to packages that control services, it might be
sufficient to restart the services to pick up the updates, but a system reboot ensures that all previous
package and library updates are complete.

To update a single package on an Amazon Linux instance

Use this procedure to update a single package (and its dependencies) and not the entire system.

1. Run the yum update command with the name of the package you would like to update.

[ec2-user ~]$ sudo yum update openssl

2. Review the package information listed, type y, and press Enter to accept the update or updates.
Sometimes there will be more than one package listed if there are package dependencies that must
be resolved. The yum output shows the status of the update while it is running.
3. (Optional) Reboot your instance to ensure that you are using the latest packages and libraries from
your update; kernel updates are not loaded until a reboot occurs. Updates to any glibc libraries
should also be followed by a reboot. For updates to packages that control services, it might be
sufficient to restart the services to pick up the updates, but a system reboot ensures that all previous
package and library updates are complete.

Add repositories on an Amazon Linux instance

By default, Amazon Linux instances launch with two repositories enabled: amzn-main and amzn-
updates. While there are many packages available in these repositories that are updated by Amazon
Web Services, there may be a package that you wish to install that is contained in another repository.
Important
This information applies to Amazon Linux. For information about other distributions, see their
specific documentation.

To install a package from a different repository with yum, you need to add the repository information
to the /etc/yum.conf file or to its own repository.repo file in the /etc/yum.repos.d directory.
You can do this manually, but most yum repositories provide their own repository.repo file at their
repository URL.

652
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Manage software

To determine what yum repositories are already installed

• List the installed yum repositories with the following command:

[ec2-user ~]$ yum repolist all

The resulting output lists the installed repositories and reports the status of each. Enabled
repositories display the number of packages they contain.

To add a yum repository to /etc/yum.repos.d

1. Find the location of the .repo file. This will vary depending on the repository you are adding. In this
example, the .repo file is at https://www.example.com/repository.repo.
2. Add the repository with the yum-config-manager command.

[ec2-user ~]$ sudo yum-config-manager --add-repo https://

www.example.com/repository.repo
Loaded plugins: priorities, update-motd, upgrade-helper
adding repo from: https://www.example.com/repository.repo
grabbing file https://www.example.com/repository.repo to /etc/
yum.repos.d/repository.repo
repository.repo | 4.0 kB 00:00
repo saved to /etc/yum.repos.d/repository.repo

After you install a repository, you must enable it as described in the next procedure.

To enable a yum repository in /etc/yum.repos.d

• Use the yum-config-manager command with the --enable repository flag. The following
command enables the Extra Packages for Enterprise Linux (EPEL) repository from the Fedora project.
By default, this repository is present in /etc/yum.repos.d on Amazon Linux AMI instances, but it
is not enabled.

[ec2-user ~]$ sudo yum-config-manager --enable epel

Note
To enable the EPEL repository on Amazon Linux 2, use the following command:

[ec2-user ~]$ sudo yum install https://dl.fedoraproject.org/pub/epel/epel-

release-latest-7.noarch.rpm

For information on enabling the EPEL repository on other distributions, such as Red Hat and
CentOS, see the EPEL documentation at https://fedoraproject.org/wiki/EPEL.

Find software packages on an Amazon Linux instance

You can use the yum search command to search the descriptions of packages that are available in your
configured repositories. This is especially helpful if you don't know the exact name of the package you
want to install. Simply append the keyword search to the command; for multiple word searches, wrap
the search query with quotation marks.
Important
This information applies to Amazon Linux. For information about other distributions, see their
specific documentation.

653
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Manage software

[ec2-user ~]$ sudo yum search "find"

The following is example output for Amazon Linux 2.

Loaded plugins: extras_suggestions, langpacks, priorities, update-motd

============================== N/S matched: find ===============================
findutils.x86_64 : The GNU versions of find utilities (find and xargs)
gedit-plugin-findinfiles.x86_64 : gedit findinfiles plugin
ocaml-findlib-devel.x86_64 : Development files for ocaml-findlib
perl-File-Find-Rule.noarch : Perl module implementing an alternative interface to
File::Find
robotfindskitten.x86_64 : A game/zen simulation. You are robot. Your job is to find kitten.
mlocate.x86_64 : An utility for finding files by name
ocaml-findlib.x86_64 : Objective CAML package manager and build helper
perl-Devel-Cycle.noarch : Find memory cycles in objects
perl-Devel-EnforceEncapsulation.noarch : Find access violations to blessed objects
perl-File-Find-Rule-Perl.noarch : Common rules for searching for Perl things
perl-File-HomeDir.noarch : Find your home and other directories on any platform
perl-IPC-Cmd.noarch : Finding and running system commands made easy
perl-Perl-MinimumVersion.noarch : Find a minimum required version of perl for Perl code
texlive-xesearch.noarch : A string finder for XeTeX
valgrind.x86_64 : Tool for finding memory management bugs in programs
valgrind.i686 : Tool for finding memory management bugs in programs

The following is example output for Amazon Linux.

Loaded plugins: priorities, security, update-motd, upgrade-helper

============================== N/S Matched: find ===============================
findutils.x86_64 : The GNU versions of find utilities (find and xargs)
perl-File-Find-Rule.noarch : Perl module implementing an alternative interface to
File::Find
perl-Module-Find.noarch : Find and use installed modules in a (sub)category
libpuzzle.i686 : Library to quickly find visually similar images (gif, png, jpg)
libpuzzle.x86_64 : Library to quickly find visually similar images (gif, png, jpg)
mlocate.x86_64 : An utility for finding files by name

Multiple word search queries in quotation marks only return results that match the exact query. If you
don't see the expected package, simplify your search to one keyword and then scan the results. You can
also try keyword synonyms to broaden your search.

For more information about packages for Amazon Linux 2 and Amazon Linux, see the following:

• Package repository (p. 181)

• Extras library (Amazon Linux 2) (p. 183)

Install software packages on an Amazon Linux instance

The yum package manager is a great tool for installing software, because it can search all of your
enabled repositories for different software packages and also handle any dependencies in the software
installation process.
Important
This information applies to Amazon Linux. For information about other distributions, see their
specific documentation.

To install a package from a repository, use the yum install package command, replacing package with
the name of the software to install. For example, to install the links text-based web browser, enter the
following command.

654
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Manage software

[ec2-user ~]$ sudo yum install links

You can also use yum install to install RPM package files that you have downloaded from the Internet.
To do this, simply append the path name of an RPM file to the installation command instead of a
repository package name.

[ec2-user ~]$ sudo yum install my-package.rpm

Prepare to compile software on an Amazon Linux instance

There is a wealth of open-source software available on the Internet that has not been pre-compiled
and made available for download from a package repository. You may eventually discover a software
package that you need to compile yourself, from its source code. For your system to be able to compile
software, you need to install several development tools, such as make, gcc, and autoconf.
Important
This information applies to Amazon Linux. For information about other distributions, see their
specific documentation.

Because software compilation is not a task that every Amazon EC2 instance requires, these tools are not
installed by default, but they are available in a package group called "Development Tools" that is easily
added to an instance with the yum groupinstall command.

[ec2-user ~]$ sudo yum groupinstall "Development Tools"

Software source code packages are often available for download (from web sites such as https://
github.com/ and http://sourceforge.net/) as a compressed archive file, called a tarball. These tarballs will
usually have the .tar.gz file extension. You can decompress these archives with the tar command.

[ec2-user ~]$ tar -xzf software.tar.gz

After you have decompressed and unarchived the source code package, you should look for a README or
INSTALL file in the source code directory that can provide you with further instructions for compiling
and installing the source code.

To retrieve source code for Amazon Linux packages

Amazon Web Services provides the source code for maintained packages. You can download the source
code for any installed packages with the yumdownloader --source command.

• Run the yumdownloader --source package command to download the source code for package.
For example, to download the source code for the htop package, enter the following command.

[ec2-user ~]$ yumdownloader --source htop

Loaded plugins: priorities, update-motd, upgrade-helper

Enabling amzn-updates-source repository
Enabling amzn-main-source repository
amzn-main-source
| 1.9 kB 00:00:00
amzn-updates-source
| 1.9 kB 00:00:00
(1/2): amzn-updates-source/latest/primary_db
| 52 kB 00:00:00
(2/2): amzn-main-source/latest/primary_db
| 734 kB 00:00:00
htop-1.0.1-2.3.amzn1.src.rpm

655
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Manage users

The location of the source RPM is in the directory from which you ran the command.

Manage user accounts on your Amazon Linux

instance
Each Linux instance launches with a default Linux system user account. The default user name is
determined by the AMI that was specified when you launched the instance.

• For Amazon Linux 2 or the Amazon Linux AMI, the user name is ec2-user.
• For a CentOS AMI, the user name is centos.
• For a Debian AMI, the user name is admin.
• For a Fedora AMI, the user name is ec2-user or fedora.
• For a RHEL AMI, the user name is ec2-user or root.
• For a SUSE AMI, the user name is ec2-user or root.
• For an Ubuntu AMI, the user name is ubuntu.
• Otherwise, if ec2-user and root don't work, check with the AMI provider.

Note
Linux system users should not be confused with AWS Identity and Access Management (IAM)
users. For more information, see IAM Users and Groups in the IAM User Guide.

Contents
• Considerations (p. 656)
• Create a user account (p. 656)
• Remove a user account (p. 658)

Considerations
Using the default user account is adequate for many applications. However, you may choose to add
user accounts so that individuals can have their own files and workspaces. Furthermore, creating user
accounts for new users is much more secure than granting multiple (possibly inexperienced) users access
to the default user account, because the default user account can cause a lot of damage to a system
when used improperly. For more information, see Tips for Securing Your EC2 Instance.

To enable users SSH access to your EC2 instance using a Linux system user account, you must share the
SSH key with the user. Alternatively, you can use EC2 Instance Connect to provide access to users without
the need to share and manage SSH keys. For more information, see Connect to your Linux instance using
EC2 Instance Connect (p. 604).

Create a user account

First create the user account, and then add the SSH public key that allows the user to connect to and log
into the instance.

To create a user account

1. Create a new key pair (p. 1045). You must provide the .pem file to the user for whom you are
creating the user account. They must use this file to connect to the instance.
2. Retrieve the public key from the key pair that you created in the previous step.

656
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Manage users

$ ssh-keygen -y -f /path_to_key_pair/key-pair-name.pem

The command returns the public key, as shown in the following example.

ssh-rsa
AAAAB3NzaC1yc2EAAAADAQABAAABAQClKsfkNkuSevGj3eYhCe53pcjqP3maAhDFcvBS7O6Vhz2ItxCih
+PnDSUaw+WNQn/mZphTk/a/gU8jEzoOWbkM4yxyb/wB96xbiFveSFJuOp/
d6RJhJOI0iBXrlsLnBItntckiJ7FbtxJMXLvvwJryDUilBMTjYtwB+QhYXUMOzce5Pjz5/i8SeJtjnV3iAoG/
cQk+0FzZqaeJAAHco+CY/5WrUBkrHmFJr6HcXkvJdWPkYQS3xqC0+FmUZofz221CBt5IMucxXPkX4rWi
+z7wB3RbBQoQzd8v7yeb7OzlPnWOyN0qFU0XA246RA8QFYiCNYwI3f05p6KLxEXAMPLE

3. Connect to the instance.

4. Use the adduser command to create the user account and add it to the system (with an entry in the
/etc/passwd file). The command also creates a group and a home directory for the account. In this
example, the user account is named newuser.

• Amazon Linux and Amazon Linux 2

[ec2-user ~]$ sudo adduser newuser

• Ubuntu

Include the --disabled-password parameter to create the user account without a password.

[ubuntu ~]$ sudo adduser newuser --disabled-password

5. Switch to the new account so that the directory and file that you create will have the proper
ownership.

[ec2-user ~]$ sudo su - newuser

The prompt changes from ec2-user to newuser to indicate that you have switched the shell
session to the new account.
6. Add the SSH public key to the user account. First create a directory in the user's home directory
for the SSH key file, then create the key file, and finally paste the public key into the key file, as
described in the following sub-steps.

a. Create a .ssh directory in the newuser home directory and change its file permissions to 700
(only the owner can read, write, or open the directory).

[newuser ~]$ mkdir .ssh

[newuser ~]$ chmod 700 .ssh

Important
Without these exact file permissions, the user will not be able to log in.
b. Create a file named authorized_keys in the .ssh directory and change its file permissions to
600 (only the owner can read or write to the file).

[newuser ~]$ touch .ssh/authorized_keys

[newuser ~]$ chmod 600 .ssh/authorized_keys

657
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Processor state control

Important
Without these exact file permissions, the user will not be able to log in.
c. Open the authorized_keys file using your favorite text editor (such as vim or nano).

[newuser ~]$ nano .ssh/authorized_keys

Paste the public key that you retrieved in Step 2 into the file and save the changes.
Important
Ensure that you paste the public key in one continuous line. The public key must not be
split over multiple lines.

The user should now be able to log into the newuser account on your instance, using the
private key that corresponds to the public key that you added to the authorized_keys file.
For more information about the different methods of connecting to a Linux instance, see
Connect to your Linux instance (p. 598).

Remove a user account

If a user account is no longer needed, you can remove that account so that it can no longer be used.

Use the userdel command to remove the user account from the system. When you specify the -r
parameter, the user's home directory and mail spool are deleted. To keep the user's home directory and
mail spool, omit the -r parameter.

[ec2-user ~]$ sudo userdel -r olduser

Processor state control for your EC2 instance

C-states control the sleep levels that a core can enter when it is idle. C-states are numbered starting
with C0 (the shallowest state where the core is totally awake and executing instructions) and go to C6
(the deepest idle state where a core is powered off). P-states control the desired performance (in CPU
frequency) from a core. P-states are numbered starting from P0 (the highest performance setting where
the core is allowed to use Intel Turbo Boost Technology to increase frequency if possible), and they
go from P1 (the P-state that requests the maximum baseline frequency) to P15 (the lowest possible
frequency).

The following instance types provide the ability for an operating system to control processor C-states
and P-states:

• General purpose: m4.10xlarge | m4.16xlarge | m5.metal | m5d.metal

• Compute optimized: c4.8xlarge | c5.metal | c5n.metal
• Memory optimized: r4.8xlarge | r4.16xlarge | r5.metal | r5d.metal | u-6tb1.metal
| u-9tb1.metal | u-12tb1.metal | u-18tb1.metal | u-24tb1.metal x1.16xlarge |
x1.32xlarge | x1e.8xlarge | x1e.16xlarge | x1e.32xlarge | z1d.metal
• Storage optimized: d2.8xlarge | i3.8xlarge | i3.16xlarge | i3.metal | i3en.metal |
h1.8xlarge | h1.16xlarge
• Accelerated computing: f1.16xlarge | g3.16xlarge | g4dn.metal | p2.16xlarge | p3.16xlarge

The following instance types provide the ability for an operating system to control processor C-states:

• General purpose: m5.12xlarge | m5.24xlarge | m5d.12xlarge | m5d.24xlarge | m5n.12xlarge |

m5n.24xlarge | m5dn.12xlarge | m5dn.24xlarge

658
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Processor state control

• Compute optimized: c5.9xlarge | c5.12xlarge | c5.18xlarge | c5.24xlarge | c5a.24xlarge |

c5ad.24xlarge | c5d.9xlarge | c5d.12xlarge | c5d.18xlarge | c5d.24xlarge | c5n.9xlarge
| c5n.18xlarge
• Memory optimized: r5.12xlarge | r5.24xlarge | r5d.12xlarge | r5d.24xlarge |
r5n.12xlarge | r5n.24xlarge | r5dn.12xlarge | r5dn.24xlarge | z1d.6xlarge |
z1d.12xlarge
• Storage optimized: d3en.12xlarge | i3en.12xlarge | i3en.24xlarge
• Accelerated computing: inf1.24xlarge | p3dn.24xlarge

AWS Graviton processors have built-in power saving modes and operate at a fixed frequency. Therefore,
they do not provide the ability for the operating system to control C-states and P-states.

You might want to change the C-state or P-state settings to increase processor performance consistency,
reduce latency, or tune your instance for a specific workload. The default C-state and P-state settings
provide maximum performance, which is optimal for most workloads. However, if your application would
benefit from reduced latency at the cost of higher single- or dual-core frequencies, or from consistent
performance at lower frequencies as opposed to bursty Turbo Boost frequencies, consider experimenting
with the C-state or P-state settings that are available to these instances.

The following sections describe the different processor state configurations and how to monitor the
effects of your configuration. These procedures were written for, and apply to Amazon Linux; however,
they may also work for other Linux distributions with a Linux kernel version of 3.9 or newer. For more
information about other Linux distributions and processor state control, see your system-specific
documentation.
Note
The examples on this page use the turbostat utility (which is available on Amazon Linux by
default) to display processor frequency and C-state information, and the stress command (which
can be installed by running sudo yum install -y stress) to simulate a workload.
If the output does not display the C-state information, include the --debug option in the
command (sudo turbostat --debug stress <options>).

Contents
• Highest performance with maximum Turbo Boost frequency (p. 659)
• High performance and low latency by limiting deeper C-states (p. 660)
• Baseline performance with the lowest variability (p. 662)

Highest performance with maximum Turbo Boost frequency

This is the default processor state control configuration for the Amazon Linux AMI, and it is
recommended for most workloads. This configuration provides the highest performance with lower
variability. Allowing inactive cores to enter deeper sleep states provides the thermal headroom required
for single or dual core processes to reach their maximum Turbo Boost potential.

The following example shows a c4.8xlarge instance with two cores actively performing work reaching
their maximum processor Turbo Boost frequency.

[ec2-user ~]$ sudo turbostat stress -c 2 -t 10

stress: info: [30680] dispatching hogs: 2 cpu, 0 io, 0 vm, 0 hdd
stress: info: [30680] successful run completed in 10s
pk cor CPU %c0 GHz TSC SMI %c1 %c3 %c6 %c7 %pc2 %pc3 %pc6 %pc7
Pkg_W RAM_W PKG_% RAM_%
5.54 3.44 2.90 0 9.18 0.00 85.28 0.00 0.00 0.00 0.00 0.00
94.04 32.70 54.18 0.00
0 0 0 0.12 3.26 2.90 0 3.61 0.00 96.27 0.00 0.00 0.00 0.00 0.00
48.12 18.88 26.02 0.00

659
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Processor state control

0 0 18 0.12 3.26 2.90 0 3.61

0 1 1 0.12 3.26 2.90 0 4.11 0.00 95.77 0.00
0 1 19 0.13 3.27 2.90 0 4.11
0 2 2 0.13 3.28 2.90 0 4.45 0.00 95.42 0.00
0 2 20 0.11 3.27 2.90 0 4.47
0 3 3 0.05 3.42 2.90 0 99.91 0.00 0.05 0.00
0 3 21 97.84 3.45 2.90 0 2.11
...
1 1 10 0.06 3.33 2.90 0 99.88 0.01 0.06 0.00
1 1 28 97.61 3.44 2.90 0 2.32
...
10.002556 sec

In this example, vCPUs 21 and 28 are running at their maximum Turbo Boost frequency because
the other cores have entered the C6 sleep state to save power and provide both power and thermal
headroom for the working cores. vCPUs 3 and 10 (each sharing a processor core with vCPUs 21 and 28)
are in the C1 state, waiting for instruction.

In the following example, all 18 cores are actively performing work, so there is no headroom for
maximum Turbo Boost, but they are all running at the "all core Turbo Boost" speed of 3.2 GHz.

[ec2-user ~]$ sudo turbostat stress -c 36 -t 10

stress: info: [30685] dispatching hogs: 36 cpu, 0 io, 0 vm, 0 hdd
stress: info: [30685] successful run completed in 10s
pk cor CPU %c0 GHz TSC SMI %c1 %c3 %c6 %c7 %pc2 %pc3 %pc6 %pc7
Pkg_W RAM_W PKG_% RAM_%
99.27 3.20 2.90 0 0.26 0.00 0.47 0.00 0.00 0.00 0.00 0.00
228.59 31.33 199.26 0.00
0 0 0 99.08 3.20 2.90 0 0.27 0.01 0.64 0.00 0.00 0.00 0.00 0.00
114.69 18.55 99.32 0.00
0 0 18 98.74 3.20 2.90 0 0.62
0 1 1 99.14 3.20 2.90 0 0.09 0.00 0.76 0.00
0 1 19 98.75 3.20 2.90 0 0.49
0 2 2 99.07 3.20 2.90 0 0.10 0.02 0.81 0.00
0 2 20 98.73 3.20 2.90 0 0.44
0 3 3 99.02 3.20 2.90 0 0.24 0.00 0.74 0.00
0 3 21 99.13 3.20 2.90 0 0.13
0 4 4 99.26 3.20 2.90 0 0.09 0.00 0.65 0.00
0 4 22 98.68 3.20 2.90 0 0.67
0 5 5 99.19 3.20 2.90 0 0.08 0.00 0.73 0.00
0 5 23 98.58 3.20 2.90 0 0.69
0 6 6 99.01 3.20 2.90 0 0.11 0.00 0.89 0.00
0 6 24 98.72 3.20 2.90 0 0.39
...

High performance and low latency by limiting deeper C-states

C-states control the sleep levels that a core may enter when it is inactive. You may want to control
C-states to tune your system for latency versus performance. Putting cores to sleep takes time, and
although a sleeping core allows more headroom for another core to boost to a higher frequency, it takes
time for that sleeping core to wake back up and perform work. For example, if a core that is assigned
to handle network packet interrupts is asleep, there may be a delay in servicing that interrupt. You can
configure the system to not use deeper C-states, which reduces the processor reaction latency, but that
in turn also reduces the headroom available to other cores for Turbo Boost.

A common scenario for disabling deeper sleep states is a Redis database application, which stores the
database in system memory for the fastest possible query response time.

To limit deeper sleep states on Amazon Linux 2

1. Open the /etc/default/grub file with your editor of choice.

660
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Processor state control

[ec2-user ~]$ sudo vim /etc/default/grub

2. Edit the GRUB_CMDLINE_LINUX_DEFAULT line and add the intel_idle.max_cstate=1 option to

set C1 as the deepest C-state for idle cores.

GRUB_CMDLINE_LINUX_DEFAULT="console=tty0 console=ttyS0,115200n8 net.ifnames=0

biosdevname=0 nvme_core.io_timeout=4294967295 intel_idle.max_cstate=1"
GRUB_TIMEOUT=0

3. Save the file and exit your editor.

4. Run the following command to rebuild the boot configuration.

[ec2-user ~]$ sudo grub2-mkconfig -o /boot/grub2/grub.cfg

5. Reboot your instance to enable the new kernel option.

[ec2-user ~]$ sudo reboot

To limit deeper sleep states on Amazon Linux AMI

1. Open the /boot/grub/grub.conf file with your editor of choice.

[ec2-user ~]$ sudo vim /boot/grub/grub.conf

2. Edit the kernel line of the first entry and add the intel_idle.max_cstate=1 option to set C1 as
the deepest C-state for idle cores.

# created by imagebuilder
default=0
timeout=1
hiddenmenu

title Amazon Linux 2014.09 (3.14.26-24.46.amzn1.x86_64)

root (hd0,0)
kernel /boot/vmlinuz-3.14.26-24.46.amzn1.x86_64 root=LABEL=/ console=ttyS0
intel_idle.max_cstate=1
initrd /boot/initramfs-3.14.26-24.46.amzn1.x86_64.img

3. Save the file and exit your editor.

4. Reboot your instance to enable the new kernel option.

[ec2-user ~]$ sudo reboot

The following example shows a c4.8xlarge instance with two cores actively performing work at the "all
core Turbo Boost" core frequency.

[ec2-user ~]$ sudo turbostat stress -c 2 -t 10

stress: info: [5322] dispatching hogs: 2 cpu, 0 io, 0 vm, 0 hdd
stress: info: [5322] successful run completed in 10s
pk cor CPU %c0 GHz TSC SMI %c1 %c3 %c6 %c7 %pc2 %pc3 %pc6 %pc7
Pkg_W RAM_W PKG_% RAM_%
5.56 3.20 2.90 0 94.44 0.00 0.00 0.00 0.00 0.00 0.00 0.00
131.90 31.11 199.47 0.00
0 0 0 0.03 2.08 2.90 0 99.97 0.00 0.00 0.00 0.00 0.00 0.00 0.00
67.23 17.11 99.76 0.00

661
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Processor state control

0 0 18 0.01 1.93 2.90 0 99.99

0 1 1 0.02 1.96 2.90 0 99.98 0.00 0.00 0.00
0 1 19 99.70 3.20 2.90 0 0.30
...
1 1 10 0.02 1.97 2.90 0 99.98 0.00 0.00 0.00
1 1 28 99.67 3.20 2.90 0 0.33
1 2 11 0.04 2.63 2.90 0 99.96 0.00 0.00 0.00
1 2 29 0.02 2.11 2.90 0 99.98
...

In this example, the cores for vCPUs 19 and 28 are running at 3.2 GHz, and the other cores are in the C1
C-state, awaiting instruction. Although the working cores are not reaching their maximum Turbo Boost
frequency, the inactive cores will be much faster to respond to new requests than they would be in the
deeper C6 C-state.

Baseline performance with the lowest variability

You can reduce the variability of processor frequency with P-states. P-states control the desired
performance (in CPU frequency) from a core. Most workloads perform better in P0, which requests
Turbo Boost. But you may want to tune your system for consistent performance rather than bursty
performance that can happen when Turbo Boost frequencies are enabled.

Intel Advanced Vector Extensions (AVX or AVX2) workloads can perform well at lower frequencies, and
AVX instructions can use more power. Running the processor at a lower frequency, by disabling Turbo
Boost, can reduce the amount of power used and keep the speed more consistent. For more information
about optimizing your instance configuration and workload for AVX, see http://www.intel.com/
content/dam/www/public/us/en/documents/white-papers/performance-xeon-e5-v3-advanced-vector-
extensions-paper.pdf.

This section describes how to limit deeper sleep states and disable Turbo Boost (by requesting the P1 P-
state) to provide low-latency and the lowest processor speed variability for these types of workloads.

To limit deeper sleep states and disable Turbo Boost on Amazon Linux 2

1. Open the /etc/default/grub file with your editor of choice.

[ec2-user ~]$ sudo vim /etc/default/grub

2. Edit the GRUB_CMDLINE_LINUX_DEFAULT line and add the intel_idle.max_cstate=1 option to

set C1 as the deepest C-state for idle cores.

GRUB_CMDLINE_LINUX_DEFAULT="console=tty0 console=ttyS0,115200n8 net.ifnames=0

biosdevname=0 nvme_core.io_timeout=4294967295 intel_idle.max_cstate=1"
GRUB_TIMEOUT=0

3. Save the file and exit your editor.

4. Run the following command to rebuild the boot configuration.

[ec2-user ~]$ grub2-mkconfig -o /boot/grub2/grub.cfg

5. Reboot your instance to enable the new kernel option.

[ec2-user ~]$ sudo reboot

6. When you need the low processor speed variability that the P1 P-state provides, run the following
command to disable Turbo Boost.

[ec2-user ~]$ sudo sh -c "echo 1 > /sys/devices/system/cpu/intel_pstate/no_turbo"

662
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Processor state control

7. When your workload is finished, you can re-enable Turbo Boost with the following command.

[ec2-user ~]$ sudo sh -c "echo 0 > /sys/devices/system/cpu/intel_pstate/no_turbo"

To limit deeper sleep states and disable Turbo Boost on Amazon Linux AMI

1. Open the /boot/grub/grub.conf file with your editor of choice.

[ec2-user ~]$ sudo vim /boot/grub/grub.conf

2. Edit the kernel line of the first entry and add the intel_idle.max_cstate=1 option to set C1 as
the deepest C-state for idle cores.

# created by imagebuilder
default=0
timeout=1
hiddenmenu

title Amazon Linux 2014.09 (3.14.26-24.46.amzn1.x86_64)

root (hd0,0)
kernel /boot/vmlinuz-3.14.26-24.46.amzn1.x86_64 root=LABEL=/ console=ttyS0
intel_idle.max_cstate=1
initrd /boot/initramfs-3.14.26-24.46.amzn1.x86_64.img

3. Save the file and exit your editor.

4. Reboot your instance to enable the new kernel option.

[ec2-user ~]$ sudo reboot

5. When you need the low processor speed variability that the P1 P-state provides, run the following
command to disable Turbo Boost.

[ec2-user ~]$ sudo sh -c "echo 1 > /sys/devices/system/cpu/intel_pstate/no_turbo"

6. When your workload is finished, you can re-enable Turbo Boost with the following command.

[ec2-user ~]$ sudo sh -c "echo 0 > /sys/devices/system/cpu/intel_pstate/no_turbo"

The following example shows a c4.8xlarge instance with two vCPUs actively performing work at the
baseline core frequency, with no Turbo Boost.

[ec2-user ~]$ sudo turbostat stress -c 2 -t 10

stress: info: [5389] dispatching hogs: 2 cpu, 0 io, 0 vm, 0 hdd
stress: info: [5389] successful run completed in 10s
pk cor CPU %c0 GHz TSC SMI %c1 %c3 %c6 %c7 %pc2 %pc3 %pc6 %pc7
Pkg_W RAM_W PKG_% RAM_%
5.59 2.90 2.90 0 94.41 0.00 0.00 0.00 0.00 0.00 0.00 0.00
128.48 33.54 200.00 0.00
0 0 0 0.04 2.90 2.90 0 99.96 0.00 0.00 0.00 0.00 0.00 0.00 0.00
65.33 19.02 100.00 0.00
0 0 18 0.04 2.90 2.90 0 99.96
0 1 1 0.05 2.90 2.90 0 99.95 0.00 0.00 0.00
0 1 19 0.04 2.90 2.90 0 99.96
0 2 2 0.04 2.90 2.90 0 99.96 0.00 0.00 0.00
0 2 20 0.04 2.90 2.90 0 99.96
0 3 3 0.05 2.90 2.90 0 99.95 0.00 0.00 0.00
0 3 21 99.95 2.90 2.90 0 0.05

663
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Set the time

...
1 1 28 99.92 2.90 2.90 0 0.08
1 2 11 0.06 2.90 2.90 0 99.94 0.00 0.00 0.00
1 2 29 0.05 2.90 2.90 0 99.95

The cores for vCPUs 21 and 28 are actively performing work at the baseline processor speed of 2.9
GHz, and all inactive cores are also running at the baseline speed in the C1 C-state, ready to accept
instructions.

Set the time for your Linux instance

A consistent and accurate time reference is crucial for many server tasks and processes. Most system logs
include a time stamp that you can use to determine when problems occur and in what order the events
take place. If you use the AWS CLI or an AWS SDK to make requests from your instance, these tools sign
requests on your behalf. If your instance's date and time are not set correctly, the date in the signature
may not match the date of the request, and AWS rejects the request.

Amazon provides the Amazon Time Sync Service, which is accessible from all EC2 instances, and is also
used by other AWS services. This service uses a fleet of satellite-connected and atomic reference clocks
in each Region to deliver accurate current time readings of the Coordinated Universal Time (UTC) global
standard through Network Time Protocol (NTP). The Amazon Time Sync Service automatically smooths
any leap seconds that are added to UTC.

The Amazon Time Sync Service is available through NTP at the 169.254.169.123 IP address for any
instance running in a VPC. Your instance does not require access to the internet, and you do not have
to configure your security group rules or your network ACL rules to allow access. The latest versions of
Amazon Linux 2 and Amazon Linux AMIs synchronize with the Amazon Time Sync Service by default.

Use the following procedures to configure the Amazon Time Sync Service on your instance using the
chrony client. Alternatively, you can use external NTP sources. For more information about NTP and
public time sources, see http://www.ntp.org/. An instance needs access to the internet for the external
NTP time sources to work.

Topics
• Configure the Amazon Time Sync Service on Amazon Linux AMI (p. 664)
• Configure the Amazon Time Sync Service on Ubuntu (p. 666)
• Configure the Amazon Time Sync Service on SUSE Linux (p. 667)
• Change the time zone on Amazon Linux (p. 667)

Configure the Amazon Time Sync Service on Amazon Linux AMI

Note
On Amazon Linux 2, the default chrony configuration is already set up to use the Amazon Time
Sync Service IP address.

With the Amazon Linux AMI, you must edit the chrony configuration file to add a server entry for the
Amazon Time Sync Service.

To configure your instance to use the Amazon Time Sync Service

1. Connect to your instance and uninstall the NTP service.

[ec2-user ~]$ sudo yum erase 'ntp*'

2. Install the chrony package.

664
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Set the time

[ec2-user ~]$ sudo yum install chrony

3. Open the /etc/chrony.conf file using a text editor (such as vim or nano). Verify that the file
includes the following line:

server 169.254.169.123 prefer iburst minpoll 4 maxpoll 4

If the line is present, then the Amazon Time Sync Service is already configured and you can go to the
next step. If not, add the line after any other server or pool statements that are already present in
the file, and save your changes.
4. Restart the chrony daemon (chronyd).

[ec2-user ~]$ sudo service chronyd restart

Starting chronyd: [ OK ]

Note
On RHEL and CentOS (up to version 6), the service name is chrony instead of chronyd.
5. Use the chkconfig command to configure chronyd to start at each system boot.

[ec2-user ~]$ sudo chkconfig chronyd on

6. Verify that chrony is using the 169.254.169.123 IP address to synchronize the time.

[ec2-user ~]$ chronyc sources -v

210 Number of sources = 7

.-- Source mode '^' = server, '=' = peer, '#' = local clock.
/ .- Source state '*' = current synced, '+' = combined , '-' = not combined,
| / '?' = unreachable, 'x' = time may be in error, '~' = time too variable.
|| .- xxxx [ yyyy ] +/- zzzz
|| Reachability register (octal) -. | xxxx = adjusted offset,
|| Log2(Polling interval) --. | | yyyy = measured offset,
|| \ | | zzzz = estimated error.
|| | | \
MS Name/IP address Stratum Poll Reach LastRx Last sample
===============================================================================
^* 169.254.169.123 3 6 17 43 -30us[ -226us] +/- 287us
^- ec2-12-34-231-12.eu-west> 2 6 17 43 -388us[ -388us] +/- 11ms
^- tshirt.heanet.ie 1 6 17 44 +178us[ +25us] +/- 1959us
^? tbag.heanet.ie 0 6 0 - +0ns[ +0ns] +/- 0ns
^? bray.walcz.net 0 6 0 - +0ns[ +0ns] +/- 0ns
^? 2a05:d018:c43:e312:ce77:> 0 6 0 - +0ns[ +0ns] +/- 0ns
^? 2a05:d018:dab:2701:b70:b> 0 6 0 - +0ns[ +0ns] +/- 0ns

In the output that's returned, ^* indicates the preferred time source.

7. Verify the time synchronization metrics that are reported by chrony.

[ec2-user ~]$ chronyc tracking

Reference ID : A9FEA97B (169.254.169.123)

Stratum : 4

665
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Set the time

Ref time (UTC) : Wed Nov 22 13:18:34 2017

System time : 0.000000626 seconds slow of NTP time
Last offset : +0.002852759 seconds
RMS offset : 0.002852759 seconds
Frequency : 1.187 ppm fast
Residual freq : +0.020 ppm
Skew : 24.388 ppm
Root delay : 0.000504752 seconds
Root dispersion : 0.001112565 seconds
Update interval : 64.4 seconds
Leap status : Normal

Configure the Amazon Time Sync Service on Ubuntu

You must edit the chrony configuration file to add a server entry for the Amazon Time Sync Service.

To configure your instance to use the Amazon Time Sync Service

1. Connect to your instance and use apt to install the chrony package.

ubuntu:~$ sudo apt install chrony

Note
If necessary, update your instance first by running sudo apt update.
2. Open the /etc/chrony/chrony.conf file using a text editor (such as vim or nano). Add the
following line before any other server or pool statements that are already present in the file, and
save your changes:

server 169.254.169.123 prefer iburst minpoll 4 maxpoll 4

3. Restart the chrony service.

ubuntu:~$ sudo /etc/init.d/chrony restart

[ ok ] Restarting chrony (via systemctl): chrony.service.

4. Verify that chrony is using the 169.254.169.123 IP address to synchronize the time.

ubuntu:~$ chronyc sources -v

210 Number of sources = 7

.-- Source mode '^' = server, '=' = peer, '#' = local clock.
/ .- Source state '*' = current synced, '+' = combined , '-' = not combined,
| / '?' = unreachable, 'x' = time may be in error, '~' = time too variable.
|| .- xxxx [ yyyy ] +/- zzzz
|| Reachability register (octal) -. | xxxx = adjusted offset,
|| Log2(Polling interval) --. | | yyyy = measured offset,
|| \ | | zzzz = estimated error.
|| | | \
MS Name/IP address Stratum Poll Reach LastRx Last sample
===============================================================================
^* 169.254.169.123 3 6 17 12 +15us[ +57us] +/- 320us
^- tbag.heanet.ie 1 6 17 13 -3488us[-3446us] +/- 1779us
^- ec2-12-34-231-12.eu-west- 2 6 17 13 +893us[ +935us] +/- 7710us
^? 2a05:d018:c43:e312:ce77:6 0 6 0 10y +0ns[ +0ns] +/- 0ns

666
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Set the time

^? 2a05:d018:d34:9000:d8c6:5 0 6 0 10y +0ns[ +0ns] +/- 0ns

^? tshirt.heanet.ie 0 6 0 10y +0ns[ +0ns] +/- 0ns
^? bray.walcz.net 0 6 0 10y +0ns[ +0ns] +/- 0ns

In the output that's returned, ^* indicates the preferred time source.

5. Verify the time synchronization metrics that are reported by chrony.

ubuntu:~$ chronyc tracking

Reference ID : 169.254.169.123 (169.254.169.123)

Stratum : 4
Ref time (UTC) : Wed Nov 29 07:41:57 2017
System time : 0.000000011 seconds slow of NTP time
Last offset : +0.000041659 seconds
RMS offset : 0.000041659 seconds
Frequency : 10.141 ppm slow
Residual freq : +7.557 ppm
Skew : 2.329 ppm
Root delay : 0.000544 seconds
Root dispersion : 0.000631 seconds
Update interval : 2.0 seconds
Leap status : Normal

Configure the Amazon Time Sync Service on SUSE Linux

Install chrony from https://software.opensuse.org/package/chrony.

Open the /etc/chrony.conf file using a text editor (such as vim or nano). Verify that the file contains
the following line:

server 169.254.169.123 prefer iburst minpoll 4 maxpoll 4

If this line is not present, add it. Comment out any other server or pool lines. Open yast and enable the
chrony service.

Change the time zone on Amazon Linux

Amazon Linux instances are set to the UTC (Coordinated Universal Time) time zone by default. You can
change the time on an instance to the local time or to another time zone in your network.
Important
This information applies to Amazon Linux. For information about other distributions, see their
specific documentation.

To change the time zone on an instance

1. Identify the time zone to use on the instance. The /usr/share/zoneinfo directory contains a
hierarchy of time zone data files. Browse the directory structure at that location to find a file for
your time zone.

[ec2-user ~]$ ls /usr/share/zoneinfo

Africa Chile GB Indian Mideast posixrules US
America CST6CDT GB-Eire Iran MST PRC UTC
Antarctica Cuba GMT iso3166.tab MST7MDT PST8PDT WET
Arctic EET GMT0 Israel Navajo right W-SU

667
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

...

Some of the entries at this location are directories (such as America), and these directories contain
time zone files for specific cities. Find your city (or a city in your time zone) to use for the instance.
2. Update the /etc/sysconfig/clock file with the new time zone. In this example, we use the time
zone data file for Los Angeles, /usr/share/zoneinfo/America/Los_Angeles.

a. Open the /etc/sysconfig/clock file with your favorite text editor (such as vim or nano).
You need to use sudo with your editor command because /etc/sysconfig/clock is owned
by root.

[ec2-user ~]$ sudo nano /etc/sysconfig/clock

b. Locate the ZONE entry, and change it to the time zone file (omitting the /usr/share/
zoneinfo section of the path). For example, to change to the Los Angeles time zone, change
the ZONE entry to the following:

ZONE="America/Los_Angeles"

Note
Do not change the UTC=true entry to another value. This entry is for the hardware
clock, and does not need to be adjusted when you're setting a different time zone on
your instance.
c. Save the file and exit the text editor.
3. Create a symbolic link between /etc/localtime and the time zone file so that the instance finds
the time zone file when it references local time information.

[ec2-user ~]$ sudo ln -sf /usr/share/zoneinfo/America/Los_Angeles /etc/localtime

4. Reboot the system to pick up the new time zone information in all services and applications.

[ec2-user ~]$ sudo reboot

5. (Optional) Confirm that the current time zone is updated to the new time zone by using the date
command. The current time zone appears in the output. In the following example, the current time
zone is PDT, which refers to the Los Angeles time zone.

[ec2-user ~]$ date

Sun Aug 16 05:45:16 PDT 2020

Optimize CPU options

Amazon EC2 instances support multithreading, which enables multiple threads to run concurrently on
a single CPU core. Each thread is represented as a virtual CPU (vCPU) on the instance. An instance has
a default number of CPU cores, which varies according to instance type. For example, an m5.xlarge
instance type has two CPU cores and two threads per core by default—four vCPUs in total.
Note
Each vCPU is a thread of a CPU core, except for T2 instances and instances powered by AWS
Graviton2 processors.

In most cases, there is an Amazon EC2 instance type that has a combination of memory and number
of vCPUs to suit your workloads. However, you can specify the following CPU options to optimize your
instance for specific workloads or business needs:

668
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

• Number of CPU cores: You can customize the number of CPU cores for the instance. You might do
this to potentially optimize the licensing costs of your software with an instance that has sufficient
amounts of RAM for memory-intensive workloads but fewer CPU cores.
• Threads per core: You can disable multithreading by specifying a single thread per CPU core. You
might do this for certain workloads, such as high performance computing (HPC) workloads.

You can specify these CPU options during instance launch. There is no additional or reduced charge
for specifying CPU options. You're charged the same as instances that are launched with default CPU
options.

Contents
• Rules for specifying CPU options (p. 669)
• CPU cores and threads per CPU core per instance type (p. 669)
• Specify CPU options for your instance (p. 692)
• View the CPU options for your instance (p. 693)

Rules for specifying CPU options

To specify the CPU options for your instance, be aware of the following rules:

• CPU options can only be specified during instance launch and cannot be modified after launch.
• When you launch an instance, you must specify both the number of CPU cores and threads per core in
the request. For example requests, see Specify CPU options for your instance (p. 692).
• The number of vCPUs for the instance is the number of CPU cores multiplied by the threads per core.
To specify a custom number of vCPUs, you must specify a valid number of CPU cores and threads per
core for the instance type. You cannot exceed the default number of vCPUs for the instance. For more
information, see CPU cores and threads per CPU core per instance type (p. 669).
• To disable multithreading, specify one thread per core.
• When you change the instance type (p. 311) of an existing instance, the CPU options automatically
change to the default CPU options for the new instance type.
• The specified CPU options persist after you stop, start, or reboot an instance.

CPU cores and threads per CPU core per instance type
The following tables list the instance types that support specifying CPU options.

Contents
• Accelerated computing instances (p. 670)
• Compute optimized instances (p. 672)
• General purpose instances (p. 677)
• Memory optimized instances (p. 683)
• Storage optimized instances (p. 690)

669
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Accelerated computing instances

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

f1.2xlarge 8 4 2 1, 2, 3, 4 1, 2

f1.4xlarge 16 8 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8

f1.16xlarge 64 32 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32

g3.4xlarge 16 8 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8

g3.8xlarge 32 16 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

g3.16xlarge 64 32 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32

g3s.xlarge 4 2 2 1, 2 1, 2

g4ad.4xlarge 16 8 2 2, 4, 8 1, 2

g4ad.8xlarge 32 16 2 2, 4, 8, 16 1, 2

g4ad.16xlarge 64 32 2 2, 4, 8, 16, 32 1, 2

g4dn.xlarge 4 2 2 1, 2 1, 2

g4dn.2xlarge 8 4 2 1, 2, 3, 4 1, 2

g4dn.4xlarge 16 8 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8

g4dn.8xlarge 32 16 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

g4dn.12xlarge 48 24 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24

g4dn.16xlarge 64 32 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32

inf1.xlarge 4 2 2 2 1, 2

670
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

inf1.2xlarge 8 4 2 2, 4 1, 2

inf1.6xlarge 24 12 2 2, 4, 6, 8, 10, 1, 2
12

inf1.24xlarge 96 48 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32, 34,
36, 38, 40, 42,
44, 46, 48

p2.xlarge 4 2 2 1, 2 1, 2

p2.8xlarge 32 16 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

p2.16xlarge 64 32 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32

p3.2xlarge 8 4 2 1, 2, 3, 4 1, 2

p3.8xlarge 32 16 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

p3.16xlarge 64 32 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32

p3dn.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

p4d.24xlarge 96 48 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32, 34,
36, 38, 40, 42,
44, 46, 48

671
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Compute optimized instances

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

c4.large 2 1 2 1 1, 2

c4.xlarge 4 2 2 1, 2 1, 2

c4.2xlarge 8 4 2 1, 2, 3, 4 1, 2

c4.4xlarge 16 8 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8

c4.8xlarge 36 18 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18

c5.large 2 1 2 1 1, 2

c5.xlarge 4 2 2 2 1, 2

c5.2xlarge 8 4 2 2, 4 1, 2

c5.4xlarge 16 8 2 2, 4, 6, 8 1, 2

c5.9xlarge 36 18 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18

c5.12xlarge 48 24 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24

c5.18xlarge 72 36 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36

c5.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

c5a.large 2 1 2 1 1, 2

c5a.xlarge 4 2 2 1, 2 1, 2

c5a.2xlarge 8 4 2 1, 2, 3, 4 1, 2

c5a.4xlarge 16 8 2 1, 2, 3, 4, 8 1, 2

c5a.8xlarge 32 16 2 1, 2, 3, 4, 8, 12, 1, 2
16

c5a.12xlarge 48 24 2 1, 2, 3, 4, 8, 12, 1, 2
16, 20, 24

672
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

c5a.16xlarge 64 32 2 1, 2, 3, 4, 8, 12, 1, 2
16, 20, 24, 28,
32

c5a.24xlarge 96 48 2 1, 2, 3, 4, 8, 12, 1, 2
16, 20, 24, 28,
32, 36, 40, 44,
48

c5ad.large 2 1 2 1 1, 2

c5ad.xlarge 4 2 2 1, 2 1, 2

c5ad.2xlarge 8 4 2 1, 2, 3, 4 1, 2

c5ad.4xlarge 16 8 2 1, 2, 3, 4, 8 1, 2

c5ad.8xlarge 32 16 2 1, 2, 3, 4, 8, 12, 1, 2
16

c5ad.12xlarge 48 24 2 1, 2, 3, 4, 8, 12, 1, 2
16, 20, 24

c5ad.16xlarge 64 32 2 1, 2, 3, 4, 8, 12, 1, 2
16, 20, 24, 28,
32

c5ad.24xlarge 96 48 2 1, 2, 3, 4, 8, 12, 1, 2
16, 20, 24, 28,
32, 36, 40, 44,
48

c5d.large 2 1 2 1 1, 2

c5d.xlarge 4 2 2 2 1, 2

c5d.2xlarge 8 4 2 2, 4 1, 2

c5d.4xlarge 16 8 2 2, 4, 6, 8 1, 2

c5d.9xlarge 36 18 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18

c5d.12xlarge 48 24 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24

c5d.18xlarge 72 36 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36

673
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

c5d.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

c5n.large 2 1 2 1 1, 2

c5n.xlarge 4 2 2 2 1, 2

c5n.2xlarge 8 4 2 2, 4 1, 2

c5n.4xlarge 16 8 2 2, 4, 6, 8 1, 2

c5n.9xlarge 36 18 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18

c5n.18xlarge 72 36 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36

c6g.medium 1 1 1 1 1

c6g.large 2 2 1 1, 2 1

c6g.xlarge 4 4 1 1, 2, 3, 4 1

c6g.2xlarge 8 8 1 1, 2, 3, 4, 5, 6, 1
7, 8

c6g.4xlarge 16 16 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

c6g.8xlarge 32 32 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32

674
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

c6g.12xlarge 48 48 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48

c6g.16xlarge 64 64 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48, 49, 50, 51,
52, 53, 54, 55,
56, 57, 58, 59,
60, 61, 62, 63,
64

c6gd.medium 1 1 1 1 1

c6gd.large 2 2 1 1, 2 1

c6gd.xlarge 4 4 1 1, 2, 3, 4 1

c6gd.2xlarge 8 8 1 1, 2, 3, 4, 5, 6, 1
7, 8

c6gd.4xlarge 16 16 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

c6gd.8xlarge 32 32 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32

675
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

c6gd.12xlarge 48 48 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48

c6gd.16xlarge 64 64 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48, 49, 50, 51,
52, 53, 54, 55,
56, 57, 58, 59,
60, 61, 62, 63,
64

c6gn.medium 1 1 1 1 1

c6gn.large 2 2 1 1, 2 1

c6gn.xlarge 4 4 1 1, 2, 3, 4 1

c6gn.2xlarge 8 8 1 1, 2, 3, 4, 5, 6, 1
7, 8

c6gn.4xlarge 16 16 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

c6gn.8xlarge 32 32 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32

676
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

c6gn.12xlarge 48 48 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48

c6gn.16xlarge 64 64 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48, 49, 50, 51,
52, 53, 54, 55,
56, 57, 58, 59,
60, 61, 62, 63,
64

General purpose instances

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

m4.large 2 1 2 1 1, 2

m4.xlarge 4 2 2 1, 2 1, 2

m4.2xlarge 8 4 2 1, 2, 3, 4 1, 2

m4.4xlarge 16 8 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8

m4.10xlarge 40 20 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20

m4.16xlarge 64 32 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,

677
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core
20, 22, 24, 26,
28, 30, 32

m5.large 2 1 2 1 1, 2

m5.xlarge 4 2 2 2 1, 2

m5.2xlarge 8 4 2 2, 4 1, 2

m5.4xlarge 16 8 2 2, 4, 6, 8 1, 2

m5.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

m5.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

m5.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

m5.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

m5a.large 2 1 2 1 1, 2

m5a.xlarge 4 2 2 2 1, 2

m5a.2xlarge 8 4 2 2, 4 1, 2

m5a.4xlarge 16 8 2 2, 4, 6, 8 1, 2

m5a.8xlarge 32 16 2 4, 6, 8, 10, 12, 1, 2

14, 16

m5a.12xlarge 48 24 2 6, 12, 18, 24 1, 2

m5a.16xlarge 64 32 2 8, 10, 12, 14, 1, 2

16, 18, 20, 22,
24, 26, 28, 30,
32

m5a.24xlarge 96 48 2 12, 18, 24, 36, 1, 2

48

m5ad.large 2 1 2 1 1, 2

m5ad.xlarge 4 2 2 2 1, 2

m5ad.2xlarge 8 4 2 2, 4 1, 2

678
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

m5ad.4xlarge 16 8 2 2, 4, 6, 8 1, 2

m5ad.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

m5ad.12xlarge 48 24 2 6, 12, 18, 24 1, 2

m5ad.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

m5ad.24xlarge 96 48 2 12, 18, 24, 36, 1, 2

48

m5d.large 2 1 2 1 1, 2

m5d.xlarge 4 2 2 2 1, 2

m5d.2xlarge 8 4 2 2, 4 1, 2

m5d.4xlarge 16 8 2 2, 4, 6, 8 1, 2

m5d.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

m5d.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

m5d.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

m5d.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

m5dn.large 2 1 2 1 1, 2

m5dn.xlarge 4 2 2 2 1, 2

m5dn.2xlarge 8 4 2 2, 4 1, 2

m5dn.4xlarge 16 8 2 2, 4, 6, 8 1, 2

m5dn.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

m5dn.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

679
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

m5dn.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

m5dn.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

m5n.large 2 1 2 1 1, 2

m5n.xlarge 4 2 2 2 1, 2

m5n.2xlarge 8 4 2 2, 4 1, 2

m5n.4xlarge 16 8 2 2, 4, 6, 8 1, 2

m5n.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

m5n.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

m5n.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

m5n.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

m5zn.large 2 1 2 1 1, 2

m5zn.xlarge 4 2 2 1, 2 1, 2

m5zn.2xlarge 8 4 2 2, 4 1, 2

m5zn.3xlarge 12 6 2 2, 4, 6 1, 2

m5zn.6xlarge 24 12 2 2, 4, 6, 8, 10, 1, 2
12

m5zn.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

m6g.medium 1 1 1 1 1

680
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

m6g.large 2 2 1 1, 2 1

m6g.xlarge 4 4 1 1, 2, 3, 4 1

m6g.2xlarge 8 8 1 1, 2, 3, 4, 5, 6, 1
7, 8

m6g.4xlarge 16 16 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

m6g.8xlarge 32 32 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32

m6g.12xlarge 48 48 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48

m6g.16xlarge 64 64 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48, 49, 50, 51,
52, 53, 54, 55,
56, 57, 58, 59,
60, 61, 62, 63,
64

m6gd.medium 1 1 1 1 1

m6gd.large 2 2 1 1, 2 1

681
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

m6gd.xlarge 4 4 1 1, 2, 3, 4 1

m6gd.2xlarge 8 8 1 1, 2, 3, 4, 5, 6, 1
7, 8

m6gd.4xlarge 16 16 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

m6gd.8xlarge 32 32 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32

m6gd.12xlarge 48 48 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48

m6gd.16xlarge 64 64 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48, 49, 50, 51,
52, 53, 54, 55,
56, 57, 58, 59,
60, 61, 62, 63,
64

t2.nano 1 1 1 1 1

t2.micro 1 1 1 1 1

t2.small 1 1 1 1 1

682
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

t2.medium 2 2 1 1, 2 1

t2.large 2 2 1 1, 2 1

t2.xlarge 4 4 1 1, 2, 3, 4 1

t2.2xlarge 8 8 1 1, 2, 3, 4, 5, 6, 1
7, 8

t3.nano 2 1 2 1 1, 2

t3.micro 2 1 2 1 1, 2

t3.small 2 1 2 1 1, 2

t3.medium 2 1 2 1 1, 2

t3.large 2 1 2 1 1, 2

t3.xlarge 4 2 2 2 1, 2

t3.2xlarge 8 4 2 2, 4 1, 2

t3a.nano 2 1 2 1 1, 2

t3a.micro 2 1 2 1 1, 2

t3a.small 2 1 2 1 1, 2

t3a.medium 2 1 2 1 1, 2

t3a.large 2 1 2 1 1, 2

t3a.xlarge 4 2 2 2 1, 2

t3a.2xlarge 8 4 2 2, 4 1, 2

Memory optimized instances

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

r4.large 2 1 2 1 1, 2

r4.xlarge 4 2 2 1, 2 1, 2

r4.2xlarge 8 4 2 1, 2, 3, 4 1, 2

r4.4xlarge 16 8 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8

r4.8xlarge 32 16 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

683
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

r4.16xlarge 64 32 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32

r5.large 2 1 2 1 1, 2

r5.xlarge 4 2 2 2 1, 2

r5.2xlarge 8 4 2 2, 4 1, 2

r5.4xlarge 16 8 2 2, 4, 6, 8 1, 2

r5.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

r5.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

r5.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

r5.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

r5a.large 2 1 2 1 1, 2

r5a.xlarge 4 2 2 2 1, 2

r5a.2xlarge 8 4 2 2, 4 1, 2

r5a.4xlarge 16 8 2 2, 4, 6, 8 1, 2

r5a.8xlarge 32 16 2 4, 6, 8, 10, 12, 1, 2

14, 16

r5a.12xlarge 48 24 2 6, 12, 18, 24 1, 2

r5a.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

r5a.24xlarge 96 48 2 12, 18, 24, 36, 1, 2

48

r5ad.large 2 1 2 1 1, 2

r5ad.xlarge 4 2 2 2 1, 2

684
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

r5ad.2xlarge 8 4 2 2, 4 1, 2

r5ad.4xlarge 16 8 2 2, 4, 6, 8 1, 2

r5ad.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

r5ad.12xlarge 48 24 2 6, 12, 18, 24 1, 2

r5ad.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

r5ad.24xlarge 96 48 2 12, 18, 24, 36, 1, 2

48

r5b.large 2 1 2 1 1, 2

r5b.xlarge 4 2 2 2 1, 2

r5b.2xlarge 8 4 2 2, 4 1, 2

r5b.4xlarge 16 8 2 2, 4, 6, 8 1, 2

r5b.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

r5b.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

r5b.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

r5b.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

r5d.large 2 1 2 1 1, 2

r5d.xlarge 4 2 2 2 1, 2

r5d.2xlarge 8 4 2 2, 4 1, 2

r5d.4xlarge 16 8 2 2, 4, 6, 8 1, 2

r5d.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

685
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

r5d.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

r5d.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

r5d.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

r5dn.large 2 1 2 1 1, 2

r5dn.xlarge 4 2 2 2 1, 2

r5dn.2xlarge 8 4 2 2, 4 1, 2

r5dn.4xlarge 16 8 2 2, 4, 6, 8 1, 2

r5dn.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

r5dn.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

r5dn.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

r5dn.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

r5n.large 2 1 2 1 1, 2

r5n.xlarge 4 2 2 2 1, 2

r5n.2xlarge 8 4 2 2, 4 1, 2

r5n.4xlarge 16 8 2 2, 4, 6, 8 1, 2

r5n.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

686
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

r5n.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

r5n.16xlarge 64 32 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32

r5n.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

r6g.medium 1 1 1 1 1

r6g.large 2 2 1 1, 2 1

r6g.xlarge 4 4 1 1, 2, 3, 4 1

r6g.2xlarge 8 8 1 1, 2, 3, 4, 5, 6, 1
7, 8

r6g.4xlarge 16 16 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

r6g.8xlarge 32 32 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32

r6g.12xlarge 48 48 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48

687
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

r6g.16xlarge 64 64 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48, 49, 50, 51,
52, 53, 54, 55,
56, 57, 58, 59,
60, 61, 62, 63,
64

r6gd.medium 1 1 1 1 1

r6gd.large 2 2 1 1, 2 1

r6gd.xlarge 4 4 1 1, 2, 3, 4 1

r6gd.2xlarge 8 8 1 1, 2, 3, 4, 5, 6, 1
7, 8

r6gd.4xlarge 16 16 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

r6gd.8xlarge 32 32 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32

r6gd.12xlarge 48 48 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48

688
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

r6gd.16xlarge 64 64 1 1, 2, 3, 4, 5, 6, 1
7, 8, 9, 10, 11,
12, 13, 14, 15,
16, 17, 18, 19,
20, 21, 22, 23,
24, 25, 26, 27,
28, 29, 30, 31,
32, 33, 34, 35,
36, 37, 38, 39,
40, 41, 42, 43,
44, 45, 46, 47,
48, 49, 50, 51,
52, 53, 54, 55,
56, 57, 58, 59,
60, 61, 62, 63,
64

x1.16xlarge 64 32 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32

x1.32xlarge 128 64 2 4, 8, 12, 16, 20, 1, 2

24, 28, 32, 36,
40, 44, 48, 52,
56, 60, 64

x1e.xlarge 4 2 2 1, 2 1, 2

x1e.2xlarge 8 4 2 1, 2, 3, 4 1, 2

x1e.4xlarge 16 8 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8

x1e.8xlarge 32 16 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

x1e.16xlarge 64 32 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32

x1e.32xlarge 128 64 2 4, 8, 12, 16, 20, 1, 2

24, 28, 32, 36,
40, 44, 48, 52,
56, 60, 64

z1d.large 2 1 2 1 1, 2

z1d.xlarge 4 2 2 2 1, 2

z1d.2xlarge 8 4 2 2, 4 1, 2

689
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

z1d.3xlarge 12 6 2 2, 4, 6 1, 2

z1d.6xlarge 24 12 2 2, 4, 6, 8, 10, 1, 2
12

z1d.12xlarge 48 24 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24

Storage optimized instances

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

d2.xlarge 4 2 2 1, 2 1, 2

d2.2xlarge 8 4 2 1, 2, 3, 4 1, 2

d2.4xlarge 16 8 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8

d2.8xlarge 36 18 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18

d3.xlarge 4 2 2 1, 2 1, 2

d3.2xlarge 8 4 2 2, 4 1, 2

d3.4xlarge 16 8 2 2, 4, 6, 8 1, 2

d3.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

d3en.large 2 1 2 1 1, 2

d3en.xlarge 4 2 2 1, 2 1, 2

d3en.2xlarge 8 4 2 2, 4 1, 2

d3en.4xlarge 16 8 2 2, 4, 6, 8 1, 2

d3en.6xlarge 24 12 2 2, 4, 6, 8, 10, 1, 2
12

d3en.8xlarge 32 16 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16

d3en.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

h1.2xlarge 8 4 2 1, 2, 3, 4 1, 2

690
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Instance type Default vCPUs Default CPU Default Valid CPU Valid threads
cores threads per cores per core
core

h1.4xlarge 16 8 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8

h1.8xlarge 32 16 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

h1.16xlarge 64 32 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32

i3.large 2 1 2 1 1, 2

i3.xlarge 4 2 2 1, 2 1, 2

i3.2xlarge 8 4 2 1, 2, 3, 4 1, 2

i3.4xlarge 16 8 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8

i3.8xlarge 32 16 2 1, 2, 3, 4, 5, 6, 1, 2
7, 8, 9, 10, 11,
12, 13, 14, 15,
16

i3.16xlarge 64 32 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24, 26,
28, 30, 32

i3en.large 2 1 2 1 1, 2

i3en.xlarge 4 2 2 2 1, 2

i3en.2xlarge 8 4 2 2, 4 1, 2

i3en.3xlarge 12 6 2 2, 4, 6 1, 2

i3en.6xlarge 24 12 2 2, 4, 6, 8, 10, 1, 2
12

i3en.12xlarge 48 24 2 2, 4, 6, 8, 10, 1, 2
12, 14, 16, 18,
20, 22, 24

i3en.24xlarge 96 48 2 4, 6, 8, 10, 12, 1, 2

14, 16, 18, 20,
22, 24, 26, 28,
30, 32, 34, 36,
38, 40, 42, 44,
46, 48

691
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

Specify CPU options for your instance

You can specify CPU options during instance launch. The following examples are for an r4.4xlarge
instance type, which has the following default values (p. 683):

• Default CPU cores: 8

• Default threads per core: 2
• Default vCPUs: 16 (8 * 2)
• Valid number of CPU cores: 1, 2, 3, 4, 5, 6, 7, 8
• Valid number of threads per core: 1, 2

Disable multithreading
To disable multithreading, specify one thread per core.

To disable multithreading during instance launch (console)

1. Follow the Launch an instance using the Launch Instance Wizard (p. 526) procedure.
2. On the Configure Instance Details page, for CPU options, choose Specify CPU options.
3. For Core count, choose the number of required CPU cores. In this example, to specify the default
CPU core count for an r4.4xlarge instance, choose 8.
4. To disable multithreading, for Threads per core, choose 1.
5. Continue as prompted by the wizard. When you've finished reviewing your options on the Review
Instance Launch page, choose Launch. For more information, see Launch an instance using the
Launch Instance Wizard (p. 526).

To disable multithreading during instance launch (AWS CLI)

Use the run-instances AWS CLI command and specify a value of 1 for ThreadsPerCore for the --cpu-
options parameter. For CoreCount, specify the number of CPU cores. In this example, to specify the
default CPU core count for an r4.4xlarge instance, specify a value of 8.

aws ec2 run-instances --image-id ami-1a2b3c4d --instance-type r4.4xlarge --cpu-options

"CoreCount=8,ThreadsPerCore=1" --key-name MyKeyPair

Specify a custom number of vCPUs

You can customize the number of CPU cores and threads per core for the instance.

To specify a custom number of vCPUs during instance launch (console)

The following example launches an r4.4xlarge instance with six vCPUs.

1. Follow the Launch an instance using the Launch Instance Wizard (p. 526) procedure.
2. On the Configure Instance Details page, for CPU options, choose Specify CPU options.
3. To get six vCPUs, specify three CPU cores and two threads per core, as follows:

• For Core count, choose 3.

• For Threads per core, choose 2.
4. Continue as prompted by the wizard. When you've finished reviewing your options on the Review
Instance Launch page, choose Launch. For more information, see Launch an instance using the
Launch Instance Wizard (p. 526).

692
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize CPU options

To specify a custom number of vCPUs during instance launch (AWS CLI)

The following example launches an r4.4xlarge instance with six vCPUs.

Use the run-instances AWS CLI command and specify the number of CPU cores and number of threads
in the --cpu-options parameter. You can specify three CPU cores and two threads per core to get six
vCPUs.

aws ec2 run-instances --image-id ami-1a2b3c4d --instance-type r4.4xlarge --cpu-options

"CoreCount=3,ThreadsPerCore=2" --key-name MyKeyPair

Alternatively, specify six CPU cores and one thread per core (disable multithreading) to get six vCPUs:

aws ec2 run-instances --image-id ami-1a2b3c4d --instance-type r4.4xlarge --cpu-options

"CoreCount=6,ThreadsPerCore=1" --key-name MyKeyPair

View the CPU options for your instance

You can view the CPU options for an existing instance in the Amazon EC2 console or by describing the
instance using the AWS CLI.

New console

To view the CPU options for an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the left navigation pane, choose Instances and select the instance.
3. On the Details tab, under Host and placement group, find Number of vCPUs.
4. To view core count and threads per core, choose the value for Number of vCPUs.

Old console

To view the CPU options for an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the left navigation pane, choose Instances and select the instance.
3. Choose Description and find Number of vCPUs.
4. To view core count and threads per core, choose the value for Number of vCPUs.

To view the CPU options for an instance (AWS CLI)

Use the describe-instances command.

aws ec2 describe-instances --instance-ids i-123456789abcde123

...
"Instances": [
{
"Monitoring": {
"State": "disabled"
},
"PublicDnsName": "ec2-198-51-100-5.eu-central-1.compute.amazonaws.com",
"State": {
"Code": 16,
"Name": "running"
},

693
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the hostname

"EbsOptimized": false,
"LaunchTime": "2018-05-08T13:40:33.000Z",
"PublicIpAddress": "198.51.100.5",
"PrivateIpAddress": "172.31.2.206",
"ProductCodes": [],
"VpcId": "vpc-1a2b3c4d",
"CpuOptions": {
"CoreCount": 34,
"ThreadsPerCore": 1
},
"StateTransitionReason": "",
...
}
]
...

In the output that's returned, the CoreCount field indicates the number of cores for the instance. The
ThreadsPerCore field indicates the number of threads per core.

Alternatively, connect to your instance and use a tool such as lscpu to view the CPU information for your
instance.

You can use AWS Config to record, assess, audit, and evaluate configuration changes for instances,
including terminated instances. For more information, see Getting Started with AWS Config in the AWS
Config Developer Guide.

Change the hostname of your Amazon Linux instance

When you launch an instance, it is assigned a hostname that is a form of the private, internal IPv4
address. A typical Amazon EC2 private DNS name looks something like this: ip-12-34-56-78.us-
west-2.compute.internal, where the name consists of the internal domain, the service (in this case,
compute), the region, and a form of the private IPv4 address. Part of this hostname is displayed at the
shell prompt when you log into your instance (for example, ip-12-34-56-78). Each time you stop and
restart your Amazon EC2 instance (unless you are using an Elastic IP address), the public IPv4 address
changes, and so does your public DNS name, system hostname, and shell prompt.
Important
This information applies to Amazon Linux. For information about other distributions, see their
specific documentation.

Change the system hostname

If you have a public DNS name registered for the IP address of your instance (such as
webserver.mydomain.com), you can set the system hostname so your instance identifies itself as a
part of that domain. This also changes the shell prompt so that it displays the first portion of this name
instead of the hostname supplied by AWS (for example, ip-12-34-56-78). If you do not have a public
DNS name registered, you can still change the hostname, but the process is a little different.

To change the system hostname to a public DNS name

Follow this procedure if you already have a public DNS name registered.

1. • For Amazon Linux 2: Use the hostnamectl command to set your hostname to reflect the fully
qualified domain name (such as webserver.mydomain.com).

[ec2-user ~]$ sudo hostnamectl set-hostname webserver.mydomain.com

• For Amazon Linux AMI: On your instance, open the /etc/sysconfig/network configuration file
in your favorite text editor and change the HOSTNAME entry to reflect the fully qualified domain
name (such as webserver.mydomain.com).

694
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the hostname

HOSTNAME=webserver.mydomain.com

2. Reboot the instance to pick up the new hostname.

[ec2-user ~]$ sudo reboot

Alternatively, you can reboot using the Amazon EC2 console (on the Instances page, select the
instance and choose Instance state, Reboot instance).
3. Log into your instance and verify that the hostname has been updated. Your prompt should show
the new hostname (up to the first ".") and the hostname command should show the fully-qualified
domain name.

[ec2-user@webserver ~]$ hostname

webserver.mydomain.com

To change the system hostname without a public DNS name

1. • For Amazon Linux 2: Use the hostnamectl command to set your hostname to reflect the desired
system hostname (such as webserver).

[ec2-user ~]$ sudo hostnamectl set-hostname webserver.localdomain

• For Amazon Linux AMI: On your instance, open the /etc/sysconfig/network configuration
file in your favorite text editor and change the HOSTNAME entry to reflect the desired system
hostname (such as webserver).

HOSTNAME=webserver.localdomain

2. Open the /etc/hosts file in your favorite text editor and change the entry beginning with
127.0.0.1 to match the example below, substituting your own hostname.

127.0.0.1 webserver.localdomain webserver localhost4 localhost4.localdomain4

3. Reboot the instance to pick up the new hostname.

[ec2-user ~]$ sudo reboot

Alternatively, you can reboot using the Amazon EC2 console (on the Instances page, select the
instance and choose Instance state, Reboot instance).
4. Log into your instance and verify that the hostname has been updated. Your prompt should show
the new hostname (up to the first ".") and the hostname command should show the fully-qualified
domain name.

[ec2-user@webserver ~]$ hostname

webserver.localdomain

Change the shell prompt without affecting the hostname

If you do not want to modify the hostname for your instance, but you would like to have a more useful
system name (such as webserver) displayed than the private name supplied by AWS (for example,

695
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the hostname

ip-12-34-56-78), you can edit the shell prompt configuration files to display your system nickname
instead of the hostname.

To change the shell prompt to a host nickname

1. Create a file in /etc/profile.d that sets the environment variable called NICKNAME to the value
you want in the shell prompt. For example, to set the system nickname to webserver, run the
following command.

[ec2-user ~]$ sudo sh -c 'echo "export NICKNAME=webserver" > /etc/profile.d/prompt.sh'

2. Open the /etc/bashrc (Red Hat) or /etc/bash.bashrc (Debian/Ubuntu) file in your favorite text
editor (such as vim or nano). You need to use sudo with the editor command because /etc/bashrc
and /etc/bash.bashrc are owned by root.
3. Edit the file and change the shell prompt variable (PS1) to display your nickname instead of
the hostname. Find the following line that sets the shell prompt in /etc/bashrc or /etc/
bash.bashrc (several surrounding lines are shown below for context; look for the line that starts
with [ "$PS1"):

# Turn on checkwinsize
shopt -s checkwinsize
[ "$PS1" = "\\s-\\v\\\$ " ] && PS1="[\u@\h \W]\\$ "
# You might want to have e.g. tty in prompt (e.g. more virtual machines)
# and console windows

Change the \h (the symbol for hostname) in that line to the value of the NICKNAME variable.

# Turn on checkwinsize
shopt -s checkwinsize
[ "$PS1" = "\\s-\\v\\\$ " ] && PS1="[\u@$NICKNAME \W]\\$ "
# You might want to have e.g. tty in prompt (e.g. more virtual machines)
# and console windows

4. (Optional) To set the title on shell windows to the new nickname, complete the following steps.

a. Create a file named /etc/sysconfig/bash-prompt-xterm.

[ec2-user ~]$ sudo touch /etc/sysconfig/bash-prompt-xterm

b. Make the file executable using the following command.

[ec2-user ~]$ sudo chmod +x /etc/sysconfig/bash-prompt-xterm

c. Open the /etc/sysconfig/bash-prompt-xterm file in your favorite text editor (such as vim
or nano). You need to use sudo with the editor command because /etc/sysconfig/bash-
prompt-xterm is owned by root.
d. Add the following line to the file.

echo -ne "\033]0;${USER}@${NICKNAME}:${PWD/#$HOME/~}\007"

5. Log out and then log back in to pick up the new nickname value.

Change the hostname on other Linux distributions

The procedures on this page are intended for use with Amazon Linux only. For more information about
other Linux distributions, see their specific documentation and the following articles:

696
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Set up dynamic DNS

• How do I assign a static hostname to a private Amazon EC2 instance running RHEL 7 or Centos 7?

Set up dynamic DNS on Your Amazon Linux instance

When you launch an EC2 instance, it is assigned a public IP address and a public DNS (Domain Name
System) name that you can use to reach it from the Internet. Because there are so many hosts in
the Amazon Web Services domain, these public names must be quite long for each name to remain
unique. A typical Amazon EC2 public DNS name looks something like this: ec2-12-34-56-78.us-
west-2.compute.amazonaws.com, where the name consists of the Amazon Web Services domain, the
service (in this case, compute), the region, and a form of the public IP address.

Dynamic DNS services provide custom DNS host names within their domain area that can be easy to
remember and that can also be more relevant to your host's use case; some of these services are also free
of charge. You can use a dynamic DNS provider with Amazon EC2 and configure the instance to update
the IP address associated with a public DNS name each time the instance starts. There are many different
providers to choose from, and the specific details of choosing a provider and registering a name with
them are outside the scope of this guide.
Important
This information applies to Amazon Linux. For information about other distributions, see their
specific documentation.

To use dynamic DNS with Amazon EC2

1. Sign up with a dynamic DNS service provider and register a public DNS name with their service. This
procedure uses the free service from noip.com/free as an example.
2. Configure the dynamic DNS update client. After you have a dynamic DNS service provider and
a public DNS name registered with their service, point the DNS name to the IP address for your
instance. Many providers (including noip.com) allow you to do this manually from your account page
on their website, but many also support software update clients. If an update client is running on
your EC2 instance, your dynamic DNS record is updated each time the IP address changes, as after
a shutdown and restart. In this example, you install the noip2 client, which works with the service
provided by noip.com.

a. Enable the Extra Packages for Enterprise Linux (EPEL) repository to gain access to the noip2
client.
Note
Amazon Linux instances have the GPG keys and repository information for the EPEL
repository installed by default; however, Red Hat and CentOS instances must first
install the epel-release package before you can enable the EPEL repository. For
more information and to download the latest version of this package, see https://
fedoraproject.org/wiki/EPEL.

• For Amazon Linux 2:

[ec2-user ~]$ sudo yum install https://dl.fedoraproject.org/pub/epel/epel-

release-latest-7.noarch.rpm

• For Amazon Linux AMI:

[ec2-user ~]$ sudo yum-config-manager --enable epel

b. Install the noip package.

[ec2-user ~]$ sudo yum install -y noip

697
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Run commands at launch

c. Create the configuration file. Enter the login and password information when prompted and
answer the subsequent questions to configure the client.

[ec2-user ~]$ sudo noip2 -C

3. Enable the noip service.

• For Amazon Linux 2:

[ec2-user ~]$ sudo systemctl enable noip.service

• For Amazon Linux AMI:

[ec2-user ~]$ sudo chkconfig noip on

4. Start the noip service.

• For Amazon Linux 2:

[ec2-user ~]$ sudo systemctl start noip.service

• For Amazon Linux AMI:

[ec2-user ~]$ sudo service noip start

This command starts the client, which reads the configuration file (/etc/no-ip2.conf) that you
created earlier and updates the IP address for the public DNS name that you chose.
5. Verify that the update client has set the correct IP address for your dynamic DNS name. Allow a few
minutes for the DNS records to update, and then try to connect to your instance using SSH with the
public DNS name that you configured in this procedure.

Run commands on your Linux instance at launch

When you launch an instance in Amazon EC2, you have the option of passing user data to the instance
that can be used to perform common automated configuration tasks and even run scripts after
the instance starts. You can pass two types of user data to Amazon EC2: shell scripts and cloud-init
directives. You can also pass this data into the launch wizard as plain text, as a file (this is useful for
launching instances using the command line tools), or as base64-encoded text (for API calls).

If you are interested in more complex automation scenarios, consider using AWS CloudFormation and
AWS OpsWorks. For more information, see the AWS CloudFormation User Guide and the AWS OpsWorks
User Guide.

For information about running commands on your Windows instance at launch, see Running Commands
on Your Windows Instance at Launch and Managing Windows Instance Configuration in the Amazon EC2
User Guide for Windows Instances.

In the following examples, the commands from the Install a LAMP Web Server on Amazon Linux 2 (p. 38)
are converted to a shell script and a set of cloud-init directives that run when the instance launches. In
each example, the following tasks are performed by the user data:

• The distribution software packages are updated.

• The necessary web server, php, and mariadb packages are installed.
• The httpd service is started and turned on via systemctl.

698
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Run commands at launch

• The ec2-user is added to the apache group.

• The appropriate ownership and file permissions are set for the web directory and the files contained
within it.
• A simple web page is created to test the web server and PHP engine.

Contents
• Prerequisites (p. 699)
• User data and shell scripts (p. 699)
• User data and the console (p. 700)
• User data and cloud-init directives (p. 701)
• User data and the AWS CLI (p. 702)

Prerequisites
The following examples assume that your instance has a public DNS name that is reachable from the
Internet. For more information, see Step 1: Launch an instance (p. 33). You must also configure your
security group to allow SSH (port 22), HTTP (port 80), and HTTPS (port 443) connections. For more
information about these prerequisites, see Set up to use Amazon EC2 (p. 28).

Also, these instructions are intended for use with Amazon Linux 2, and the commands and directives
may not work for other Linux distributions. For more information about other distributions, such as their
support for cloud-init, see their specific documentation.

User data and shell scripts

If you are familiar with shell scripting, this is the easiest and most complete way to send instructions to
an instance at launch. Adding these tasks at boot time adds to the amount of time it takes to boot the
instance. You should allow a few minutes of extra time for the tasks to complete before you test that the
user script has finished successfully.
Important
By default, user data scripts and cloud-init directives run only during the boot cycle when you
first launch an instance. You can update your configuration to ensure that your user data scripts
and cloud-init directives run every time you restart your instance. For more information, see
How can I execute user data with every restart of my EC2 instance? in the AWS Knowledge
Center.

User data shell scripts must start with the #! characters and the path to the interpreter you want to read
the script (commonly /bin/bash). For a great introduction on shell scripting, see the BASH Programming
HOW-TO at the Linux Documentation Project (tldp.org).

Scripts entered as user data are run as the root user, so do not use the sudo command in the script.
Remember that any files you create will be owned by root; if you need non-root users to have file
access, you should modify the permissions accordingly in the script. Also, because the script is not run
interactively, you cannot include commands that require user feedback (such as yum update without the
-y flag).

If you use an AWS API, including the AWS CLI, in a user data script, you must use an instance profile when
launching the instance. An instance profile provides the appropriate AWS credentials required by the user
data script to issue the API call. For more information, see Using instance profiles in the IAM User Guide.
The permissions you assign to the IAM role depend on which services you are calling with the API. For
more information, see IAM roles for Amazon EC2.

The cloud-init output log file (/var/log/cloud-init-output.log) captures console output so it is

easy to debug your scripts following a launch if the instance does not behave the way you intended.

699
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Run commands at launch

When a user data script is processed, it is copied to and run from /var/lib/cloud/
instances/instance-id/. The script is not deleted after it is run. Be sure to delete the user data
scripts from /var/lib/cloud/instances/instance-id/ before you create an AMI from the
instance. Otherwise, the script will exist in this directory on any instance launched from the AMI.

User data and the console

You can specify instance user data when you launch the instance. If the root volume of the instance is an
EBS volume, you can also stop the instance and update its user data.

Specify instance user data at launch

Follow the procedure for launching an instance at Launch an instance using the Launch Instance
Wizard (p. 526), but when you get to the section called “Step 3: Configure Instance Details” (p. 528) in
that procedure, copy your shell script in the User data field, and then complete the launch procedure.

In the example script below, the script creates and configures our web server.

#!/bin/bash
yum update -y
amazon-linux-extras install -y lamp-mariadb10.2-php7.2 php7.2
yum install -y httpd mariadb-server
systemctl start httpd
systemctl enable httpd
usermod -a -G apache ec2-user
chown -R ec2-user:apache /var/www
chmod 2775 /var/www
find /var/www -type d -exec chmod 2775 {} \;
find /var/www -type f -exec chmod 0664 {} \;
echo "<?php phpinfo(); ?>" > /var/www/html/phpinfo.php

Allow enough time for the instance to launch and run the commands in your script, and then check to
see that your script has completed the tasks that you intended.

For our example, in a web browser, enter the URL of the PHP test file the script created. This URL is the
public DNS address of your instance followed by a forward slash and the file name.

http://my.public.dns.amazonaws.com/phpinfo.php

You should see the PHP information page. If you are unable to see the PHP information page, check that
the security group you are using contains a rule to allow HTTP (port 80) traffic. For more information, see
Add rules to a security group (p. 1066).

(Optional) If your script did not accomplish the tasks you were expecting it to, or if you just want to
verify that your script completed without errors, examine the cloud-init output log file at /var/log/
cloud-init-output.log and look for error messages in the output.

For additional debugging information, you can create a Mime multipart archive that includes a cloud-init
data section with the following directive:

output : { all : '| tee -a /var/log/cloud-init-output.log' }

This directive sends command output from your script to /var/log/cloud-init-output.log. For
more information about cloud-init data formats and creating Mime multi part archive, see cloud-init
Formats.

View and update the instance user data

To update the instance user data, you must first stop the instance. If the instance is running, you can view
the user data but you cannot modify it.

700
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Run commands at launch

Warning
When you stop an instance, the data on any instance store volumes is erased. To keep data from
instance store volumes, be sure to back it up to persistent storage.

New console

To modify instance user data

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Instance state, Stop instance. If this option is disabled, either
the instance is already stopped or its root device is an instance store volume.
4. When prompted for confirmation, choose Stop. It can take a few minutes for the instance to
stop.
5. With the instance still selected, choose Actions, Instance settings, Edit user data.
6. Modify the user data as needed, and then choose Save.
7. Restart the instance. The new user data is visible on your instance after you restart it; however,
user data scripts are not run.

Old console

To modify instance user data

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, Instance State, Stop. If this option is disabled, either
the instance is already stopped or its root device is an instance store volume.
4. When prompted for confirmation, choose Yes, Stop. It can take a few minutes for the instance
to stop.
5. With the instance still selected, choose Actions, Instance Settings, View/Change User Data.
6. In the View/Change User Data dialog box, update the user data, and then choose Save.
7. Restart the instance. The new user data is visible on your instance after you restart it; however,
user data scripts are not run.

User data and cloud-init directives

The cloud-init package configures specific aspects of a new Amazon Linux instance when it is launched;
most notably, it configures the .ssh/authorized_keys file for the ec2-user so you can log in with
your own private key. For more information, see cloud-init (p. 184).

The cloud-init user directives can be passed to an instance at launch the same way that a script
is passed, although the syntax is different. For more information about cloud-init, go to http://
cloudinit.readthedocs.org/en/latest/index.html.
Important
By default, user data scripts and cloud-init directives run only during the boot cycle when you
first launch an instance. You can update your configuration to ensure that your user data scripts
and cloud-init directives run every time you restart your instance. For more information, see
How can I execute user data with every restart of my EC2 instance? in the AWS Knowledge
Center.

Adding these tasks at boot time adds to the amount of time it takes to boot an instance. You should
allow a few minutes of extra time for the tasks to complete before you test that your user data directives
have completed.

701
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Run commands at launch

To pass cloud-init directives to an instance with user data

1. Follow the procedure for launching an instance at Launch an instance using the Launch
Instance Wizard (p. 526), but when you get to the section called “Step 3: Configure Instance
Details” (p. 528) in that procedure, enter your cloud-init directive text in the User data field, and
then complete the launch procedure.

In the example below, the directives create and configure a web server on Amazon Linux 2. The
#cloud-config line at the top is required in order to identify the commands as cloud-init
directives.

#cloud-config
repo_update: true
repo_upgrade: all

packages:
- httpd
- mariadb-server

runcmd:
- [ sh, -c, "amazon-linux-extras install -y lamp-mariadb10.2-php7.2 php7.2" ]
- systemctl start httpd
- sudo systemctl enable httpd
- [ sh, -c, "usermod -a -G apache ec2-user" ]
- [ sh, -c, "chown -R ec2-user:apache /var/www" ]
- chmod 2775 /var/www
- [ find, /var/www, -type, d, -exec, chmod, 2775, {}, \; ]
- [ find, /var/www, -type, f, -exec, chmod, 0664, {}, \; ]
- [ sh, -c, 'echo "<?php phpinfo(); ?>" > /var/www/html/phpinfo.php' ]

2. Allow enough time for the instance to launch and run the directives in your user data, and then
check to see that your directives have completed the tasks you intended.

For our example, in a web browser, enter the URL of the PHP test file the directives created. This URL
is the public DNS address of your instance followed by a forward slash and the file name.

http://my.public.dns.amazonaws.com/phpinfo.php

You should see the PHP information page. If you are unable to see the PHP information page,
check that the security group you are using contains a rule to allow HTTP (port 80) traffic. For more
information, see Add rules to a security group (p. 1066).
3. (Optional) If your directives did not accomplish the tasks you were expecting them to, or if you
just want to verify that your directives completed without errors, examine the output log file at /
var/log/cloud-init-output.log and look for error messages in the output. For additional
debugging information, you can add the following line to your directives:

output : { all : '| tee -a /var/log/cloud-init-output.log' }

This directive sends runcmd output to /var/log/cloud-init-output.log.

User data and the AWS CLI

You can use the AWS CLI to specify, modify, and view the user data for your instance. For information
about viewing user data from your instance using instance metadata, see Retrieve instance user
data (p. 719).

702
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Run commands at launch

On Windows, you can use the AWS Tools for Windows PowerShell instead of using the AWS CLI. For more
information, see User Data and the Tools for Windows PowerShell in the Amazon EC2 User Guide for
Windows Instances.

Example: Specify user data at launch

To specify user data when you launch your instance, use the run-instances command with the --user-
data parameter. With run-instances, the AWS CLI performs base64 encoding of the user data for you.

The following example shows how to specify a script as a string on the command line:

aws ec2 run-instances --image-id ami-abcd1234 --count 1 --instance-type m3.medium \

--key-name my-key-pair --subnet-id subnet-abcd1234 --security-group-ids sg-abcd1234 \
--user-data echo user data

The following example shows how to specify a script using a text file. Be sure to use the file:// prefix
to specify the file.

aws ec2 run-instances --image-id ami-abcd1234 --count 1 --instance-type m3.medium \

--key-name my-key-pair --subnet-id subnet-abcd1234 --security-group-ids sg-abcd1234 \
--user-data file://my_script.txt

The following is an example text file with a shell script.

#!/bin/bash
yum update -y
service httpd start
chkconfig httpd on

Example: Modify the user data of a stopped instance

You can modify the user data of a stopped instance using the modify-instance-attribute command. With
modify-instance-attribute, the AWS CLI does not perform base64 encoding of the user data for you.

• On a Linux computer, use the base64 command to encode the user data.

base64 my_script.txt >my_script_base64.txt

• On a Windows computer, use the certutil command to encode the user data. Before you can use this
file with the AWS CLI, you must remove the first (BEGIN CERTIFICATE) and last (END CERTIFICATE)
lines.

certutil -encode my_script.txt my_script_base64.txt

notepad my_script_base64.txt

Use the --attribute and --value parameters to use the encoded text file to specify the user data. Be
sure to use the file:// prefix to specify the file.

aws ec2 modify-instance-attribute --instance-id i-1234567890abcdef0 --attribute userData --

value file://my_script_base64.txt

Example: Clear the user data of a stopped instance

To delete the existing user data, use the modify-instance-attribute command as follows:

aws ec2 modify-instance-attribute --instance-id i-1234567890abcdef0 --user-data Value=

703
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Example: View user data

To retrieve the user data for an instance, use the describe-instance-attribute command. With describe-
instance-attribute, the AWS CLI does not perform base64 decoding of the user data for you.

aws ec2 describe-instance-attribute --instance-id i-1234567890abcdef0 --attribute userData

The following is example output with the user data base64 encoded.

{
"UserData": {
"Value":
"IyEvYmluL2Jhc2gKeXVtIHVwZGF0ZSAteQpzZXJ2aWNlIGh0dHBkIHN0YXJ0CmNoa2NvbmZpZyBodHRwZCBvbg=="
},
"InstanceId": "i-1234567890abcdef0"
}

• On a Linux computer , use the --query option to get the encoded user data and the base64
command to decode it.

aws ec2 describe-instance-attribute --instance-id i-1234567890abcdef0 --attribute

userData --output text --query "UserData.Value" | base64 --decode

• On a Windows computer, use the --query option to get the coded user data and the certutil
command to decode it. Note that the encoded output is stored in a file and the decoded output is
stored in another file.

aws ec2 describe-instance-attribute --instance-id i-1234567890abcdef0 --attribute

userData --output text --query "UserData.Value" >my_output.txt
certutil -decode my_output.txt my_output_decoded.txt
type my_output_decoded.txt

The following is example output.

#!/bin/bash
yum update -y
service httpd start
chkconfig httpd on

Instance metadata and user data

Instance metadata is data about your instance that you can use to configure or manage the running
instance. Instance metadata is divided into categories, for example, host name, events, and security
groups.

You can also use instance metadata to access user data that you specified when launching your instance.
For example, you can specify parameters for configuring your instance, or include a simple script. You
can build generic AMIs and use user data to modify the configuration files supplied at launch time. For
example, if you run web servers for various small businesses, they can all use the same generic AMI and
retrieve their content from the Amazon S3 bucket that you specify in the user data at launch. To add a
new customer at any time, create a bucket for the customer, add their content, and launch your AMI with
the unique bucket name provided to your code in the user data. If you launch more than one instance
at the same time, the user data is available to all instances in that reservation. Each instance that is part
of the same reservation has a unique ami-launch-index number, allowing you to write code that
controls what to do. For example, the first host might elect itself as the original node in a cluster. For a
detailed AMI launch example, see Example: AMI launch index value (p. 720).

704
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

EC2 instances can also include dynamic data, such as an instance identity document that is generated
when the instance is launched. For more information, see Dynamic data categories (p. 730).
Important
Although you can only access instance metadata and user data from within the instance itself,
the data is not protected by authentication or cryptographic methods. Anyone who has direct
access to the instance, and potentially any software running on the instance, can view its
metadata. Therefore, you should not store sensitive data, such as passwords or long-lived
encryption keys, as user data.

Contents
• Configure the instance metadata service (p. 705)
• Retrieve instance metadata (p. 711)
• Work with instance user data (p. 718)
• Retrieve dynamic data (p. 720)
• Example: AMI launch index value (p. 720)
• Instance metadata categories (p. 723)
• Instance identity documents (p. 730)

Configure the instance metadata service

You can access instance metadata from a running instance using one of the following methods:

• Instance Metadata Service Version 1 (IMDSv1) – a request/response method

• Instance Metadata Service Version 2 (IMDSv2) – a session-oriented method

By default, you can use either IMDSv1 or IMDSv2, or both. The instance metadata service distinguishes
between IMDSv1 and IMDSv2 requests based on whether, for any given request, either the PUT or GET
headers, which are unique to IMDSv2, are present in that request.

You can configure the instance metadata service on each instance such that local code or users must use
IMDSv2. When you specify that IMDSv2 must be used, IMDSv1 no longer works. For more information,
see Configure the instance metadata options (p. 708).

To retrieve instance metadata, see Retrieve instance metadata (p. 711).

How Instance Metadata Service Version 2 works

IMDSv2 uses session-oriented requests. With session-oriented requests, you create a session token
that defines the session duration, which can be a minimum of one second and a maximum of six hours.
During the specified duration, you can use the same session token for subsequent requests. After the
specified duration expires, you must create a new session token to use for future requests.

The following example uses a Linux shell script and IMDSv2 to retrieve the top-level instance metadata
items. The example command:

• Creates a session token lasting six hours (21,600 seconds) using the PUT request
• Stores the session token header in a variable named TOKEN
• Requests the top-level metadata items using the token

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-

metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/meta-data/

705
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

After you've created a token, you can reuse it until it expires. In the following example command, which
gets the ID of the AMI used to launch the instance, the token that is stored in $TOKEN in the previous
example is reused.

[ec2-user ~]$ curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/

meta-data/ami-id

When you use IMDSv2 to request instance metadata, the request must include the following:

1. Use a PUT request to initiate a session to the instance metadata service. The PUT request returns a
token that must be included in subsequent GET requests to the instance metadata service. The token
is required to access metadata using IMDSv2.
2. Include the token in all GET requests to the instance metadata service. When token usage is set to
required, requests without a valid token or with an expired token receive a 401 - Unauthorized
HTTP error code. For information about changing the token usage requirement, see modify-instance-
metadata-options in the AWS CLI Command Reference.
• The token is an instance-specific key. The token is not valid on other EC2 instances and will be
rejected if you attempt to use it outside of the instance on which it was generated.
• The PUT request must include a header that specifies the time to live (TTL) for the token, in seconds,
up to a maximum of six hours (21,600 seconds). The token represents a logical session. The TTL
specifies the length of time that the token is valid and, therefore, the duration of the session.
• After a token expires, to continue accessing instance metadata, you must create a new session using
another PUT.
• You can choose to reuse a token or create a new token with every request. For a small number of
requests, it might be easier to generate and immediately use a token each time you need to access
the instance metadata service. But for efficiency, you can specify a longer duration for the token
and reuse it rather than having to write a PUT request every time you need to request instance
metadata. There is no practical limit on the number of concurrent tokens, each representing its own
session. IMDSv2 is, however, still constrained by normal instance metadata service connection and
throttling limits. For more information, see Query throttling (p. 717).

HTTP GET and HEAD methods are allowed in IMDSv2 instance metadata requests. PUT requests are
rejected if they contain an X-Forwarded-For header.

By default, the response to PUT requests has a response hop limit (time to live) of 1 at the IP protocol
level. You can adjust the hop limit using the modify-instance-metadata-options command if you
need to make it larger. For example, you might need a larger hop limit for backward compatibility with
container services running on the instance. For more information, see modify-instance-metadata-options
in the AWS CLI Command Reference.

Transition to using Instance Metadata Service Version 2

Use of Instance Metadata Service Version 2 (IMDSv2) is optional. Instance Metadata Service Version
1 (IMDSv1) will continue to be supported indefinitely. If you choose to migrate to using IMDSv2, we
recommend that you use the following tools and transition path.

Tools for helping with the transition to IMDSv2

If your software uses IMDSv1, use the following tools to help reconfigure your software to use IMDSv2.

• AWS software: The latest versions of the AWS SDKs and CLIs support IMDSv2. To use IMDSv2, make
sure that your EC2 instances have the latest versions of the AWS SDKs and CLIs. For information
about updating the CLI, see Upgrading to the latest version of the AWS CLI in the AWS Command Line
Interface User Guide.
• CloudWatch: IMDSv2 uses token-backed sessions, while IMDSv1 does not. The MetadataNoToken
CloudWatch metric tracks the number of calls to the instance metadata service that are using IMDSv1.

706
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

By tracking this metric to zero, you can determine if and when all of your software has been upgraded
to use IMDSv2. For more information, see Instance metrics (p. 764).
• Updates to EC2 APIs and CLIs: For existing instances, you can use the modify-instance-metadata-
options CLI command (or the ModifyInstanceMetadataOptions API) to require the use of IMDSv2.
For new instances, you can use the run-instances CLI command (or the RunInstances API) and the
metadata-options parameter to launch new instances that require the use of IMDSv2.

To require the use of IMDSv2 on all new instances launched by Auto Scaling groups, your Auto Scaling
groups can use either a launch template or a launch configuration. When you create a launch template
or create a launch configuration, you must configure the MetadataOptions parameters to require the
use IMDSv2. After you configure the launch template or launch configuration, the Auto Scaling group
launches new instances using the new launch template or launch configuration, but existing instances
are not affected.

Use the modify-instance-metadata-options CLI command (or the ModifyInstanceMetadataOptions

API) to require the use of IMDSv2 on the existing instances, or terminate the instances and the Auto
Scaling group will launch new replacement instances with the instance metadata options settings that
are defined in the launch template or launch configuration.

For Auto Scaling groups that use launch configurations, you can replace the launch configurations with
launch templates.
• IAM policies and SCPs: You can use an IAM condition to enforce that IAM users can't launch an
instance unless it uses IMDSv2. You can also use IAM conditions to enforce that IAM users can't modify
running instances to re-enable IMDSv1, and to enforce that the instance metadata service is available
on the instance.

The ec2:MetadataHttpTokens, ec2:MetadataHttpPutResponseHopLimit, and

ec2:MetadataHttpEndpoint IAM condition keys can be used to control the use of the RunInstances
and the ModifyInstanceMetadataOptions API and corresponding CLI. If a policy is created, and a
parameter in the API call does not match the state specified in the policy using the condition key, the
API or CLI call fails with an UnauthorizedOperation response. These condition keys can be used
either in IAM policies or AWS Organizations service control policies (SCPs).

Furthermore, you can choose an additional layer of protection to enforce the change from IMDSv1
to IMDSv2. At the access management layer with respect to the APIs called via EC2 Role credentials,
you can use a new condition key in either IAM policies or AWS Organizations service control
policies (SCPs). Specifically, by using the policy condition key ec2:RoleDelivery with a value
of 2.0 in your IAM policies, API calls made with EC2 Role credentials obtained from IMDSv1 will
receive an UnauthorizedOperation response. The same thing can be achieved more broadly
with that condition required by an SCP. This ensures that credentials delivered via IMDSv1 cannot
actually be used to call APIs because any API calls not matching the specified condition will
receive an UnauthorizedOperation error. For example IAM policies, see Work with instance
metadata (p. 1021). For more information, see Service Control Policies in the AWS Organizations User
Guide.

Recommended path to requiring IMDSv2 access

Using the above tools, we recommend that you follow this path for transitioning to IMDSv2:

Step 1: At the start

Update the SDKs, CLIs, and your software that use Role credentials on their EC2 instances to IMDSv2-
compatible versions. For information about updating the CLI, see Upgrading to the latest version of the
AWS CLI in the AWS Command Line Interface User Guide.

Then, change your software that directly accesses instance metadata (in other words, that does not use
an SDK) using the IMDSv2 requests.

707
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Step 2: During the transition

Track your transition progress by using the CloudWatch metric MetadataNoToken. This metric shows
the number of calls to the instance metadata service that are using IMDSv1 on your instances. For more
information, see Instance metrics (p. 764).

Step 3: When everything is ready on all instances

Everything is ready on all instances when the CloudWatch metric MetadataNoToken records zero
IMDSv1 usage. At this stage, you can do the following:

• For existing instances: You can require IMDSv2 use through the modify-instance-metadata-options
command. You can make these changes on running instances; you do not need to restart your
instances.
• For new instances: When launching a new instance, you can do one of the following:
• In the Amazon EC2 console launch instance wizard, set Metadata accessible to Enabled and
Metadata version to V2. For more information, see Step 3: Configure Instance Details (p. 528).
• Use the run-instances command to specify that only IMDSv2 is to be used.

Updating instance metadata options for existing instances is available only through the API or AWS CLI.
It is currently not available in the Amazon EC2 console. For more information, see Configure the instance
metadata options (p. 708).

Step 4: When all of your instances are transitioned to IMDSv2

The ec2:MetadataHttpTokens, ec2:MetadataHttpPutResponseHopLimit, and
ec2:MetadataHttpEndpoint IAM condition keys can be used to control the use of the RunInstances
and the ModifyInstanceMetadataOptions API and corresponding CLI. If a policy is created, and a
parameter in the API call does not match the state specified in the policy using the condition key, the API
or CLI call fails with an UnauthorizedOperation response. For example IAM policies, see Work with
instance metadata (p. 1021).

Configure the instance metadata options

Instance metadata options allow you to configure new or existing instances to do the following:

• Require the use of IMDSv2 when requesting instance metadata

• Specify the PUT response hop limit
• Turn off access to instance metadata

You can also use IAM condition keys in an IAM policy or SCP to do the following:

• Allow an instance to launch only if it's configured to require the use of IMDSv2
• Restrict the number of allowed hops
• Turn off access to instance metadata

You can configure instance metadata options when launching new instances from the Amazon EC2
console. For more information, see Step 3: Configure Instance Details (p. 528).

To configure the instance metadata options on new or existing instances, you can use the AWS SDK or
AWS CLI. For more information, see run-instances and modify-instance-metadata-options in the AWS CLI
Command Reference.
Note
You should proceed cautiously and conduct careful testing before making any changes. Take
note of the following:

708
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

• If you enforce the use of IMDSv2, applications or agents that use IMDSv1 for instance
metadata access will break.
• If you turn off all access to instance metadata, applications or agents that rely on instance
metadata access to function will break.

Topics
• Configure instance metadata options for new instances (p. 709)
• Configure instance metadata options for existing instances (p. 710)

Configure instance metadata options for new instances

You can require the use of IMDSv2 on an instance when you launch it. You can also create an IAM policy
that prevents users from launching new instances unless they require IMDSv2 on the new instance.

Console

To require the use of IMDSv2 on a new instance

• When launching a new instance in the Amazon EC2 console, select the following options on the
Configure Instance Details page:

• Under Advanced Details, for Metadata accessible, select Enabled.

• For Metadata version, select V2 (token required).

For more information, see Step 3: Configure Instance Details (p. 528).
AWS CLI

To require the use of IMDSv2 on a new instance

The following run-instances example launches a c3.large instance with --metadata-options

set to HttpTokens=required. When you specify a value for HttpTokens, you must also set
HttpEndpoint to enabled. Because the secure token header is set to required for metadata
retrieval requests, this opts in the instance to require using IMDSv2 when requesting instance
metadata.

aws ec2 run-instances

--image-id ami-0abcdef1234567890
--instance-type c3.large
...
--metadata-options "HttpEndpoint=enabled,HttpTokens=required"

To enforce the use of IMDSv2 on all new instances

To ensure that IAM users can only launch instances that require the use of IMDSv2 when requesting
instance metadata, you can specify that the condition to require IMDSv2 must be met before an instance
can be launched. For the example IAM policy, see Work with instance metadata (p. 1021).

Console

To turn off access to instance metadata

• To ensure that access to your instance metadata is turned off, regardless of which version of the
instance metadata service you are using, launch the instance in the Amazon EC2 console with
the following option selected on the Configure Instance Details page:

709
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

• Under Advanced Details, for Metadata accessible, select Disabled.

For more information, see Step 3: Configure Instance Details (p. 528).
AWS CLI

To turn off access to instance metadata

To ensure that access to your instance metadata is turned off, regardless of which version of the
instance metadata service you are using, launch the instance with --metadata-options set to
HttpEndpoint=disabled. You can turn access on later by using the modify-instance-metadata-
options command.

aws ec2 run-instances

--image-id ami-0abcdef1234567890
--instance-type c3.large
...
--metadata-options "HttpEndpoint=disabled"

Configure instance metadata options for existing instances

You can require the use IMDSv2 on an existing instance. You can also change the PUT response hop limit
and turn off access to instance metadata on an existing instance. You can also create an IAM policy that
prevents users from modifying the instance metadata options on an existing instance.

To require the use of IMDSv2

You can opt in to require that IMDSv2 is used when requesting instance metadata. Use the modify-
instance-metadata-options CLI command and set the http-tokens parameter to required. When you
specify a value for http-tokens, you must also set http-endpoint to enabled.

aws ec2 modify-instance-metadata-options \

--instance-id i-1234567898abcdef0 \
--http-tokens required \
--http-endpoint enabled

To change the PUT response hop limit

For existing instances, you can change the settings of the PUT response hop limit. Use the modify-
instance-metadata-options CLI command and set the http-put-response-hop-limit parameter
to the required number of hops. In the following example, the hop limit is set to 3. Note that when
specifying a value for http-put-response-hop-limit, you must also set http-endpoint to
enabled.

aws ec2 modify-instance-metadata-options \

--instance-id i-1234567898abcdef0 \
--http-put-response-hop-limit 3 \
--http-endpoint enabled

To restore the use of IMDSv1 on an instance using IMDSv2

You can use the modify-instance-metadata-options CLI command with http-tokens set to optional
to restore the use of IMDSv1 when requesting instance metadata.

aws ec2 modify-instance-metadata-options \

--instance-id i-1234567898abcdef0 \

710
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

--http-tokens optional \
--http-endpoint enabled

To turn off access to instance metadata

You can turn off access to your instance metadata by disabling the HTTP endpoint of the instance
metadata service, regardless of which version of the instance metadata service you are using. You can
reverse this change at any time by enabling the HTTP endpoint. Use the modify-instance-metadata-
options CLI command and set the http-endpoint parameter to disabled.

aws ec2 modify-instance-metadata-options \

--instance-id i-1234567898abcdef0 \
--http-endpoint disabled

To control the use of modify-instance-metadata-options

To control which IAM users can modify the instance metadata options, specify a policy that prevents
all users other than users with a specified role to use the ModifyInstanceMetadataOptions API. For the
example IAM policy, see Work with instance metadata (p. 1021).

Retrieve instance metadata

Because your instance metadata is available from your running instance, you do not need to use the
Amazon EC2 console or the AWS CLI. This can be helpful when you're writing scripts to run from your
instance. For example, you can access the local IP address of your instance from instance metadata to
manage a connection to an external application.

Instance metadata is divided into categories. For a description of each instance metadata category, see
Instance metadata categories (p. 723).

To view all categories of instance metadata from within a running instance, use the following URI.

http://169.254.169.254/latest/meta-data/

The IP address 169.254.169.254 is a link-local address and is valid only from the instance. For more
information, see Link-local address on Wikipedia.

Note that you are not billed for HTTP requests used to retrieve instance metadata and user data.

The command format is different, depending on whether you use IMDSv1 or IMDSv2. By default, you can
use both instance metadata services. To require the use of IMDSv2, see Configure the instance metadata
service (p. 705).

You can use a tool such as cURL, as shown in the following example.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/meta-
data/

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/

711
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Note
The AWS SDKs use IMDSv2 calls by default. If the IMDSv2 call receives no response, the SDK
retries the call and, if still unsuccessful, uses IMDSv1. This can result in a delay. In a container
environment, if the hop limit is 1, the IMDSv2 response does not return because going to the
container is considered an additional network hop. To avoid the process of falling back to
IMDSv1 and the resultant delay, in a container environment we recommend that you set the hop
limit to 2. For more information, see Configure the instance metadata options (p. 708).

You can also download the Instance Metadata Query tool, which allows you to query the instance
metadata using Instance Metadata Service Version 1 without having to enter the full URI or category
names.

Responses and error messages

All instance metadata is returned as text (HTTP content type text/plain).

A request for a specific metadata resource returns the appropriate value, or a 404 - Not Found HTTP
error code if the resource is not available.

A request for a general metadata resource (the URI ends with a /) returns a list of available resources, or
a 404 - Not Found HTTP error code if there is no such resource. The list items are on separate lines,
terminated by line feeds (ASCII 10).

For requests made using Instance Metadata Service Version 2, the following HTTP error codes can be
returned:

• 400 - Missing or Invalid Parameters – The PUT request is not valid.

• 401 - Unauthorized – The GET request uses an invalid token. The recommended action is to
generate a new token.
• 403 - Forbidden – The request is not allowed or the instance metadata service is turned off.

Examples of retrieving instance metadata

Examples
• Get the available versions of the instance metadata (p. 712)
• Get the top-level metadata items (p. 713)
• Get the list of available public keys (p. 715)
• Show the formats in which public key 0 is available (p. 716)
• Get public key 0 (in the OpenSSH key format) (p. 716)
• Get the subnet ID for an instance (p. 717)

Get the available versions of the instance metadata

This example gets the available versions of the instance metadata. These versions do not necessarily
correlate with an Amazon EC2 API version. The earlier versions are available to you in case you have
scripts that rely on the structure and information present in a previous version.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/
1.0

712
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

2007-01-19
2007-03-01
2007-08-29
2007-10-10
2007-12-15
2008-02-01
2008-09-01
2009-04-04
2011-01-01
2011-05-01
2012-01-12
2014-02-25
2014-11-05
2015-10-20
2016-04-19
2016-06-30
2016-09-02
latest

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/

1.0
2007-01-19
2007-03-01
2007-08-29
2007-10-10
2007-12-15
2008-02-01
2008-09-01
2009-04-04
2011-01-01
2011-05-01
2012-01-12
2014-02-25
2014-11-05
2015-10-20
2016-04-19
2016-06-30
2016-09-02
latest

Get the top-level metadata items

This example gets the top-level metadata items. For more information, see Instance metadata
categories (p. 723).

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/meta-
data/
ami-id
ami-launch-index
ami-manifest-path
block-device-mapping/
events/
hostname
iam/
instance-action
instance-id

713
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

instance-life-cycle
instance-type
local-hostname
local-ipv4
mac
metrics/
network/
placement/
profile
public-hostname
public-ipv4
public-keys/
reservation-id
security-groups
services/

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/

ami-id
ami-launch-index
ami-manifest-path
block-device-mapping/
events/
hostname
iam/
instance-action
instance-id
instance-type
local-hostname
local-ipv4
mac
metrics/
network/
placement/
profile
public-hostname
public-ipv4
public-keys/
reservation-id
security-groups
services/

The following examples get the values of some of the top-level metadata items that were obtained in
the preceding example. The IMDSv2 requests use the stored token that was created in the preceding
example command, assuming it has not expired.

IMDSv2

[ec2-user ~]$ curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/

latest/meta-data/ami-id
ami-0abcdef1234567890

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/ami-id

ami-0abcdef1234567890

714
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

IMDSv2

[ec2-user ~]$ curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/

latest/meta-data/reservation-id
r-0efghijk987654321

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/reservation-id

r-0efghijk987654321

IMDSv2

[ec2-user ~]$ curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/

latest/meta-data/local-hostname
ip-10-251-50-12.ec2.internal

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/local-hostname

ip-10-251-50-12.ec2.internal

IMDSv2

[ec2-user ~]$ curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/

latest/meta-data/public-hostname
ec2-203-0-113-25.compute-1.amazonaws.com

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/public-hostname

ec2-203-0-113-25.compute-1.amazonaws.com

Get the list of available public keys

This example gets the list of available public keys.

IMDSv2

[ec2-user ~]$ `curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-

metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/meta-
data/public-keys/
0=my-public-key

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/public-keys/

715
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

0=my-public-key

Show the formats in which public key 0 is available

This example shows the formats in which public key 0 is available.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/meta-
data/public-keys/0/
openssh-key

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/public-keys/0/

openssh-key

Get public key 0 (in the OpenSSH key format)

This example gets public key 0 (in the OpenSSH key format).

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/meta-
data/public-keys/0/openssh-key
ssh-rsa MIICiTCCAfICCQD6m7oRw0uXOjANBgkqhkiG9w0BAQUFADCBiDELMAkGA1UEBhMC
VVMxCzAJBgNVBAgTAldBMRAwDgYDVQQHEwdTZWF0dGxlMQ8wDQYDVQQKEwZBbWF6
b24xFDASBgNVBAsTC0lBTSBDb25zb2xlMRIwEAYDVQQDEwlUZXN0Q2lsYWMxHzAd
BgkqhkiG9w0BCQEWEG5vb25lQGFtYXpvbi5jb20wHhcNMTEwNDI1MjA0NTIxWhcN
MTIwNDI0MjA0NTIxWjCBiDELMAkGA1UEBhMCVVMxCzAJBgNVBAgTAldBMRAwDgYD
VQQHEwdTZWF0dGxlMQ8wDQYDVQQKEwZBbWF6b24xFDASBgNVBAsTC0lBTSBDb25z
b2xlMRIwEAYDVQQDEwlUZXN0Q2lsYWMxHzAdBgkqhkiG9w0BCQEWEG5vb25lQGFt
YXpvbi5jb20wgZ8wDQYJKoZIhvcNAQEBBQADgY0AMIGJAoGBAMaK0dn+a4GmWIWJ
21uUSfwfEvySWtC2XADZ4nB+BLYgVIk60CpiwsZ3G93vUEIO3IyNoH/f0wYK8m9T
rDHudUZg3qX4waLG5M43q7Wgc/MbQITxOUSQv7c7ugFFDzQGBzZswY6786m86gpE
Ibb3OhjZnzcvQAaRHhdlQWIMm2nrAgMBAAEwDQYJKoZIhvcNAQEFBQADgYEAtCu4
nUhVVxYUntneD9+h8Mg9q6q+auNKyExzyLwaxlAoo7TJHidbtS4J5iNmZgXL0Fkb
FFBjvSfpJIlJ00zbhNYS5f6GuoEDmFJl0ZxBHjJnyp378OD8uTs7fLvjx79LjSTb
NYiytVbZPQUQ5Yaxu2jXnimvw3rrszlaEXAMPLE my-public-key

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/public-keys/0/openssh-key

ssh-rsa MIICiTCCAfICCQD6m7oRw0uXOjANBgkqhkiG9w0BAQUFADCBiDELMAkGA1UEBhMC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716
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

nUhVVxYUntneD9+h8Mg9q6q+auNKyExzyLwaxlAoo7TJHidbtS4J5iNmZgXL0Fkb
FFBjvSfpJIlJ00zbhNYS5f6GuoEDmFJl0ZxBHjJnyp378OD8uTs7fLvjx79LjSTb
NYiytVbZPQUQ5Yaxu2jXnimvw3rrszlaEXAMPLE my-public-key

Get the subnet ID for an instance

This example gets the subnet ID for an instance.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/meta-
data/network/interfaces/macs/02:29:96:8f:6a:2d/subnet-id
subnet-be9b61d7

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/network/interfaces/

macs/02:29:96:8f:6a:2d/subnet-id
subnet-be9b61d7

Query throttling
We throttle queries to the instance metadata service on a per-instance basis, and we place limits on the
number of simultaneous connections from an instance to the instance metadata service.

If you're using the instance metadata service to retrieve AWS security credentials, avoid querying for
credentials during every transaction or concurrently from a high number of threads or processes, as
this might lead to throttling. Instead, we recommend that you cache the credentials until they start
approaching their expiry time.

If you are throttled while accessing the instance metadata service, retry your query with an exponential
backoff strategy.

Limit instance metadata service access

You can consider using local firewall rules to disable access from some or all processes to the instance
metadata service.

Using iptables to limit access

The following example uses Linux iptables and its owner module to prevent the Apache webserver
(based on its default installation user ID of apache) from accessing 169.254.169.254. It uses a deny rule
to reject all instance metadata requests (whether IMDSv1 or IMDSv2) from any process running as that
user.

$ sudo iptables --append OUTPUT --proto tcp --destination 169.254.169.254 --match owner --
uid-owner apache --jump REJECT

Or, you can consider only allowing access to particular users or groups, by using allow rules. Allow rules
might be easier to manage from a security perspective, because they require you to make a decision
about what software needs access to instance metadata. If you use allow rules, it's less likely you will
accidentally allow software to access the metadata service (that you did not intend to have access) if
you later change the software or configuration on an instance. You can also combine group usage with

717
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

allow rules, so that you can add and remove users from a permitted group without needing to change
the firewall rule.

The following example prevents access to the instance metadata service by all processes, except for
processes running in the user account trustworthy-user.

$ sudo iptables --append OUTPUT --proto tcp --destination 169.254.169.254 --match owner !
--uid-owner trustworthy-user --jump REJECT

Note

• To use local firewall rules, you need to adapt the preceding example commands to suit your
needs.
• By default, iptables rules are not persistent across system reboots. They can be made to be
persistent by using OS features, not described here.
• The iptables owner module only matches group membership if the group is the primary
group of a given local user. Other groups are not matched.

Using PF or IPFW to limit access

If you are using FreeBSD or OpenBSD, you can also consider using PF or IPFW. The following examples
limit access to the instance metadata service to just the root user.

PF

$ block out inet proto tcp from any to 169.254.169.254

$ pass out inet proto tcp from any to 169.254.169.254 user root

IPFW

$ allow tcp from any to 169.254.169.254 uid root

$ deny tcp from any to 169.254.169.254

Note
The order of the PF and IPFW commands matter. PF defaults to last matching rule and IPFW
defaults to first matching rule.

Work with instance user data

When working with instance user data, keep the following in mind:

• User data must be base64-encoded. The Amazon EC2 console can perform the base64-encoding for
you or accept base64-encoded input.
• User data is limited to 16 KB, in raw form, before it is base64-encoded. The size of a string of length n
after base64-encoding is ceil(n/3)*4.
• User data must be base64-decoded when you retrieve it. If you retrieve the data using instance
metadata or the console, it's decoded for you automatically.
• User data is treated as opaque data: what you give is what you get back. It is up to the instance to be
able to interpret it.
• If you stop an instance, modify its user data, and start the instance, the updated user data is not run
when you start the instance.

718
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Specify instance user data at launch

You can specify user data when you launch an instance. For more information, see Launch an
instance using the Launch Instance Wizard (p. 526) and Run commands on your Linux instance at
launch (p. 698).

Modify instance user data

You can modify user data for an instance in the stopped state if the root volume is an EBS volume. For
more information, see View and update the instance user data (p. 700).

Retrieve instance user data

To retrieve user data from within a running instance, use the following URI.

http://169.254.169.254/latest/user-data

A request for user data returns the data as it is (content type application/octet-stream).

This example returns user data that was provided as comma-separated text.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/user-
data
1234,john,reboot,true | 4512,richard, | 173,,,

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/user-data

1234,john,reboot,true | 4512,richard, | 173,,,

This example returns user data that was provided as a script.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/user-
data
#!/bin/bash
yum update -y
service httpd start
chkconfig httpd on

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/user-data

#!/bin/bash
yum update -y
service httpd start
chkconfig httpd on

To retrieve user data for an instance from your own computer, see User data and the AWS CLI (p. 702)

719
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Retrieve dynamic data

To retrieve dynamic data from within a running instance, use the following URI.

http://169.254.169.254/latest/dynamic/

This example shows how to retrieve the high-level instance identity categories.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/dynamic/
instance-identity/
rsa2048
pkcs7
document
signature
dsa2048

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/dynamic/instance-identity/

rsa2048
pkcs7
document
signature
dsa2048

For more information about dynamic data and examples of how to retrieve it, see Instance identity
documents (p. 730).

Example: AMI launch index value

This example demonstrates how you can use both user data and instance metadata to configure your
instances.

Alice wants to launch four instances of her favorite database AMI, with the first acting as the original
instance and the remaining three acting as replicas. When she launches them, she wants to add user data
about the replication strategy for each replica. She is aware that this data will be available to all four
instances, so she needs to structure the user data in a way that allows each instance to recognize which
parts are applicable to it. She can do this using the ami-launch-index instance metadata value, which
will be unique for each instance.

Here is the user data that Alice has constructed.

replicate-every=1min | replicate-every=5min | replicate-every=10min

The replicate-every=1min data defines the first replica's configuration, replicate-every=5min

defines the second replica's configuration, and so on. Alice decides to provide this data as an ASCII string
with a pipe symbol (|) delimiting the data for the separate instances.

Alice launches four instances using the run-instances command, specifying the user data.

aws ec2 run-instances \

720
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

--image-id ami-0abcdef1234567890 \
--count 4 \
--instance-type t2.micro \
--user-data "replicate-every=1min | replicate-every=5min | replicate-every=10min"

After they're launched, all instances have a copy of the user data and the common metadata shown here:

• AMI ID: ami-0abcdef1234567890

• Reservation ID: r-1234567890abcabc0
• Public keys: none
• Security group name: default
• Instance type: t2.micro

However, each instance has certain unique metadata.

Instance 1

Metadata Value

instance-id i-1234567890abcdef0

ami-launch-index 0

public-hostname ec2-203-0-113-25.compute-1.amazonaws.com

public-ipv4 67.202.51.223

local-hostname ip-10-251-50-12.ec2.internal

local-ipv4 10.251.50.35

Instance 2

Metadata Value

instance-id i-0598c7d356eba48d7

ami-launch-index 1

public-hostname ec2-67-202-51-224.compute-1.amazonaws.com

public-ipv4 67.202.51.224

local-hostname ip-10-251-50-36.ec2.internal

local-ipv4 10.251.50.36

Instance 3

Metadata Value

instance-id i-0ee992212549ce0e7

ami-launch-index 2

public-hostname ec2-67-202-51-225.compute-1.amazonaws.com

721
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Metadata Value

public-ipv4 67.202.51.225

local-hostname ip-10-251-50-37.ec2.internal

local-ipv4 10.251.50.37

Instance 4

Metadata Value

instance-id i-1234567890abcdef0

ami-launch-index 3

public-hostname ec2-67-202-51-226.compute-1.amazonaws.com

public-ipv4 67.202.51.226

local-hostname ip-10-251-50-38.ec2.internal

local-ipv4 10.251.50.38

Alice can use the ami-launch-index value to determine which portion of the user data is applicable to
a particular instance.

1. She connects to one of the instances, and retrieves the ami-launch-index for that instance to
ensure it is one of the replicas:
IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/meta-data/api/token"

-H "X-aws-ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/meta-
data/ami-launch-index
2

For the following steps, the IMDSv2 requests use the stored token from the preceding IMDSv2
command, assuming the token has not expired.
IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/ami-launch-index

2

2. She saves the ami-launch-index as a variable.

IMDSv2

[ec2-user ~]$ ami_launch_index=`curl -H "X-aws-ec2-metadata-token: $TOKEN" -v

http://169.254.169.254/latest/meta-data/ami-launch-index`

IMDSv1

[ec2-user ~]$ ami_launch_index=`curl http://169.254.169.254/latest/meta-data/ami-

launch-index`

3. She saves the user data as a variable.

722
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

IMDSv2

[ec2-user ~]$ user_data=`curl -H "X-aws-ec2-metadata-token: $TOKEN" -v

http://169.254.169.254/latest/user-data`

IMDSv1

[ec2-user ~]$ user_data=`curl http://169.254.169.254/latest/user-data`

4. Finally, Alice uses the cut command to extract the portion of the user data that is applicable to that
instance.
IMDSv2

[ec2-user ~]$ echo $user_data | cut -d"|" -f"$ami_launch_index"

replicate-every=5min

IMDSv1

[ec2-user ~]$ echo $user_data | cut -d"|" -f"$ami_launch_index"

replicate-every=5min

Instance metadata categories

The following table lists the categories of instance metadata.
Note
When Amazon EC2 releases a new instance metadata category, the instance metadata for
the new category might not be available for existing instances. To ensure that the instance
metadata is available for an existing instance, you need to stop and then start (p. 623) the
instance.
Important
Some of the category names in the following table are placeholders for data that is unique to
your instance. For example, mac represents the MAC address for the network interface. You must
replace the placeholders with the actual values.

Data Description Release date

ami-id The AMI ID used to launch the Version 1.0

instance.

ami-launch-index If you started more than one Version 1.0

instance at the same time, this value
indicates the order in which the
instance was launched. The value of
the first instance launched is 0.

ami-manifest-path The path to the AMI manifest file in Version 1.0

Amazon S3. If you used an Amazon
EBS-backed AMI to launch the
instance, the returned result is
unknown.

ancestor-ami-ids The AMI IDs of any instances that 2007-10-10

were rebundled to create this AMI.

723
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Data Description Release date

This value will only exist if the
AMI manifest file contained an
ancestor-amis key.

block-device-mapping/ami The virtual device that contains the 2007-12-15

root/boot file system.

block-device-mapping/ebs The virtual devices associated with 2007-12-15

N any Amazon EBS volumes. Amazon
EBS volumes are only available in
metadata if they were present at
launch time or when the instance
was last started. The N indicates the
index of the Amazon EBS volume
(such as ebs1 or ebs2).

block-device-mapping/eph The virtual devices for any non- 2007-12-15

emeral NVMe instance store volumes.
N The N indicates the index of each
volume. The number of instance
store volumes in the block device
mapping might not match the actual
number of instance store volumes
for the instance. The instance
type determines the number of
instance store volumes that are
available to an instance. If the
number of instance store volumes in
a block device mapping exceeds the
number available to an instance, the
additional instance store volumes are
ignored.

block-device-mapping/root The virtual devices or partitions 2007-12-15

associated with the root devices
or partitions on the virtual device,
where the root (/ or C:) file system is
associated with the given instance.

block-device-mapping/swap The virtual devices associated with 2007-12-15

swap. Not always present.

elastic-gpus/ If there is an Elastic GPU attached to 2016-11-30

associations/elastic-gpu-id the instance, contains a JSON string
with information about the Elastic
GPU, including its ID and connection
information.

elastic-inference/ If there is an Elastic Inference 2018-11-29

associations/eia-id accelerator attached to the instance,
contains a JSON string with
information about the Elastic
Inference accelerator, including its ID
and type.

724
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Data Description Release date

events/maintenance/history If there are completed or canceled 2018-08-17

maintenance events for the instance,
contains a JSON string with
information about the events. For
more information, see To view event
history about completed or canceled
events (p. 754).

events/maintenance/sched If there are active maintenance 2018-08-17

uled events for the instance, contains a
JSON string with information about
the events. For more information, see
View scheduled events (p. 750).

events/recommendations/r The approximate time, in 2020-11-04

ebalance UTC, when the EC2 instance
rebalance recommendation
notification is emitted for the
instance. The following is an
example of the metadata for
this category: {"noticeTime":
"2020-11-05T08:22:00Z"}.
This category is available
only after the notification is
emitted. For more information,
see EC2 instance rebalance
recommendations (p. 449).

hostname The private IPv4 DNS hostname of Version 1.0

the instance. In cases where multiple
network interfaces are present, this
refers to the eth0 device (the device
for which the device number is 0).

iam/info If there is an IAM role associated with 2012-01-12

the instance, contains information
about the last time the instance
profile was updated, including
the instance's LastUpdated
date, InstanceProfileArn, and
InstanceProfileId. Otherwise, not
present.

iam/security-credentials/ If there is an IAM role associated 2012-01-12

role-name with the instance, role-name is the
name of the role, and role-name
contains the temporary security
credentials associated with the role
(for more information, see Retrieve
security credentials from instance
metadata (p. 1033)). Otherwise, not
present.

725
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Data Description Release date

identity-credentials/ec2/ [Internal use only] Information 2018-05-23

info about the credentials in identity-
credentials/ec2/security-
credentials/ec2-instance.
These credentials are used by AWS
features such as EC2 Instance
Connect, and do not have any
additional AWS API permissions or
privileges beyond identifying the
instance.

identity-credentials/ec2/ [Internal use only] Credentials 2018-05-23

security-credentials/ec2- that allow on-instance software to
instance identify itself to AWS to support
features such as EC2 Instance
Connect. These credentials do
not have any additional AWS API
permissions or privileges.

instance-action Notifies the instance that it should 2008-09-01

reboot in preparation for bundling.
Valid values: none | shutdown |
bundle-pending.

instance-id The ID of this instance. Version 1.0

instance-life-cycle The purchasing option of this 2019-10-01

instance. For more information, see
Instance purchasing options (p. 320).

instance-type The type of instance. For more 2007-08-29

information, see Instance
types (p. 203).

kernel-id The ID of the kernel launched with 2008-02-01

this instance, if applicable.

local-hostname The private IPv4 DNS hostname of 2007-01-19

the instance. In cases where multiple
network interfaces are present, this
refers to the eth0 device (the device
for which the device number is 0).

local-ipv4 The private IPv4 address of the Version 1.0

instance. In cases where multiple
network interfaces are present, this
refers to the eth0 device (the device
for which the device number is 0).

mac The instance's media access control 2011-01-01

(MAC) address. In cases where
multiple network interfaces are
present, this refers to the eth0 device
(the device for which the device
number is 0).

metrics/vhostmd No longer available. 2011-05-01

726
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Data Description Release date

network/interfaces/macs/ The unique device number associated 2011-01-01

mac/device-number with that interface. The device
number corresponds to the device
name; for example, a device-
number of 2 is for the eth2 device.
This category corresponds to the
DeviceIndex and device-index
fields that are used by the Amazon
EC2 API and the EC2 commands for
the AWS CLI.

network/interfaces/macs/ The ID of the network interface. 2011-01-01

mac/interface-id

network/interfaces/macs/ The private IPv4 addresses that 2011-01-01

mac/ipv4-associations/pu are associated with each public
blic-ip IP address and assigned to that
interface.

network/interfaces/macs/ The IPv6 addresses associated with 2016-06-30

mac/ipv6s the interface. Returned only for
instances launched into a VPC.

network/interfaces/macs/ The interface's local hostname. 2011-01-01

mac/local-hostname

network/interfaces/macs/ The private IPv4 addresses 2011-01-01

mac/local-ipv4s associated with the interface.

network/interfaces/macs/ The instance's MAC address. 2011-01-01

mac/mac

network/interfaces/ The index of the network card. Some 2020-11-01

macs/mac/network-card-index instance types support multiple
network cards.

network/interfaces/macs/ The ID of the owner of the network 2011-01-01

mac/owner-id interface. In multiple-interface
environments, an interface can be
attached by a third party, such as
Elastic Load Balancing. Traffic on
an interface is always billed to the
interface owner.

network/interfaces/macs/ The interface's public DNS (IPv4). 2011-01-01

mac/public-hostname This category is only returned if the
enableDnsHostnames attribute is
set to true. For more information,
see Using DNS with Your VPC.

network/interfaces/macs/ The public IP address or Elastic 2011-01-01

mac/public-ipv4s IP addresses associated with the
interface. There may be multiple IPv4
addresses on an instance.

network/interfaces/macs/ Security groups to which the network 2011-01-01

mac/security-groups interface belongs.

727
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Data Description Release date

network/interfaces/macs/ The IDs of the security groups to 2011-01-01

mac/security-group-ids which the network interface belongs.

network/interfaces/macs/ The ID of the subnet in which the 2011-01-01

mac/subnet-id interface resides.

network/interfaces/macs/ The IPv4 CIDR block of the subnet in 2011-01-01

mac/subnet-ipv4-cidr-block which the interface resides.

network/interfaces/macs/ The IPv6 CIDR block of the subnet in 2016-06-30

mac/subnet-ipv6-cidr-blocks which the interface resides.

network/interfaces/macs/ The ID of the VPC in which the 2011-01-01

mac/vpc-id interface resides.

network/interfaces/macs/ The primary IPv4 CIDR block of the 2011-01-01

mac/vpc-ipv4-cidr-block VPC.

network/interfaces/macs/ The IPv4 CIDR blocks for the VPC. 2016-06-30

mac/vpc-ipv4-cidr-blocks

network/interfaces/macs/ The IPv6 CIDR block of the VPC in 2016-06-30

mac/vpc-ipv6-cidr-blocks which the interface resides.

placement/availability-zone The Availability Zone in which the 2008-02-01

instance launched.

placement/availability-z The static Availability Zone ID in 2020-08-24

one-id which the instance is launched. The
Availability Zone ID is consistent
across accounts. However, it might be
different from the Availability Zone,
which can vary by account.

placement/group-name The name of the placement group in 2020-08-24

which the instance is launched.

placement/host-id The ID of the host on which the 2020-08-24

instance is launched. Applicable only
to Dedicated Hosts.

placement/partition-number The number of the partition in which 2020-08-24

the instance is launched.

placement/region The AWS Region in which the 2020-08-24

instance is launched.

product-codes AWS Marketplace product codes 2007-03-01

associated with the instance, if any.

public-hostname The instance's public DNS. This 2007-01-19

category is only returned if the
enableDnsHostnames attribute is
set to true. For more information,
see Using DNS with Your VPC in the
Amazon VPC User Guide.

728
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Data Description Release date

public-ipv4 The public IPv4 address. If an Elastic 2007-01-19

IP address is associated with the
instance, the value returned is the
Elastic IP address.

public-keys/0/openssh-key Public key. Only available if supplied Version 1.0

at instance launch time.

ramdisk-id The ID of the RAM disk specified at 2007-10-10

launch time, if applicable.

reservation-id The ID of the reservation. Version 1.0

security-groups The names of the security groups Version 1.0

applied to the instance.

After launch, you can change the

security groups of the instances.
Such changes are reflected here and
in network/interfaces/macs/mac/
security-groups.

services/domain The domain for AWS resources for 2014-02-25

the Region.

services/partition The partition that the resource is 2015-10-20

in. For standard AWS Regions, the
partition is aws. If you have resources
in other partitions, the partition is
aws-partitionname. For example,
the partition for resources in the
China (Beijing) Region is aws-cn.

spot/instance-action The action (hibernate, stop, or 2016-11-15

terminate) and the approximate
time, in UTC, when the action will
occur. This item is present only if
the Spot Instance has been marked
for hibernate, stop, or terminate.
For more information, see instance-
action (p. 458).

spot/termination-time The approximate time, in UTC, that 2014-11-05

the operating system for your Spot
Instance will receive the shutdown
signal. This item is present and
contains a time value (for example,
2015-01-05T18:02:00Z) only if the
Spot Instance has been marked
for termination by Amazon EC2.
The termination-time item is not
set to a time if you terminated
the Spot Instance yourself. For
more information, see termination-
time (p. 458).

729
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Dynamic data categories

The following table lists the categories of dynamic data.

Data Description Release

date

fws/instance- Value showing whether the customer has enabled detailed 2009-04-04
monitoring one-minute monitoring in CloudWatch. Valid values:
enabled | disabled

instance-identity/ JSON containing instance attributes, such as instance- 2009-04-04

document id, private IP address, etc. See Instance identity
documents (p. 730).

instance-identity/ Used to verify the document's authenticity and 2009-04-04

pkcs7 content against the signature. See Instance identity
documents (p. 730).

instance-identity/ Data that can be used by other parties to verify 2009-04-04

signature its origin and authenticity. See Instance identity
documents (p. 730).

Instance identity documents

Each instance that you launch has an instance identity document that provides information about the
instance itself. You can use the instance identity document to validate the attributes of the instance.

The instance identity document is generated when the instance is launched and it is exposed (in plaintext
JSON format) through the Instance Metadata Service. The IP address 169.254.169.254 is a link-local
address and is valid only from the instance. For more information, see Link-local address on Wikipedia.

You can retrieve the instance identity document from a running instance at any time. The instance
identity document includes the following information:

Data Description

devpayProductCodes Deprecated.

The AWS Marketplace product code of the AMI used to launch the instance.
marketplaceProductCodes

availabilityZone The Availability Zone in which the instance is running.

privateIp The private IPv4 address of the instance.

version The version of the instance identity document format.

instanceId The ID of the instance.

billingProducts The billing product code of the AMI used to launch the instance.

instanceType The instance type of the instance.

accountId The ID of the AWS account that launched the instance.

imageId The ID of the AMI used to launch the instance.

pendingTime The date and time that the instance was launched.

730
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

Data Description

architecture The architecture of the AMI used to launch the instance (i386 | x86_64 |
arm64).

kernelId The ID of the kernel associated with the instance, if applicable.

ramdiskId The ID of the RAM disk associated with the instance, if applicable.

region The Region in which the instance is running.

Retrieve the plaintext instance identity document

To retrieve the plaintext instance identity document

Connect to the instance and run one of the following commands depending on the Instance Metadata
Service (IMDS) version used by the instance.

IMDSv2

$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-metadata-

token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/dynamic/
instance-identity/document

IMDSv1

$ curl http://169.254.169.254/latest/dynamic/instance-identity/document

The following is example output.

{
"devpayProductCodes" : null,
"marketplaceProductCodes" : [ "1abc2defghijklm3nopqrs4tu" ],
"availabilityZone" : "us-west-2b",
"privateIp" : "10.158.112.84",
"version" : "2017-09-30",
"instanceId" : "i-1234567890abcdef0",
"billingProducts" : null,
"instanceType" : "t2.micro",
"accountId" : "123456789012",
"imageId" : "ami-5fb8c835",
"pendingTime" : "2016-11-19T16:32:11Z",
"architecture" : "x86_64",
"kernelId" : null,
"ramdiskId" : null,
"region" : "us-west-2"
}

Verify the instance identity document

If you intend to use the contents of the instance identity document for an important purpose, you should
verify its contents and authenticity before using it.

The plaintext instance identity document is accompanied by three hashed and encrypted signatures. You
can use these signatures to verify the origin and authenticity of the instance identity document and the
information that it includes. The following signatures are provided:

731
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

• Base64-encoded signature—This is a base64-encoded SHA256 hash of the instance identity document

that is encrypted using an RSA key pair.
• PKCS7 signature—This is a SHA1 hash of the instance identity document that is encrypted using a DSA
key pair.
• RSA-2048 signature—This is a SHA256 hash of the instance identity document that is encrypted using
an RSA-2048 key pair.

Each signature is available at a different endpoint in the instance metadata. You can use any one of these
signatures depending on your hashing and encryption requirements. To verify the signatures, you must
use the corresponding AWS public certificate.
Important
To validate the instance identity document using the base64-encoded signature or RSA2048
signature, you must request the corresponding AWS public certificate from AWS Support.

The following topics provide detailed steps for validating the instance identity document using each
signature.

• Use the PKCS7 signature to verify the instance identity document (p. 732)
• Use the base64-encoded signature to verify the instance identity document (p. 735)
• Use the RSA-2048 signature to verify the instance identity document (p. 736)

Use the PKCS7 signature to verify the instance identity document

This topic explains how to verify the instance identity document using the PKCS7 signature and the AWS
DSA public certificate.

To verify the instance identity document using the PKCS7 signature and the AWS DSA public
certificate

1. Connect to the instance.

2. Retrieve the PKCS7 signature from the instance metadata and add it to a file named pkcs7.

a. Add the -----BEGIN PKCS7----- header to the pkcs7 file.

$ echo "-----BEGIN PKCS7-----" > pkcs7

b. Retrieve the PKCS7 signature from the instance metadata and append it to the pkcs7 file. Use
one of the following commands depending on the IMDS version used by the instance.

IMDSv2

$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-

metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/
dynamic/instance-identity/pkcs7 >> pkcs7

IMDSv1

$ curl -s http://169.254.169.254/latest/dynamic/instance-identity/pkcs7
>> pkcs7

c. Append the -----END PKCS7----- footer to a new line in the pkcs7 file.

$ echo "" >> pkcs7

732
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

$ echo "-----END PKCS7-----" >> pkcs7

3. Add the contents of the instance identity document from the instance metadata to a file named
document. Use one of the following commands depending on the IMDS version used by the
instance.

IMDSv2

$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-

metadata-token-ttl-seconds: 21600"` \ && curl -H "X-aws-ec2-metadata-token: $TOKEN"
-v http://169.254.169.254/latest/dynamic/instance-identity/document > document

IMDSv1

$ curl -s http://169.254.169.254/latest/dynamic/instance-identity/document
> document

4. Add the AWS DSA public certificate to a file named certificate.

a. Create the certificate file.

$ touch certificate

b. Open the certificate file using your preferred text editor and add the contents of the AWS
DSA public certificate. Choose the correct certificate for the AWS Region that your instance is in.
Important
If the AWS DSA public certificate for your Region is not listed below, contact AWS
Support.

Other AWS Regions

The following AWS public certificate is for all AWS Regions, except Hong Kong, Bahrain,
China, and GovCloud.

-----BEGIN CERTIFICATE-----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-----END CERTIFICATE-----

Hong Kong Region

The AWS public certificate for the Hong Kong Region is as follows.

-----BEGIN CERTIFICATE-----
MIIC7zCCAq4CCQCO7MJe5Y3VLjAJBgcqhkjOOAQDMFwxCzAJBgNVBAYTAlVTMRkw

733
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data
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-----END CERTIFICATE-----

Bahrain Region

The AWS public certificate for the Bahrain Region is as follows.

-----BEGIN CERTIFICATE-----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-----END CERTIFICATE-----

GovCloud Regions

The AWS public certificate for the AWS GovCloud Regions is as follows.

-----BEGIN CERTIFICATE-----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-----END CERTIFICATE-----

734
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

China Regions

The AWS public certificate for the China (Beijing) and China (Ningxia) Regions is as follows.

-----BEGIN CERTIFICATE-----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-----END CERTIFICATE-----

c. Save and close the file.

5. Use the OpenSSL smime command to verify the signature. Include the -verify option to indicate
that the signature needs to be verified, and the -noverify option to indicate that the certificate
does not need to be verified.

$ openssl smime -verify -in pkcs7 -inform PEM -content document -certfile certificate -
noverify

If the signature is valid, the Verification successful message appears. If the signature cannot
be verified, contact AWS Support.

Use the base64-encoded signature to verify the instance identity document

This topic explains how to verify the instance identity document using the base64-encoded signature
and the AWS RSA public certificate.
Important
To validate the instance identity document using the base64-encoded signature, you must
request the AWS RSA public certificate from AWS Support.

To validate the instance identity document using the base64-encoded signature and the AWS
RSA public certificate

1. Connect to the instance.

2. Retrieve the base64-encoded signature from the instance metadata, convert it to binary, and add
it to a file named signature. Use one of the following commands depending on the IMDS version
used by the instance.

IMDSv2

$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-

metadata-token-ttl-seconds: 21600"` \ && curl -H "X-aws-ec2-metadata-token: $TOKEN"

735
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

-v http://169.254.169.254/latest/dynamic/instance-identity/signature | base64 -d
> signature

IMDSv1

$ curl -s http://169.254.169.254/latest/dynamic/instance-identity/signature |
base64 -d > signature

3. Retrieve the plaintext instance identity document from the instance metadata and add it to a file
named document. Use one of the following commands depending on the IMDS version used by the
instance.

IMDSv2

$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-

metadata-token-ttl-seconds: 21600"` \ && curl -H "X-aws-ec2-metadata-token: $TOKEN"
-v http://169.254.169.254/latest/dynamic/instance-identity/document > document

IMDSv1

$ curl -s http://169.254.169.254/latest/dynamic/instance-identity/document
> document

4. Add the AWS RSA public certificate that you received from AWS Support to a file named
certificate.
5. Extract the public key from the certificate that you received from AWS Support and save it to a file
named key.

$ openssl x509 -pubkey -noout -in certificate > key

6. Use OpenSSL dgst command to verify the instance identity document.

$ openssl dgst -sha256 -verify key -signature signature document

If the signature is valid, the Verified OK message appears. If the signature cannot be verified,
contact AWS Support.

Use the RSA-2048 signature to verify the instance identity document

This topic explains how to verify the instance identity document using the RSA-2048 signature and the
AWS RSA-2048 public certificate.
Important
To validate the instance identity document using the RSA-2048 signature, you must request the
AWS RSA-2048 public certificate from AWS Support.

To verify the instance identity document using the RSA-2048 signature and the AWS
RSA-2048 public certificate

1. Connect to the instance.

2. Retrieve the RSA-2048 signature from the instance metadata and add it to a file named rsa2048.

a. Add the -----BEGIN PKCS7----- header to the rsa2048 file.

$ echo "-----BEGIN PKCS7-----" > rsa2048

736
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance metadata and user data

b. Retrieve the RSA-2048 signature from the instance metadata and append it to the rsa2048 file.
Use one of the following commands depending on the IMDS version used by the instance.

IMDSv2

$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-

metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/
dynamic/instance-identity/rsa2048 >> rsa2048

IMDSv1

$ curl -s http://169.254.169.254/latest/dynamic/instance-identity/rsa2048
>> rsa2048

c. Append the -----END PKCS7----- footer to a new line in the rsa2048 file.

$ echo "" >> rsa2048

$ echo "-----END PKCS7-----" >> rsa2048

3. Add the contents of the instance identity document from the instance metadata to a file named
document. Use one of the following commands depending on the IMDS version used by the
instance.

IMDSv2

$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-

metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/
dynamic/instance-identity/document > document

IMDSv1

$ curl -s http://169.254.169.254/latest/dynamic/instance-identity/document
> document

4. Add the AWS RSA-2048 public certificate to a file named certificate.

a. Create the certificate file.

$ touch certificate

b. Open the certificate file using your preferred text editor and add the contents of the AWS
RSA-2048 public certificate that you received from AWS Support.
c. Save and close the file.
5. Use the OpenSSL smime command to verify the signature. Include the -verify option to indicate
that the signature needs to be verified, and the -noverify option to indicate that the certificate
does not need to be verified.

$ openssl smime -verify -in rsa2048 -inform PEM -content document -certfile certificate
-noverify

If the signature is valid, the Verification successful message appears. If the signature cannot
be verified, contact AWS Support.

737
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Elastic Inference

Amazon Elastic Inference

Amazon Elastic Inference (EI) is a resource you can attach to your Amazon EC2 CPU instances to
accelerate your deep learning (DL) inference workloads. Amazon EI accelerators come in multiple sizes
and are a cost-effective method to build intelligent capabilities into applications running on Amazon EC2
instances.

Amazon EI distributes model operations defined by TensorFlow, Apache MXNet, and the Open Neural
Network Exchange (ONNX) format through MXNet between low-cost, DL inference accelerators and the
CPU of the instance.

For more information about Amazon Elastic Inference, see the Amazon EI Developer Guide.

Identify EC2 Linux instances

Your application might need to determine whether it is running on an EC2 instance.

For information about identifying Windows instances, see Identify EC2 Windows Instances in the Amazon
EC2 User Guide for Windows Instances.

Inspect the instance identity document

For a definitive and cryptographically verified method of identifying an EC2 instance, check the instance
identity document, including its signature. These documents are available on every EC2 instance at the
local, non-routable address http://169.254.169.254/latest/dynamic/instance-identity/.
For more information, see Instance identity documents (p. 730).

Inspect the system UUID

You can get the system UUID and look for the presence of the characters "ec2" or "EC2" in the beginning
octet of the UUID. This method to determine whether a system is an EC2 instance is quick but potentially
inaccurate because there is a small chance that a system that is not an EC2 instance could have a UUID
that starts with these characters. Furthermore, for EC2 instances that are not using Amazon Linux, the
distribution's implementation of SMBIOS might represent the UUID in little-endian format, therefore the
"EC2" characters do not appear at the beginning of the UUID.

Example : Get the UUID from the hypervisor

If /sys/hypervisor/uuid exists, you can use the following command:

[ec2-user ~]$ cat /sys/hypervisor/uuid

In the following example output, the UUID starts with "ec2", which indicates that the system is probably
an EC2 instance.

ec2e1916-9099-7caf-fd21-012345abcdef

Example : Get the UUID from DMI (HVM instances only)

On HVM instances only, you can use the Desktop Management Interface (DMI).

You can use the dmidecode tool to return the UUID. On Amazon Linux, use the following command to
install the dmidecode tool if it's not already installed on your instance:

738
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Inspect the system UUID

[ec2-user ~]$ sudo yum install dmidecode -y

Then run the following command:

[ec2-user ~]$ sudo dmidecode --string system-uuid

Alternatively, use the following command:

[ec2-user ~]$ sudo cat /sys/devices/virtual/dmi/id/product_uuid

In the following example output, the UUID starts with "EC2", which indicates that the system is probably
an EC2 instance.

EC2E1916-9099-7CAF-FD21-01234ABCDEF

In the following example output, the UUID is represented in little-endian format.

45E12AEC-DCD1-B213-94ED-01234ABCDEF

On Nitro instances, the following command can be used:

[ec2-user ~]$ cat /sys/devices/virtual/dmi/id/board_asset_tag

This returns the instance ID, which is unique to EC2 instances:

i-0af01c0123456789a

739
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Monitor Amazon EC2

Monitoring is an important part of maintaining the reliability, availability, and performance of your
Amazon Elastic Compute Cloud (Amazon EC2) instances and your AWS solutions. You should collect
monitoring data from all of the parts in your AWS solutions so that you can more easily debug a multi-
point failure if one occurs. Before you start monitoring Amazon EC2, however, you should create a
monitoring plan that should include:

• What are your goals for monitoring?

• What resources will you monitor?
• How often will you monitor these resources?
• What monitoring tools will you use?
• Who will perform the monitoring tasks?
• Who should be notified when something goes wrong?

After you have defined your monitoring goals and have created your monitoring plan, the next step is
to establish a baseline for normal Amazon EC2 performance in your environment. You should measure
Amazon EC2 performance at various times and under different load conditions. As you monitor Amazon
EC2, you should store a history of monitoring data that you've collected. You can compare current
Amazon EC2 performance to this historical data to help you to identify normal performance patterns
and performance anomalies, and devise methods to address them. For example, you can monitor CPU
utilization, disk I/O, and network utilization for your EC2 instances. When performance falls outside your
established baseline, you might need to reconfigure or optimize the instance to reduce CPU utilization,
improve disk I/O, or reduce network traffic.

To establish a baseline you should, at a minimum, monitor the following items:

Item to monitor Amazon EC2 metric Monitoring agent/CloudWatch

Logs

CPU utilization CPUUtilization (p. 763)  

Network utilization NetworkIn (p. 763)  

NetworkOut (p. 763)

Disk performance DiskReadOps (p. 763)  

DiskWriteOps (p. 763)

Disk Reads/Writes DiskReadBytes (p. 763)  

DiskWriteBytes (p. 763)

Memory utilization, disk   [Linux and Windows Server

swap utilization, disk space instances] Collect Metrics
utilization, page file utilization, and Logs from Amazon EC2
log collection Instances and On-Premises
Servers with the CloudWatch
Agent

[Migration from previous

CloudWatch Logs agent on

740
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Automated and manual monitoring

Item to monitor Amazon EC2 metric Monitoring agent/CloudWatch

Logs
Windows Server instances]
Migrate Windows Server
Instance Log Collection to the
CloudWatch Agent

Automated and manual monitoring

AWS provides various tools that you can use to monitor Amazon EC2. You can configure some of these
tools to do the monitoring for you, while some of the tools require manual intervention.

Monitoring tools
• Automated monitoring tools (p. 741)
• Manual monitoring tools (p. 742)

Automated monitoring tools

You can use the following automated monitoring tools to watch Amazon EC2 and report back to you
when something is wrong:

• System status checks – monitor the AWS systems required to use your instance to ensure that they
are working properly. These checks detect problems with your instance that require AWS involvement
to repair. When a system status check fails, you can choose to wait for AWS to fix the issue or you can
resolve it yourself (for example, by stopping and restarting or terminating and replacing an instance).
Examples of problems that cause system status checks to fail include:
• Loss of network connectivity
• Loss of system power
• Software issues on the physical host
• Hardware issues on the physical host that impact network reachability

For more information, see Status checks for your instances (p. 743).
• Instance status checks – monitor the software and network configuration of your individual instance.
These checks detect problems that require your involvement to repair. When an instance status check
fails, typically you will need to address the problem yourself (for example, by rebooting the instance
or by making modifications in your operating system). Examples of problems that may cause instance
status checks to fail include:
• Failed system status checks
• Misconfigured networking or startup configuration
• Exhausted memory
• Corrupted file system
• Incompatible kernel

For more information, see Status checks for your instances (p. 743).
• Amazon CloudWatch alarms – watch a single metric over a time period you specify, and perform
one or more actions based on the value of the metric relative to a given threshold over a number
of time periods. The action is a notification sent to an Amazon Simple Notification Service (Amazon
SNS) topic or Amazon EC2 Auto Scaling policy. Alarms invoke actions for sustained state changes only.
CloudWatch alarms will not invoke actions simply because they are in a particular state; the state

741
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Manual monitoring tools

must have changed and been maintained for a specified number of periods. For more information, see
Monitor your instances using CloudWatch (p. 761).
• Amazon CloudWatch Events – automate your AWS services and respond automatically to system
events. Events from AWS services are delivered to CloudWatch Events in near real time, and you can
specify automated actions to take when an event matches a rule you write. For more information, see
What is Amazon CloudWatch Events?.
• Amazon CloudWatch Logs – monitor, store, and access your log files from Amazon EC2 instances, AWS
CloudTrail, or other sources. For more information, see the Amazon CloudWatch Logs User Guide.
• CloudWatch agent – collect logs and system-level metrics from both hosts and guests on your
EC2 instances and on-premises servers. For more information, see Collecting Metrics and Logs
from Amazon EC2 Instances and On-Premises Servers with the CloudWatch Agent in the Amazon
CloudWatch User Guide.
• AWS Management Pack for Microsoft System Center Operations Manager – links Amazon EC2
instances and the Windows or Linux operating systems running inside them. The AWS Management
Pack is an extension to Microsoft System Center Operations Manager. It uses a designated computer in
your datacenter (called a watcher node) and the Amazon Web Services APIs to remotely discover and
collect information about your AWS resources. For more information, see AWS Management Pack for
Microsoft System Center.

Manual monitoring tools

Another important part of monitoring Amazon EC2 involves manually monitoring those items that the
monitoring scripts, status checks, and CloudWatch alarms don't cover. The Amazon EC2 and CloudWatch
console dashboards provide an at-a-glance view of the state of your Amazon EC2 environment.

• Amazon EC2 Dashboard shows:

• Service Health and Scheduled Events by Region
• Instance state
• Status checks
• Alarm status
• Instance metric details (In the navigation pane choose Instances, select an instance, and choose the
Monitoring tab)
• Volume metric details (In the navigation pane choose Volumes, select a volume, and choose the
Monitoring tab)
• Amazon CloudWatch Dashboard shows:
• Current alarms and status
• Graphs of alarms and resources
• Service health status

In addition, you can use CloudWatch to do the following:

• Graph Amazon EC2 monitoring data to troubleshoot issues and discover trends
• Search and browse all your AWS resource metrics
• Create and edit alarms to be notified of problems
• See at-a-glance overviews of your alarms and AWS resources

Best practices for monitoring

Use the following best practices for monitoring to help you with your Amazon EC2 monitoring tasks.

• Make monitoring a priority to head off small problems before they become big ones.

742
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Monitor the status of your instances

• Create and implement a monitoring plan that collects monitoring data from all of the parts in your
AWS solution so that you can more easily debug a multi-point failure if one occurs. Your monitoring
plan should address, at a minimum, the following questions:
• What are your goals for monitoring?
• What resources you will monitor?
• How often you will monitor these resources?
• What monitoring tools will you use?
• Who will perform the monitoring tasks?
• Who should be notified when something goes wrong?
• Automate monitoring tasks as much as possible.
• Check the log files on your EC2 instances.

Monitor the status of your instances

You can monitor the status of your instances by viewing status checks and scheduled events for your
instances.

A status check gives you the information that results from automated checks performed by Amazon EC2.
These automated checks detect whether specific issues are affecting your instances. The status check
information, together with the data provided by Amazon CloudWatch, gives you detailed operational
visibility into each of your instances.

You can also see status of specific events that are scheduled for your instances. The status of events
provides information about upcoming activities that are planned for your instances, such as rebooting or
retirement. They also provide the scheduled start and end time of each event.

Contents
• Status checks for your instances (p. 743)
• Scheduled events for your instances (p. 750)

Status checks for your instances

With instance status monitoring, you can quickly determine whether Amazon EC2 has detected
any problems that might prevent your instances from running applications. Amazon EC2 performs
automated checks on every running EC2 instance to identify hardware and software issues. You can view
the results of these status checks to identify specific and detectable problems. The event status data
augments the information that Amazon EC2 already provides about the state of each instance (such as
pending, running, stopping) and the utilization metrics that Amazon CloudWatch monitors (CPU
utilization, network traffic, and disk activity).

Status checks are performed every minute, returning a pass or a fail status. If all checks pass, the overall
status of the instance is OK. If one or more checks fail, the overall status is impaired. Status checks are
built into Amazon EC2, so they cannot be disabled or deleted.

When a status check fails, the corresponding CloudWatch metric for status checks is incremented. For
more information, see Status check metrics (p. 769). You can use these metrics to create CloudWatch
alarms that are triggered based on the result of the status checks. For example, you can create an alarm
to warn you if status checks fail on a specific instance. For more information, see Create and edit status
check alarms (p. 747).

You can also create an Amazon CloudWatch alarm that monitors an Amazon EC2 instance and
automatically recovers the instance if it becomes impaired due to an underlying issue. For more
information, see Recover your instance (p. 648).

743
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance status checks

Contents
• Types of status checks (p. 744)
• View status checks (p. 744)
• Report instance status (p. 747)
• Create and edit status check alarms (p. 747)

Types of status checks

There are two types of status checks: system status checks and instance status checks.

System status checks

System status checks monitor the AWS systems on which your instance runs. These checks detect
underlying problems with your instance that require AWS involvement to repair. When a system status
check fails, you can choose to wait for AWS to fix the issue, or you can resolve it yourself. For instances
backed by Amazon EBS, you can stop and start the instance yourself, which in most cases results in the
instance being migrated to a new host. For instances backed by instance store, you can terminate and
replace the instance.

The following are examples of problems that can cause system status checks to fail:

• Loss of network connectivity

• Loss of system power
• Software issues on the physical host
• Hardware issues on the physical host that impact network reachability

Instance status checks

Instance status checks monitor the software and network configuration of your individual instance.
Amazon EC2 checks the health of the instance by sending an address resolution protocol (ARP) request
to the network interface (NIC). These checks detect problems that require your involvement to repair.
When an instance status check fails, you typically must address the problem yourself (for example, by
rebooting the instance or by making instance configuration changes).

The following are examples of problems that can cause instance status checks to fail:

• Failed system status checks

• Incorrect networking or startup configuration
• Exhausted memory
• Corrupted file system
• Incompatible kernel

View status checks

Amazon EC2 provides you with several ways to view and work with status checks.

View status using the console

You can view status checks using the AWS Management Console.

744
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance status checks

New console

To view status checks (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. On the Instances page, the Status check column lists the operational status of each instance.
4. To view the status of a specific instance, select the instance, and then choose the Status Checks
tab.

If you have an instance with a failed status check and the instance has been unreachable for
over 20 minutes, choose Open support case to submit a request for assistance. To troubleshoot
system or instance status check failures yourself, see Troubleshoot instances with failed status
checks (p. 1342).
5. To review the CloudWatch metrics for status checks, select the instance, and then choose the
Monitoring tab. Scroll until you see the graphs for the following metrics:

• Status check failed (any)

• Status check failed (instance)
• Status check failed (system)

Old console

To view status checks (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. On the Instances page, the Status Checks column lists the operational status of each instance.
4. To view the status of a specific instance, select the instance, and then choose the Status Checks
tab.

745
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance status checks

If you have an instance with a failed status check and the instance has been unreachable for
over 20 minutes, choose AWS Support to submit a request for assistance. To troubleshoot
system or instance status check failures yourself, see Troubleshoot instances with failed status
checks (p. 1342).
5. To review the CloudWatch metrics for status checks, select the instance, and then choose the
Monitoring tab. Scroll until you see the graphs for the following metrics:

• Status Check Failed (Any)

• Status Check Failed (Instance)
• Status Check Failed (System)

View status using the command line

You can view status checks for running instances using the describe-instance-status (AWS CLI) command.

To view the status of all instances, use the following command.

aws ec2 describe-instance-status

To get the status of all instances with an instance status of impaired, use the following command.

aws ec2 describe-instance-status \

--filters Name=instance-status.status,Values=impaired

To get the status of a single instance, use the following command.

aws ec2 describe-instance-status \

--instance-ids i-1234567890abcdef0

Alternatively, use the following commands:

• Get-EC2InstanceStatus (AWS Tools for Windows PowerShell)

• DescribeInstanceStatus (Amazon EC2 Query API)

If you have an instance with a failed status check, see Troubleshoot instances with failed status
checks (p. 1342).

746
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance status checks

Report instance status

You can provide feedback if you are having problems with an instance whose status is not shown as
impaired, or if you want to send AWS additional details about the problems you are experiencing with an
impaired instance.

We use reported feedback to identify issues impacting multiple customers, but do not respond to
individual account issues. Providing feedback does not change the status check results that you currently
see for the instance.

Report status feedback using the console

New console

To report instance status (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance, choose the Status Checks tab, choose Actions (the second Actions menu in
the bottom half of the page), and then choose Report instance status.
4. Complete the Report instance status form, and then choose Submit.

Old console

To report instance status (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance, choose the Status Checks tab, and choose Submit feedback.
4. Complete the Report Instance Status form, and then choose Submit.

Report status feedback using the command line

Use the report-instance-status (AWS CLI) command to send feedback about the status of an impaired
instance.

aws ec2 report-instance-status \

--instances i-1234567890abcdef0 \
--status impaired \
--reason-codes code

Alternatively, use the following commands:

• Send-EC2InstanceStatus (AWS Tools for Windows PowerShell)

• ReportInstanceStatus (Amazon EC2 Query API)

Create and edit status check alarms

You can use the status check metrics (p. 769) to create CloudWatch alarms to notify you when an
instance has a failed status check.

747
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance status checks

Create a status check alarm using the console

Use the following procedure to configure an alarm that sends you a notification by email, or stops,
terminates, or recovers an instance when it fails a status check.

New console

To create a status check alarm (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance, choose the Status Checks tab, and choose Actions, Create status check
alarm.
4. On the Manage CloudWatch alarms page, under Add or edit alarm, choose Create a new
alarm.
5. For Alarm notification, turn the toggle on to configure Amazon Simple Notification Service
(Amazon SNS) notifications. Select an existing Amazon SNS topic or enter a name to create a
new topic.
6. For Alarm action, turn the toggle on to specify an action to take when the alarm is triggered.
Select the action that you'd like to take from the dropdown.
7. For Alarm thresholds, select the metric and criteria for the alarm. In Consecutive Period, set
the number of periods you want to evaluate and, in Period, enter the evaluation period duration
before triggering the alarm and sending an email.

For example, you can leave the default settings for Group samples by (Average) and Type of
data to sample (CPU utilization). You can set Alarm When to >= and enter 0.80 for Percent.
For Consecutive Period, you can enter 1. For Period, you can select 5 Minutes.
8. (Optional) For Sample metric data, choose Add to dashboard.
9. Choose Create.
Important
If you added an email address to the list of recipients or created a new topic, Amazon
SNS sends a subscription confirmation email message to each new address. Each
recipient must confirm the subscription by choosing the link contained in that message.
Alert notifications are sent only to confirmed addresses.

Old console

To create a status check alarm (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance, choose the Status Checks tab, and choose Create Status Check Alarm.
4. Select Send a notification to. Choose an existing SNS topic, or choose create topic to create
a new one. If creating a new topic, in With these recipients, enter your email address and the
addresses of any additional recipients, separated by commas.
5. (Optional) Select Take the action, and then select the action that you'd like to take.
6. In Whenever, select the status check that you want to be notified about.

If you selected Recover this instance in the previous step, select Status Check Failed (System).
7. In For at least, set the number of periods you want to evaluate and in consecutive periods,
select the evaluation period duration before triggering the alarm and sending an email.
8. (Optional) In Name of alarm, replace the default name with another name for the alarm.

748
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance status checks

9. Choose Create Alarm.

Important
If you added an email address to the list of recipients or created a new topic, Amazon
SNS sends a subscription confirmation email message to each new address. Each
recipient must confirm the subscription by choosing the link contained in that message.
Alert notifications are sent only to confirmed addresses.

If you need to make changes to an instance status alarm, you can edit it.

New console

To edit a status check alarm using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, Monitoring, Manage CloudWatch alarms.
4. On the Manage CloudWatch alarms page, under Add or edit alarm, choose Edit an existing
alarm.
5. For Search for alarm, choose the alarm to edit.
6. Make the desired changes, and then choose Update.

Old console

To edit a status check alarm using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, CloudWatch Monitoring, Add/Edit Alarms.
4. In the Alarm Details dialog box, choose the name of the alarm.
5. In the Edit Alarm dialog box, make the desired changes, and then choose Save.

Create a status check alarm using the AWS CLI

In the following example, the alarm publishes a notification to an SNS topic, arn:aws:sns:us-
west-2:111122223333:my-sns-topic, when the instance fails either the instance check or system
status check for at least two consecutive periods. The CloudWatch metric used is StatusCheckFailed.

To create a status check alarm using the AWS CLI

1. Select an existing SNS topic or create a new one. For more information, see Using the AWS CLI with
Amazon SNS in the AWS Command Line Interface User Guide.
2. Use the following list-metrics command to view the available Amazon CloudWatch metrics for
Amazon EC2.

aws cloudwatch list-metrics --namespace AWS/EC2

3. Use the following put-metric-alarm command to create the alarm.

aws cloudwatch put-metric-alarm --alarm-name StatusCheckFailed-Alarm-for-

i-1234567890abcdef0 --metric-name StatusCheckFailed --namespace AWS/EC2 --
statistic Maximum --dimensions Name=InstanceId,Value=i-1234567890abcdef0 --unit
Count --period 300 --evaluation-periods 2 --threshold 1 --comparison-operator

749
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled events

GreaterThanOrEqualToThreshold --alarm-actions arn:aws:sns:us-west-2:111122223333:my-

sns-topic

The period is the time frame, in seconds, in which Amazon CloudWatch metrics are collected. This
example uses 300, which is 60 seconds multiplied by 5 minutes. The evaluation period is the number
of consecutive periods for which the value of the metric must be compared to the threshold. This
example uses 2. The alarm actions are the actions to perform when this alarm is triggered. This
example configures the alarm to send an email using Amazon SNS.

Scheduled events for your instances

AWS can schedule events for your instances, such as a reboot, stop/start, or retirement. These events do
not occur frequently. If one of your instances will be affected by a scheduled event, AWS sends an email
to the email address that's associated with your AWS account prior to the scheduled event. The email
provides details about the event, including the start and end date. Depending on the event, you might be
able to take action to control the timing of the event.

Scheduled events are managed by AWS; you cannot schedule events for your instances. You can view the
events scheduled by AWS, customize scheduled event notifications to include or remove tags from the
email notification, perform actions when an instance is scheduled to reboot, retire, or stop.

To update the contact information for your account so that you can be sure to be notified about
scheduled events, go to the Account Settings page.

Contents
• Types of scheduled events (p. 750)
• View scheduled events (p. 750)
• Customize scheduled event notifications (p. 754)
• Work with instances scheduled to stop or retire (p. 757)
• Work with instances scheduled for reboot (p. 757)
• Work with instances scheduled for maintenance (p. 759)
• Reschedule a scheduled event (p. 759)

Types of scheduled events

Amazon EC2 can create the following types of events for your instances, where the event occurs at a
scheduled time:

• Instance stop: At the scheduled time, the instance is stopped. When you start it again, it's migrated to
a new host. Applies only to instances backed by Amazon EBS.
• Instance retirement: At the scheduled time, the instance is stopped if it is backed by Amazon EBS, or
terminated if it is backed by instance store.
• Instance reboot: At the scheduled time, the instance is rebooted.
• System reboot: At the scheduled time, the host for the instance is rebooted.
• System maintenance: At the scheduled time, the instance might be temporarily affected by network
maintenance or power maintenance.

View scheduled events

In addition to receiving notification of scheduled events in email, you can check for scheduled events
using one of the following methods.

750
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled events

New console

To view scheduled events for your instances using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. You can view scheduled events in the following screens:
• In the navigation pane, choose Events. Any resources with an associated event are
displayed. You can filter by Resource ID, Resource type, Availability zone, Event status, or
Event type.

• Alternatively, in the navigation pane, choose EC2 Dashboard. Any resources with an
associated event are displayed under Scheduled events.

• Some events are also shown for affected resources. For example, in the navigation pane,
choose Instances and select an instance. If the instance has an associated instance stop or
instance retirement event, it is displayed in the lower pane.

Old console

To view scheduled events for your instances using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. You can view scheduled events in the following screens:
• In the navigation pane, choose Events. Any resources with an associated event are
displayed. You can filter by resource type, or by specific event types. You can select the
resource to view details.

751
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled events

• Alternatively, in the navigation pane, choose EC2 Dashboard. Any resources with an
associated event are displayed under Scheduled Events.

• Some events are also shown for affected resources. For example, in the navigation pane,
choose Instances and select an instance. If the instance has an associated instance stop or
instance retirement event, it is displayed in the lower pane.

AWS CLI

To view scheduled events for your instances using the AWS CLI

Use the describe-instance-status command.

aws ec2 describe-instance-status \

--instance-id i-1234567890abcdef0 \
--query "InstanceStatuses[].Events"

The following example output shows a reboot event.

[
"Events": [
{
"InstanceEventId": "instance-event-0d59937288b749b32",
"Code": "system-reboot",
"Description": "The instance is scheduled for a reboot",
"NotAfter": "2019-03-15T22:00:00.000Z",
"NotBefore": "2019-03-14T20:00:00.000Z",
"NotBeforeDeadline": "2019-04-05T11:00:00.000Z"
}

752
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled events

]
]

The following example output shows an instance retirement event.

[
"Events": [
{
"InstanceEventId": "instance-event-0e439355b779n26",

"Code": "instance-stop",
"Description": "The instance is running on degraded hardware",
"NotBefore": "2015-05-23T00:00:00.000Z"
}
]
]

PowerShell

To view scheduled events for your instances using the AWS Tools for Windows PowerShell

Use the following Get-EC2InstanceStatus command.

PS C:\> (Get-EC2InstanceStatus -InstanceId i-1234567890abcdef0).Events

The following example output shows an instance retirement event.

Code : instance-stop
Description : The instance is running on degraded hardware
NotBefore : 5/23/2015 12:00:00 AM

Instance metadata

To view scheduled events for your instances using instance metadata

You can retrieve information about active maintenance events for your instances from the instance
metadata (p. 704) using Instance Metadata Service Version 2 or Instance Metadata Service Version 1.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" –v http://169.254.169.254/latest/meta-
data/events/maintenance/scheduled

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/events/maintenance/scheduled

The following is example output with information about a scheduled system reboot event, in JSON
format.

[
{
"NotBefore" : "21 Jan 2019 09:00:43 GMT",
"Code" : "system-reboot",
"Description" : "scheduled reboot",
"EventId" : "instance-event-0d59937288b749b32",
"NotAfter" : "21 Jan 2019 09:17:23 GMT",
"State" : "active"

753
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled events

}
]

To view event history about completed or canceled events for your instances using instance
metadata

You can retrieve information about completed or canceled events for your instances from instance
metadata (p. 704) using Instance Metadata Service Version 2 or Instance Metadata Service Version 1.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" –v http://169.254.169.254/latest/meta-
data/events/maintenance/history

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/events/maintenance/history

The following is example output with information about a system reboot event that was canceled,
and a system reboot event that was completed, in JSON format.

[
{
"NotBefore" : "21 Jan 2019 09:00:43 GMT",
"Code" : "system-reboot",
"Description" : "[Canceled] scheduled reboot",
"EventId" : "instance-event-0d59937288b749b32",
"NotAfter" : "21 Jan 2019 09:17:23 GMT",
"State" : "canceled"
},
{
"NotBefore" : "29 Jan 2019 09:00:43 GMT",
"Code" : "system-reboot",
"Description" : "[Completed] scheduled reboot",
"EventId" : "instance-event-0d59937288b749b32",
"NotAfter" : "29 Jan 2019 09:17:23 GMT",
"State" : "completed"
}
]

Customize scheduled event notifications

You can customize scheduled event notifications to include tags in the email notification. This makes it
easier to identify the affected resource (instances or Dedicated Hosts) and to prioritize actions for the
upcoming event.

When you customize event notifications to include tags, you can choose to include:

• All of the tags that are associated with the affected resource
• Only specific tags that are associated with the affected resource

For example, suppose that you assign application, costcenter, project, and owner tags to all
of your instances. You can choose to include all of the tags in event notifications. Alternatively, if you'd
like to see only the owner and project tags in event notifications, then you can choose to include only
those tags.

754
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled events

After you select the tags to include, the event notifications will include the resource ID (instance ID or
Dedicated Host ID) and the tag key and value pairs that are associated with the affected resource.

Topics
• Include tags in event notifications (p. 755)
• Remove tags from event notifications (p. 755)
• View the tags to be included in event notifications (p. 756)

Include tags in event notifications

The tags that you choose to include apply to all resources (instances and Dedicated Hosts) in the selected
Region. To customize event notifications in other Regions, first select the required Region and then
perform the following steps.

You can include tags in event notifications using one of the following methods.

New console

To include tags in event notifications

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Events.
3. Choose Actions, Manage event notifications.
4. Select Include resource tags in event notifications.
5. Do one of the following, depending on the tags that you want to include in event notifications:

• To include all of the tags associated with the affected instance or Dedicated Host, select
Include all resource tags.
• To manually select the tags to include, select Choose the tags to include, and then for
Choose the tags to include, enter the tag key and press Enter.
6. Choose Save.

AWS CLI

To include all tags in event notifications

Use the register-instance-event-notification-attributes AWS CLI command and set the

IncludeAllTagsOfInstance parameter to true.

aws ec2 register-instance-event-notification-attributes --instance-tag-attribute

"IncludeAllTagsOfInstance=true"

To include specific tags in event notifications

Use the register-instance-event-notification-attributes AWS CLI command and specify the tags to
include using the InstanceTagKeys parameter.

aws ec2 register-instance-event-notification-attributes --instance-tag-attribute

'InstanceTagKeys=["tag_key_1", "tag_key_2", "tag_key_3"]'

Remove tags from event notifications

You can remove tags from event notifications using one of the following methods.

755
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled events

New console

To remove tags from event notifications

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Events.
3. Choose Actions, Manage event notifications.
4. Do one of the following, depending on the tag that you want to remove from event
notifications.

• To remove all tags from event notifications, clear Include resource tags in event
notifications.
• To remove specific tags from event notifications, choose Remove (X) for the tags listed below
the Choose the tags to include field.
5. Choose Save.

AWS CLI

To remove all tags from event notifications

Use the deregister-instance-event-notification-attributes AWS CLI command and set the

IncludeAllTagsOfInstance parameter to false.

aws ec2 deregister-instance-event-notification-attributes --instance-tag-attribute

"IncludeAllTagsOfInstance=false"

To remove specific tags from event notifications

Use the deregister-instance-event-notification-attributes AWS CLI command and specify the tags to
remove using the InstanceTagKeys parameter.

aws ec2 deregister-instance-event-notification-attributes --instance-tag-attribute

'InstanceTagKeys=["tag_key_1", "tag_key_2", "tag_key_3"]'

View the tags to be included in event notifications

You can view the tags that are to be included in event notifications using one of the following methods.

New console

To view the tags that are to be included in event notifications

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Events.
3. Choose Actions, Manage event notifications.

AWS CLI

To view the tags that are to be included in event notifications

Use the describe-instance-event-notification-attributes AWS CLI command.

aws ec2 describe-instance-event-notification-attributes

756
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled events

Work with instances scheduled to stop or retire

When AWS detects irreparable failure of the underlying host for your instance, it schedules the instance
to stop or terminate, depending on the type of root device for the instance. If the root device is an EBS
volume, the instance is scheduled to stop. If the root device is an instance store volume, the instance is
scheduled to terminate. For more information, see Instance retirement (p. 640).
Important
Any data stored on instance store volumes is lost when an instance is stopped, hibernated, or
terminated. This includes instance store volumes that are attached to an instance that has an
EBS volume as the root device. Be sure to save data from your instance store volumes that you
might need later before the instance is stopped, hibernated, or terminated.

Actions for Instances Backed by Amazon EBS

You can wait for the instance to stop as scheduled. Alternatively, you can stop and start the instance
yourself, which migrates it to a new host. For more information about stopping your instance, in addition
to information about the changes to your instance configuration when it's stopped, see Stop and start
your instance (p. 623).

You can automate an immediate stop and start in response to a scheduled instance stop event. For more
information, see Automating Actions for EC2 Instances in the AWS Health User Guide.

Actions for Instances Backed by Instance Store

We recommend that you launch a replacement instance from your most recent AMI and migrate all
necessary data to the replacement instance before the instance is scheduled to terminate. Then, you can
terminate the original instance, or wait for it to terminate as scheduled.

Work with instances scheduled for reboot

When AWS must perform tasks such as installing updates or maintaining the underlying host, it
can schedule the instance or the underlying host for a reboot. You can reschedule most reboot
events (p. 759) so that your instance is rebooted at a specific date and time that suits you.

If you stop your linked EC2-Classic instance (p. 953), it is automatically unlinked from the VPC and the
VPC security groups are no longer associated with the instance. You can link your instance to the VPC
again after you've restarted it.

View the reboot event type

You can view whether a reboot event is an instance reboot or a system reboot using one of the following
methods.

New console

To view the type of scheduled reboot event using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Events.
3. Choose Resource type: instance from the filter list.
4. For each instance, view the value in the Event type column. The value is either system-reboot
or instance-reboot.

Old console

To view the type of scheduled reboot event using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

757
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled events

2. In the navigation pane, choose Events.

3. Choose Instance resources from the filter list.
4. For each instance, view the value in the Event Type column. The value is either system-reboot
or instance-reboot.

AWS CLI

To view the type of scheduled reboot event using the AWS CLI

Use the describe-instance-status command.

aws ec2 describe-instance-status --instance-id i-1234567890abcdef0

For scheduled reboot events, the value for Code is either system-reboot or instance-reboot.
The following example output shows a system-reboot event.

[
"Events": [
{
"InstanceEventId": "instance-event-0d59937288b749b32",
"Code": "system-reboot",
"Description": "The instance is scheduled for a reboot",
"NotAfter": "2019-03-14T22:00:00.000Z",
"NotBefore": "2019-03-14T20:00:00.000Z",
"NotBeforeDeadline": "2019-04-05T11:00:00.000Z"
}
]
]

Actions for instance reboot

You can wait for the instance reboot to occur within its scheduled maintenance window,
reschedule (p. 759) the instance reboot to a date and time that suits you, or reboot (p. 639) the
instance yourself at a time that is convenient for you.

After your instance is rebooted, the scheduled event is cleared and the event's description is updated.
The pending maintenance to the underlying host is completed, and you can begin using your instance
again after it has fully booted.

Actions for system reboot

It is not possible for you to reboot the system yourself. You can wait for the system reboot to occur
during its scheduled maintenance window, or you can reschedule (p. 759) the system reboot to a date
and time that suits you. A system reboot typically completes in a matter of minutes. After the system
reboot has occurred, the instance retains its IP address and DNS name, and any data on local instance
store volumes is preserved. After the system reboot is complete, the scheduled event for the instance is
cleared, and you can verify that the software on your instance is operating as expected.

Alternatively, if it is necessary to maintain the instance at a different time and you can't reschedule the
system reboot, then you can stop and start an Amazon EBS-backed instance, which migrates it to a new
host. However, the data on the local instance store volumes is not preserved. You can also automate
an immediate instance stop and start in response to a scheduled system reboot event. For more
information, see Automating Actions for EC2 Instances in the AWS Health User Guide. For an instance
store-backed instance, if you can't reschedule the system reboot, then you can launch a replacement
instance from your most recent AMI, migrate all necessary data to the replacement instance before the
scheduled maintenance window, and then terminate the original instance.

758
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled events

Work with instances scheduled for maintenance

When AWS must maintain the underlying host for an instance, it schedules the instance for maintenance.
There are two types of maintenance events: network maintenance and power maintenance.

During network maintenance, scheduled instances lose network connectivity for a brief period of time.
Normal network connectivity to your instance is restored after maintenance is complete.

During power maintenance, scheduled instances are taken offline for a brief period, and then rebooted.
When a reboot is performed, all of your instance's configuration settings are retained.

After your instance has rebooted (this normally takes a few minutes), verify that your application is
working as expected. At this point, your instance should no longer have a scheduled event associated
with it, or if it does, the description of the scheduled event begins with [Completed]. It sometimes takes
up to 1 hour for the instance status description to refresh. Completed maintenance events are displayed
on the Amazon EC2 console dashboard for up to a week.

Actions for Instances Backed by Amazon EBS

You can wait for the maintenance to occur as scheduled. Alternatively, you can stop and start the
instance, which migrates it to a new host. For more information about stopping your instance, in addition
to information about the changes to your instance configuration when it's stopped, see Stop and start
your instance (p. 623).

You can automate an immediate stop and start in response to a scheduled maintenance event. For more
information, see Automating Actions for EC2 Instances in the AWS Health User Guide.

Actions for instances backed by instance store

You can wait for the maintenance to occur as scheduled. Alternatively, if you want to maintain normal
operation during a scheduled maintenance window, you can launch a replacement instance from
your most recent AMI, migrate all necessary data to the replacement instance before the scheduled
maintenance window, and then terminate the original instance.

Reschedule a scheduled event

You can reschedule an event so that it occurs at a specific date and time that suits you. Only events that
have a deadline date can be rescheduled. There are other limitations for rescheduling an event (p. 761).

You can reschedule an event using one of the following methods.

New console

To reschedule an event using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Events.
3. Choose Resource type: instance from the filter list.
4. Select one or more instances, and then choose Actions, Schedule event.

Only events that have an event deadline date, indicated by a value for Deadline, can be
rescheduled. If one of the selected events does not have a deadline date, Actions, Schedule
event is disabled.
5. For New start time, enter a new date and time for the event. The new date and time must occur
before the Event deadline.
6. Choose Save.

759
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scheduled events

It might take 1-2 minutes for the updated event start time to be reflected in the console.

Old console

To reschedule an event using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Events.
3. Choose Instance resources from the filter list.
4. Select one or more instances, and then choose Actions, Schedule Event.

Only events that have an event deadline date, indicated by a value for Event Deadline, can be
rescheduled.
5. For Event start time, enter a new date and time for the event. The new date and time must
occur before the Event Deadline.
6. Choose Schedule Event.

It might take 1-2 minutes for the updated event start time to be reflected in the console.

AWS CLI

To reschedule an event using the AWS CLI

1. Only events that have an event deadline date, indicated by a value for NotBeforeDeadline,
can be rescheduled. Use the describe-instance-status command to view the
NotBeforeDeadline parameter value.

aws ec2 describe-instance-status --instance-id i-1234567890abcdef0

The following example output shows a system-reboot event that can be rescheduled because
NotBeforeDeadline contains a value.

[
"Events": [
{
"InstanceEventId": "instance-event-0d59937288b749b32",
"Code": "system-reboot",
"Description": "The instance is scheduled for a reboot",
"NotAfter": "2019-03-14T22:00:00.000Z",
"NotBefore": "2019-03-14T20:00:00.000Z",
"NotBeforeDeadline": "2019-04-05T11:00:00.000Z"
}
]
]

2. To reschedule the event, use the modify-instance-event-start-time command. Specify the new
event start time using the not-before parameter. The new event start time must fall before
the NotBeforeDeadline.

aws ec2 modify-instance-event-start-time --instance-id i-1234567890abcdef0

--instance-event-id instance-event-0d59937288b749b32 --not-
before 2019-03-25T10:00:00.000

It might take 1-2 minutes before the describe-instance-status command returns the updated
not-before parameter value.

760
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Monitor your instances using CloudWatch

Limitations
• Only events with an event deadline date can be rescheduled. The event can be rescheduled up to the
event deadline date. The Deadline column in the console and the NotBeforeDeadline field in the
AWS CLI indicate if the event has a deadline date.
• Only events that have not yet started can be rescheduled. The Start time column in the console and
the NotBefore field in the AWS CLI indicate the event start time. Events that are scheduled to start in
the next 5 minutes cannot be rescheduled.
• The new event start time must be at least 60 minutes from the current time.
• If you reschedule multiple events using the console, the event deadline date is determined by the
event with the earliest event deadline date.

Monitor your instances using CloudWatch

You can monitor your instances using Amazon CloudWatch, which collects and processes raw data
from Amazon EC2 into readable, near real-time metrics. These statistics are recorded for a period of 15
months, so that you can access historical information and gain a better perspective on how your web
application or service is performing.

By default, Amazon EC2 sends metric data to CloudWatch in 5-minute periods. To send metric data for
your instance to CloudWatch in 1-minute periods, you can enable detailed monitoring on the instance.
For more information, see Enable or turn off detailed monitoring for your instances (p. 761).

The Amazon EC2 console displays a series of graphs based on the raw data from Amazon CloudWatch.
Depending on your needs, you might prefer to get data for your instances from Amazon CloudWatch
instead of the graphs in the console.

For more information about Amazon CloudWatch, see the Amazon CloudWatch User Guide.

Contents
• Enable or turn off detailed monitoring for your instances (p. 761)
• List the available CloudWatch metrics for your instances (p. 763)
• Get statistics for metrics for your instances (p. 774)
• Graph metrics for your instances (p. 782)
• Create a CloudWatch alarm for an instance (p. 782)
• Create alarms that stop, terminate, reboot, or recover an instance (p. 784)

Enable or turn off detailed monitoring for your

instances
By default, your instance is enabled for basic monitoring. You can optionally enable detailed monitoring.
After you enable detailed monitoring, the Amazon EC2 console displays monitoring graphs with a 1-
minute period for the instance.

The following describes the data interval and charge for basic and detailed monitoring for instances.

Basic monitoring

Data is available automatically in 5-minute periods at no charge.

Detailed monitoring

Data is available in 1-minute periods for an additional charge.

761
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enable detailed monitoring

To get this level of data, you must specifically enable it for the instance. For the instances where
you've enabled detailed monitoring, you can also get aggregated data across groups of similar
instances.

Charges for detailed monitoring

If you enable detailed monitoring, you are charged per metric that is sent to CloudWatch. You are not
charged for data storage. For more information about pricing for detailed monitoring, see Paid tier on
the Amazon CloudWatch pricing page. For a pricing example, see Example 1 - EC2 Detailed Monitoring
on the Amazon CloudWatch pricing page.

Enable detailed monitoring

You can enable detailed monitoring on an instance as you launch it or after the instance is running
or stopped. Enabling detailed monitoring on an instance does not affect the monitoring of the EBS
volumes attached to the instance. For more information, see Amazon CloudWatch metrics for Amazon
EBS (p. 1252).

New console

To enable detailed monitoring for an existing instance (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, Monitoring, Manage detailed monitoring.
4. On the Detailed monitoring detail page, for Detailed monitoring, select the Enable check box.
5. Choose Save.

Old console

To enable detailed monitoring for an existing instance (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, CloudWatch Monitoring, Enable Detailed Monitoring.
4. In the Enable Detailed Monitoring dialog box, choose Yes, Enable.
5. Choose Close.

To enable detailed monitoring when launching an instance (console)

When launching an instance using the AWS Management Console, select the Monitoring check box on
the Configure Instance Details page.

To enable detailed monitoring for an existing instance (AWS CLI)

Use the following monitor-instances command to enable detailed monitoring for the specified instances.

aws ec2 monitor-instances --instance-ids i-1234567890abcdef0

To enable detailed monitoring when launching an instance (AWS CLI)

Use the run-instances command with the --monitoring flag to enable detailed monitoring.

aws ec2 run-instances --image-id ami-09092360 --monitoring Enabled=true...

762
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List available metrics

Turn off detailed monitoring

You can turn off detailed monitoring on an instance as you launch it or after the instance is running or
stopped.

New console

To turn off detailed monitoring (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, Monitoring, Manage detailed monitoring.
4. On the Detailed monitoring detail page, for Detailed monitoring, clear the Enable check box.
5. Choose Save.

Old console

To turn off detailed monitoring (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, CloudWatch Monitoring, Disable Detailed Monitoring.
4. In the Disable Detailed Monitoring dialog box, choose Yes, Disable.
5. Choose Close.

To turn off detailed monitoring (AWS CLI)

Use the following unmonitor-instances command to turn off detailed monitoring for the specified
instances.

aws ec2 unmonitor-instances --instance-ids i-1234567890abcdef0

List the available CloudWatch metrics for your

instances
Amazon EC2 sends metrics to Amazon CloudWatch. You can use the AWS Management Console, the
AWS CLI, or an API to list the metrics that Amazon EC2 sends to CloudWatch. By default, each data point
covers the 5 minutes that follow the start time of activity for the instance. If you've enabled detailed
monitoring, each data point covers the next minute of activity from the start time.

For information about getting the statistics for these metrics, see Get statistics for metrics for your
instances (p. 774).

Contents
• Instance metrics (p. 764)
• CPU credit metrics (p. 766)
• Amazon EBS metrics for Nitro-based instances (p. 767)
• Status check metrics (p. 769)
• Traffic mirroring metrics (p. 770)
• Amazon EC2 metric dimensions (p. 770)
• Amazon EC2 usage metrics (p. 770)

763
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List available metrics

• List metrics using the console (p. 771)

• List metrics using the AWS CLI (p. 773)

Instance metrics
The AWS/EC2 namespace includes the following instance metrics.

Metric Description

CPUUtilization The percentage of allocated EC2 compute units that are currently
in use on the instance. This metric identifies the processing power
required to run an application on a selected instance.

Depending on the instance type, tools in your operating system can

show a lower percentage than CloudWatch when the instance is not
allocated a full processor core.

Units: Percent

DiskReadOps Completed read operations from all instance store volumes

available to the instance in a specified period of time.

To calculate the average I/O operations per second (IOPS) for the
period, divide the total operations in the period by the number of
seconds in that period.

If there are no instance store volumes, either the value is 0 or the

metric is not reported.

Units: Count

DiskWriteOps Completed write operations to all instance store volumes available

to the instance in a specified period of time.

To calculate the average I/O operations per second (IOPS) for the
period, divide the total operations in the period by the number of
seconds in that period.

If there are no instance store volumes, either the value is 0 or the

metric is not reported.

Units: Count

DiskReadBytes Bytes read from all instance store volumes available to the instance.

This metric is used to determine the volume of the data the

application reads from the hard disk of the instance. This can be
used to determine the speed of the application.

The number reported is the number of bytes received during

the period. If you are using basic (five-minute) monitoring, you
can divide this number by 300 to find Bytes/second. If you have
detailed (one-minute) monitoring, divide it by 60.

If there are no instance store volumes, either the value is 0 or the

metric is not reported.

Units: Bytes

764
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List available metrics

Metric Description

DiskWriteBytes Bytes written to all instance store volumes available to the instance.

This metric is used to determine the volume of the data the

application writes onto the hard disk of the instance. This can be
used to determine the speed of the application.

The number reported is the number of bytes received during

the period. If you are using basic (five-minute) monitoring, you
can divide this number by 300 to find Bytes/second. If you have
detailed (one-minute) monitoring, divide it by 60.

If there are no instance store volumes, either the value is 0 or the

metric is not reported.

Units: Bytes

NetworkIn The number of bytes received on all network interfaces by the

instance. This metric identifies the volume of incoming network
traffic to a single instance.

The number reported is the number of bytes received during

the period. If you are using basic (five-minute) monitoring, you
can divide this number by 300 to find Bytes/second. If you have
detailed (one-minute) monitoring, divide it by 60.

Units: Bytes

NetworkOut The number of bytes sent out on all network interfaces by the
instance. This metric identifies the volume of outgoing network
traffic from a single instance.

The number reported is the number of bytes sent during the period.
If you are using basic (five-minute) monitoring, you can divide this
number by 300 to find Bytes/second. If you have detailed (one-
minute) monitoring, divide it by 60.

Units: Bytes

NetworkPacketsIn The number of packets received on all network interfaces by the

instance. This metric identifies the volume of incoming traffic in
terms of the number of packets on a single instance. This metric is
available for basic monitoring only.

Units: Count

Statistics: Minimum, Maximum, Average

NetworkPacketsOut The number of packets sent out on all network interfaces by the
instance. This metric identifies the volume of outgoing traffic in
terms of the number of packets on a single instance. This metric is
available for basic monitoring only.

Units: Count

Statistics: Minimum, Maximum, Average

765
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List available metrics

Metric Description

MetadataNoToken The number of times the instance metadata service was successfully
accessed using a method that does not use a token.

This metric is used to determine if there are any processes accessing

instance metadata that are using Instance Metadata Service Version
1, which does not use a token. If all requests use token-backed
sessions, i.e., Instance Metadata Service Version 2, the value is 0. For
more information, see Transition to using Instance Metadata Service
Version 2 (p. 706).

Units: Count

CPU credit metrics

The AWS/EC2 namespace includes the following CPU credit metrics for your burstable performance
instances (p. 224).

Metric Description

CPUCreditUsage The number of CPU credits spent by the instance for CPU
utilization. One CPU credit equals one vCPU running at 100%
utilization for one minute or an equivalent combination of vCPUs,
utilization, and time (for example, one vCPU running at 50%
utilization for two minutes or two vCPUs running at 25% utilization
for two minutes).

CPU credit metrics are available at a five-minute frequency only. If

you specify a period greater than five minutes, use the Sum statistic
instead of the Average statistic.

Units: Credits (vCPU-minutes)

CPUCreditBalance The number of earned CPU credits that an instance has

accrued since it was launched or started. For T2 Standard, the
CPUCreditBalance also includes the number of launch credits
that have been accrued.

Credits are accrued in the credit balance after they are earned,
and removed from the credit balance when they are spent. The
credit balance has a maximum limit, determined by the instance
size. After the limit is reached, any new credits that are earned are
discarded. For T2 Standard, launch credits do not count towards the
limit.

The credits in the CPUCreditBalance are available for the

instance to spend to burst beyond its baseline CPU utilization.

When an instance is running, credits in the CPUCreditBalance

do not expire. When a T3 or T3a instance stops, the
CPUCreditBalance value persists for seven days. Thereafter,
all accrued credits are lost. When a T2 instance stops, the
CPUCreditBalance value does not persist, and all accrued credits
are lost.

CPU credit metrics are available at a five-minute frequency only.

766
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List available metrics

Metric Description
Units: Credits (vCPU-minutes)

CPUSurplusCreditBalance The number of surplus credits that have been spent by an

unlimited instance when its CPUCreditBalance value is zero.

The CPUSurplusCreditBalance value is paid down by earned

CPU credits. If the number of surplus credits exceeds the maximum
number of credits that the instance can earn in a 24-hour period,
the spent surplus credits above the maximum incur an additional
charge.

CPU credit metrics are available at a five-minute frequency only.

Units: Credits (vCPU-minutes)

CPUSurplusCreditsCharged The number of spent surplus credits that are not paid down by
earned CPU credits, and which thus incur an additional charge.

Spent surplus credits are charged when any of the following occurs:

• The spent surplus credits exceed the maximum number of credits

that the instance can earn in a 24-hour period. Spent surplus
credits above the maximum are charged at the end of the hour.
• The instance is stopped or terminated.
• The instance is switched from unlimited to standard.

CPU credit metrics are available at a five-minute frequency only.

Units: Credits (vCPU-minutes)

Amazon EBS metrics for Nitro-based instances

The AWS/EC2 namespace includes the following Amazon EBS metrics for the Nitro-based instances that
are not bare metal instances. For the list of Nitro-based instance types, see Instances built on the Nitro
System (p. 209).

Metric values for Nitro-based instances will always be integers (whole numbers), whereas values for Xen-
based instances support decimals. Therefore, low instance CPU utilization on Nitro-based instances may
appear to be rounded down to 0.

Metric Description

EBSReadOps Completed read operations from all Amazon EBS

volumes attached to the instance in a specified period
of time.

To calculate the average read I/O operations per second

(Read IOPS) for the period, divide the total operations
in the period by the number of seconds in that period.
If you are using basic (five-minute) monitoring, you can
divide this number by 300 to calculate the Read IOPS. If
you have detailed (one-minute) monitoring, divide it by
60.

767
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List available metrics

Metric Description
Unit: Count

EBSWriteOps Completed write operations to all EBS volumes

attached to the instance in a specified period of time.

To calculate the average write I/O operations per

second (Write IOPS) for the period, divide the total
operations in the period by the number of seconds
in that period. If you are using basic (five-minute)
monitoring, you can divide this number by 300 to
calculate the Write IOPS. If you have detailed (one-
minute) monitoring, divide it by 60.

Unit: Count

EBSReadBytes Bytes read from all EBS volumes attached to the

instance in a specified period of time.

The number reported is the number of bytes read

during the period. If you are using basic (five-minute)
monitoring, you can divide this number by 300 to find
Read Bytes/second. If you have detailed (one-minute)
monitoring, divide it by 60.

Unit: Bytes

EBSWriteBytes Bytes written to all EBS volumes attached to the

instance in a specified period of time.

The number reported is the number of bytes written

during the period. If you are using basic (five-minute)
monitoring, you can divide this number by 300 to find
Write Bytes/second. If you have detailed (one-minute)
monitoring, divide it by 60.

Unit: Bytes

EBSIOBalance% Provides information about the percentage of I/O

credits remaining in the burst bucket. This metric is
available for basic monitoring only.

Instance sizes that support this metric can be found in

the table under EBS optimized by default (p. 1218): the
instances in the Instance size column that include an
asterisk (*) support this metric.

The Sum statistic is not applicable to this metric.

Unit: Percent

768
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List available metrics

Metric Description

EBSByteBalance% Provides information about the percentage of

throughput credits remaining in the burst bucket. This
metric is available for basic monitoring only.

Instance sizes that support this metric can be found in

the table under EBS optimized by default (p. 1218): the
instances in the Instance size column that include an
asterisk (*) support this metric.

The Sum statistic is not applicable to this metric.

Unit: Percent

For information about the metrics provided for your EBS volumes, see Amazon EBS metrics (p. 1252).
For information about the metrics provided for your Spot fleets, see CloudWatch metrics for Spot
Fleet (p. 428).

Status check metrics

The AWS/EC2 namespace includes the following status check metrics. By default, status check metrics
are available at a 1-minute frequency at no charge. For a newly-launched instance, status check metric
data is only available after the instance has completed the initialization state (within a few minutes
of the instance entering the running state). For more information about EC2 status checks, see Status
checks for your instances (p. 743).

Metric Description

StatusCheckFailed Reports whether the instance has passed both the instance status
check and the system status check in the last minute.

This metric can be either 0 (passed) or 1 (failed).

By default, this metric is available at a 1-minute frequency at no

charge.

Units: Count

StatusCheckFailed_Instance Reports whether the instance has passed the instance status check
in the last minute.

This metric can be either 0 (passed) or 1 (failed).

By default, this metric is available at a 1-minute frequency at no

charge.

Units: Count

StatusCheckFailed_System Reports whether the instance has passed the system status check in
the last minute.

This metric can be either 0 (passed) or 1 (failed).

By default, this metric is available at a 1-minute frequency at no

charge.

Units: Count

769
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List available metrics

Traffic mirroring metrics

The AWS/EC2 namespace includes metrics for mirrored traffic. For more information, see Monitoring
mirrored traffic using Amazon CloudWatch in the Amazon VPC Traffic Mirroring Guide.

Amazon EC2 metric dimensions

You can use the following dimensions to refine the metrics listed in the previous tables.

Dimension Description

AutoScalingGroupName This dimension filters the data you request for all instances in a
specified capacity group. An Auto Scaling group is a collection of
instances you define if you're using Auto Scaling. This dimension is
available only for Amazon EC2 metrics when the instances are in
such an Auto Scaling group. Available for instances with Detailed or
Basic Monitoring enabled.

ImageId This dimension filters the data you request for all instances running
this Amazon EC2 Amazon Machine Image (AMI). Available for
instances with Detailed Monitoring enabled.

InstanceId This dimension filters the data you request for the identified
instance only. This helps you pinpoint an exact instance from which
to monitor data.

InstanceType This dimension filters the data you request for all instances
running with this specified instance type. This helps you categorize
your data by the type of instance running. For example, you
might compare data from an m1.small instance and an m1.large
instance to determine which has the better business value for
your application. Available for instances with Detailed Monitoring
enabled.

Amazon EC2 usage metrics

You can use CloudWatch usage metrics to provide visibility into your account's usage of resources. Use
these metrics to visualize your current service usage on CloudWatch graphs and dashboards.

Amazon EC2 usage metrics correspond to AWS service quotas. You can configure alarms that alert you
when your usage approaches a service quota. For more information about CloudWatch integration with
service quotas, see Service Quotas Integration and Usage Metrics.

Amazon EC2 publishes the following metrics in the AWS/Usage namespace.

Metric Description

ResourceCount The number of the specified resources running in your account. The
resources are defined by the dimensions associated with the metric.

The most useful statistic for this metric is MAXIMUM, which

represents the maximum number of resources used during the 1-
minute period.

770
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List available metrics

The following dimensions are used to refine the usage metrics that are published by Amazon EC2.

Dimension Description

Service The name of the AWS service containing the resource. For Amazon
EC2 usage metrics, the value for this dimension is EC2.

Type The type of entity that is being reported. Currently, the only valid
value for Amazon EC2 usage metrics is Resource.

Resource The type of resource that is running. Currently, the only valid value
for Amazon EC2 usage metrics is vCPU, which returns information
on instances that are running.

Class The class of resource being tracked. For Amazon EC2 usage metrics
with vCPU as the value of the Resource dimension, the valid
values are Standard/OnDemand, F/OnDemand, G/OnDemand,
Inf/OnDemand, P/OnDemand, and X/OnDemand.

The values for this dimension define the first letter of the instance
types that are reported by the metric. For example, Standard/
OnDemand returns information about all running instances with
types that start with A, C, D, H, I, M, R, T, and Z, and G/OnDemand
returns information about all running instances with types that
start with G.

List metrics using the console

Metrics are grouped first by namespace, and then by the various dimension combinations within each
namespace. For example, you can view all metrics provided by Amazon EC2, or metrics grouped by
instance ID, instance type, image (AMI) ID, or Auto Scaling group.

To view available metrics by category (console)

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. In the navigation pane, choose Metrics.
3. Choose the EC2 metric namespace.

771
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List available metrics

4. Select a metric dimension (for example, Per-Instance Metrics).

772
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List available metrics

5. To sort the metrics, use the column heading. To graph a metric, select the check box next to the
metric. To filter by resource, choose the resource ID and then choose Add to search. To filter by
metric, choose the metric name and then choose Add to search.

List metrics using the AWS CLI

Use the list-metrics command to list the CloudWatch metrics for your instances.

To list all the available metrics for Amazon EC2 (AWS CLI)

The following example specifies the AWS/EC2 namespace to view all the metrics for Amazon EC2.

aws cloudwatch list-metrics --namespace AWS/EC2

The following is example output:

{
"Metrics": [
{
"Namespace": "AWS/EC2",
"Dimensions": [
{
"Name": "InstanceId",
"Value": "i-1234567890abcdef0"
}
],
"MetricName": "NetworkOut"
},
{
"Namespace": "AWS/EC2",
"Dimensions": [
{
"Name": "InstanceId",

773
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get statistics for metrics

"Value": "i-1234567890abcdef0"
}
],
"MetricName": "CPUUtilization"
},
{
"Namespace": "AWS/EC2",
"Dimensions": [
{
"Name": "InstanceId",
"Value": "i-1234567890abcdef0"
}
],
"MetricName": "NetworkIn"
},
...
]
}

To list all the available metrics for an instance (AWS CLI)

The following example specifies the AWS/EC2 namespace and the InstanceId dimension to view the
results for the specified instance only.

aws cloudwatch list-metrics --namespace AWS/EC2 --dimensions

Name=InstanceId,Value=i-1234567890abcdef0

To list a metric across all instances (AWS CLI)

The following example specifies the AWS/EC2 namespace and a metric name to view the results for the
specified metric only.

aws cloudwatch list-metrics --namespace AWS/EC2 --metric-name CPUUtilization

Get statistics for metrics for your instances

You can get statistics for the CloudWatch metrics for your instances.

Contents
• Statistics overview (p. 774)
• Get statistics for a specific instance (p. 775)
• Aggregate statistics across instances (p. 778)
• Aggregate statistics by Auto Scaling group (p. 780)
• Aggregate statistics by AMI (p. 781)

Statistics overview
Statistics are metric data aggregations over specified periods of time. CloudWatch provides statistics
based on the metric data points provided by your custom data or provided by other services in AWS to
CloudWatch. Aggregations are made using the namespace, metric name, dimensions, and the data point
unit of measure, within the time period you specify. The following table describes the available statistics.

Statistic Description

Minimum The lowest value observed during the specified period. You can use this value to
determine low volumes of activity for your application.

774
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get statistics for metrics

Statistic Description

Maximum The highest value observed during the specified period. You can use this value to
determine high volumes of activity for your application.

Sum All values submitted for the matching metric added together. This statistic can be
useful for determining the total volume of a metric.

Average The value of Sum / SampleCount during the specified period. By comparing this
statistic with the Minimum and Maximum, you can determine the full scope of a metric
and how close the average use is to the Minimum and Maximum. This comparison
helps you to know when to increase or decrease your resources as needed.

SampleCount The count (number) of data points used for the statistical calculation.

pNN.NN The value of the specified percentile. You can specify any percentile, using up to two
decimal places (for example, p95.45).

Get statistics for a specific instance

The following examples show you how to use the AWS Management Console or the AWS CLI to
determine the maximum CPU utilization of a specific EC2 instance.

Requirements

• You must have the ID of the instance. You can get the instance ID using the AWS Management Console
or the describe-instances command.
• By default, basic monitoring is enabled, but you can enable detailed monitoring. For more information,
see Enable or turn off detailed monitoring for your instances (p. 761).

To display the CPU utilization for a specific instance (console)

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. In the navigation pane, choose Metrics.
3. Choose the EC2 metric namespace.

775
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get statistics for metrics

4. Choose the Per-Instance Metrics dimension.

776
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get statistics for metrics

5. In the search field, enter CPUUtilization and press Enter. Choose the row for the specific
instance, which displays a graph for the CPUUtilization metric for the instance. To name the graph,
choose the pencil icon. To change the time range, select one of the predefined values or choose
custom.

6. To change the statistic or the period for the metric, choose the Graphed metrics tab. Choose the
column heading or an individual value, and then choose a different value.

To get the CPU utilization for a specific instance (AWS CLI)

Use the following get-metric-statistics command to get the CPUUtilization metric for the specified
instance, using the specified period and time interval:

777
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get statistics for metrics

aws cloudwatch get-metric-statistics --namespace AWS/EC2 --metric-name CPUUtilization --

period 3600 \
--statistics Maximum --dimensions Name=InstanceId,Value=i-1234567890abcdef0 \
--start-time 2016-10-18T23:18:00 --end-time 2016-10-19T23:18:00

The following is example output. Each value represents the maximum CPU utilization percentage for a
single EC2 instance.

{
"Datapoints": [
{
"Timestamp": "2016-10-19T00:18:00Z",
"Maximum": 0.33000000000000002,
"Unit": "Percent"
},
{
"Timestamp": "2016-10-19T03:18:00Z",
"Maximum": 99.670000000000002,
"Unit": "Percent"
},
{
"Timestamp": "2016-10-19T07:18:00Z",
"Maximum": 0.34000000000000002,
"Unit": "Percent"
},
{
"Timestamp": "2016-10-19T12:18:00Z",
"Maximum": 0.34000000000000002,
"Unit": "Percent"
},
...
],
"Label": "CPUUtilization"
}

Aggregate statistics across instances

Aggregate statistics are available for the instances that have detailed monitoring enabled. Instances
that use basic monitoring are not included in the aggregates. In addition, Amazon CloudWatch does not
aggregate data across regions. Therefore, metrics are completely separate between regions. Before you
can get statistics aggregated across instances, you must enable detailed monitoring (at an additional
charge), which provides data in 1-minute periods.

This example shows you how to use detailed monitoring to get the average CPU usage for your EC2
instances. Because no dimension is specified, CloudWatch returns statistics for all dimensions in the AWS/
EC2 namespace.
Important
This technique for retrieving all dimensions across an AWS namespace does not work for custom
namespaces that you publish to Amazon CloudWatch. With custom namespaces, you must
specify the complete set of dimensions that are associated with any given data point to retrieve
statistics that include the data point.

To display average CPU utilization across your instances (console)

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. In the navigation pane, choose Metrics.
3. Choose the EC2 namespace and then choose Across All Instances.

778
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get statistics for metrics

4. Choose the row that contains CPUUtilization, which displays a graph for the metric for all your EC2
instances. To name the graph, choose the pencil icon. To change the time range, select one of the
predefined values or choose custom.

5. To change the statistic or the period for the metric, choose the Graphed metrics tab. Choose the
column heading or an individual value, and then choose a different value.

To get average CPU utilization across your instances (AWS CLI)

Use the get-metric-statistics command as follows to get the average of the CPUUtilization metric across
your instances.

aws cloudwatch get-metric-statistics \

--namespace AWS/EC2 \
--metric-name CPUUtilization \
--period 3600 --statistics "Average" "SampleCount" \
--start-time 2016-10-11T23:18:00 \
--end-time 2016-10-12T23:18:00

The following is example output:

{
"Datapoints": [
{
"SampleCount": 238.0,
"Timestamp": "2016-10-12T07:18:00Z",
"Average": 0.038235294117647062,
"Unit": "Percent"
},
{

779
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get statistics for metrics

"SampleCount": 240.0,
"Timestamp": "2016-10-12T09:18:00Z",
"Average": 0.16670833333333332,
"Unit": "Percent"
},
{
"SampleCount": 238.0,
"Timestamp": "2016-10-11T23:18:00Z",
"Average": 0.041596638655462197,
"Unit": "Percent"
},
...
],
"Label": "CPUUtilization"
}

Aggregate statistics by Auto Scaling group

You can aggregate statistics for the EC2 instances in an Auto Scaling group. Note that Amazon
CloudWatch cannot aggregate data across regions. Metrics are completely separate between regions.

This example shows you how to retrieve the total bytes written to disk for one Auto Scaling group. The
total is computed for one-minute periods for a 24-hour interval across all EC2 instances in the specified
Auto Scaling group.

To display DiskWriteBytes for the instances in an Auto Scaling group (console)

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. In the navigation pane, choose Metrics.
3. Choose the EC2 namespace and then choose By Auto Scaling Group.
4. Choose the row for the DiskWriteBytes metric and the specific Auto Scaling group, which displays
a graph for the metric for the instances in the Auto Scaling group. To name the graph, choose the
pencil icon. To change the time range, select one of the predefined values or choose custom.
5. To change the statistic or the period for the metric, choose the Graphed metrics tab. Choose the
column heading or an individual value, and then choose a different value.

To display DiskWriteBytes for the instances in an Auto Scaling group (AWS CLI)

Use the get-metric-statistics command as follows.

aws cloudwatch get-metric-statistics --namespace AWS/EC2 --metric-name DiskWriteBytes --

period 360 \
--statistics "Sum" "SampleCount" --dimensions Name=AutoScalingGroupName,Value=my-asg --
start-time 2016-10-16T23:18:00 --end-time 2016-10-18T23:18:00

The following is example output:

{
"Datapoints": [
{
"SampleCount": 18.0,
"Timestamp": "2016-10-19T21:36:00Z",
"Sum": 0.0,
"Unit": "Bytes"
},
{
"SampleCount": 5.0,

780
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get statistics for metrics

"Timestamp": "2016-10-19T21:42:00Z",
"Sum": 0.0,
"Unit": "Bytes"
}
],
"Label": "DiskWriteBytes"
}

Aggregate statistics by AMI

You can aggregate statistics for your instances that have detailed monitoring enabled. Instances that use
basic monitoring are not included. Note that Amazon CloudWatch cannot aggregate data across regions.
Metrics are completely separate between regions.

Before you can get statistics aggregated across instances, you must enable detailed monitoring (at an
additional charge), which provides data in 1-minute periods. For more information, see Enable or turn off
detailed monitoring for your instances (p. 761).

This example shows you how to determine average CPU utilization for all instances that use a specific
Amazon Machine Image (AMI). The average is over 60-second time intervals for a one-day period.

To display the average CPU utilization by AMI (console)

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. In the navigation pane, choose Metrics.
3. Choose the EC2 namespace and then choose By Image (AMI) Id.
4. Choose the row for the CPUUtilization metric and the specific AMI, which displays a graph for the
metric for the specified AMI. To name the graph, choose the pencil icon. To change the time range,
select one of the predefined values or choose custom.
5. To change the statistic or the period for the metric, choose the Graphed metrics tab. Choose the
column heading or an individual value, and then choose a different value.

To get the average CPU utilization for an image ID (AWS CLI)

Use the get-metric-statistics command as follows.

aws cloudwatch get-metric-statistics --namespace AWS/EC2 --metric-name CPUUtilization --

period 3600 \
--statistics Average --dimensions Name=ImageId,Value=ami-3c47a355 --start-
time 2016-10-10T00:00:00 --end-time 2016-10-11T00:00:00

The following is example output. Each value represents an average CPU utilization percentage for the
EC2 instances running the specified AMI.

{
"Datapoints": [
{
"Timestamp": "2016-10-10T07:00:00Z",
"Average": 0.041000000000000009,
"Unit": "Percent"
},
{
"Timestamp": "2016-10-10T14:00:00Z",
"Average": 0.079579831932773085,
"Unit": "Percent"
},

781
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Graph metrics

{
"Timestamp": "2016-10-10T06:00:00Z",
"Average": 0.036000000000000011,
"Unit": "Percent"
},
...
],
"Label": "CPUUtilization"
}

Graph metrics for your instances

After you launch an instance, you can open the Amazon EC2 console and view the monitoring graphs for
an instance on the Monitoring tab. Each graph is based on one of the available Amazon EC2 metrics.

The following graphs are available:

• Average CPU Utilization (Percent)

• Average Disk Reads (Bytes)
• Average Disk Writes (Bytes)
• Maximum Network In (Bytes)
• Maximum Network Out (Bytes)
• Summary Disk Read Operations (Count)
• Summary Disk Write Operations (Count)
• Summary Status (Any)
• Summary Status Instance (Count)
• Summary Status System (Count)

For more information about the metrics and the data they provide to the graphs, see List the available
CloudWatch metrics for your instances (p. 763).

Graph Metrics Using the CloudWatch Console

You can also use the CloudWatch console to graph metric data generated by Amazon EC2 and other AWS
services. For more information, see Graph Metrics in the Amazon CloudWatch User Guide.

Create a CloudWatch alarm for an instance

You can create a CloudWatch alarm that monitors CloudWatch metrics for one of your instances.
CloudWatch will automatically send you a notification when the metric reaches a threshold you specify.
You can create a CloudWatch alarm using the Amazon EC2 console, or using the more advanced options
provided by the CloudWatch console.

To create an alarm using the CloudWatch console

For examples, see Creating Amazon CloudWatch Alarms in the Amazon CloudWatch User Guide.

New console

To create an alarm using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

782
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create an alarm

2. In the navigation pane, choose Instances.

3. Select the instance and choose Actions, Monitor and troubleshoot, Manage CloudWatch
alarms.
4. On the Manage CloudWatch alarms detail page, under Add or edit alarm, select Create an
alarm.
5. For Alarm notification, choose whether to turn the toggle on or off to configure Amazon
Simple Notification Service (Amazon SNS) notifications. Enter an existing Amazon SNS topic or
enter a name to create a new topic.
6. For Alarm action, choose whether to turn the toggle on or off to specify an action to take when
the alarm is triggered. Select an action from the dropdown.
7. For Alarm thresholds, select the metric and criteria for the alarm. For example, you can
leave the default settings for Group samples by (Average) and Type of data to sample (CPU
utilization). For Alarm when, choose >= and enter 0.80. For Consecutive period, enter 1. For
Period, select 5 minutes.
8. (Optional) For Sample metric data, choose Add to dashboard.
9. Choose Create.

Old console

To create an alarm using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance.
4. On the Monitoring tab located at the bottom of the page, choose Create Alarm. Or, from the
Actions dropdown, choose CloudWatch Monitoring, Add/Edit Alarm.
5. In the Create Alarm dialog box, do the following:

a. Choose create topic. For Send a notification to, enter a name for the SNS topic. For With
these recipients, enter one or more email addresses to receive notification.
b. Specify the metric and the criteria for the policy. For example, you can leave the default
settings for Whenever (Average of CPU Utilization). For Is, choose >= and enter 80 percent.
For For at least, enter 1 consecutive period of 5 Minutes.
c. Choose Create Alarm.

783
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance

You can edit your CloudWatch alarm settings from the Amazon EC2 console or the CloudWatch console.
If you want to delete your alarm, you can do so from the CloudWatch console. For more information, see
Editing or Deleting a CloudWatch Alarm in the Amazon CloudWatch User Guide.

Create alarms that stop, terminate, reboot, or recover

an instance
Using Amazon CloudWatch alarm actions, you can create alarms that automatically stop, terminate,
reboot, or recover your instances. You can use the stop or terminate actions to help you save money
when you no longer need an instance to be running. You can use the reboot and recover actions to
automatically reboot those instances or recover them onto new hardware if a system impairment occurs.

The AWSServiceRoleForCloudWatchEvents service-linked role enables AWS to perform alarm

actions on your behalf. The first time you create an alarm in the AWS Management Console, the IAM CLI,
or the IAM API, CloudWatch creates the service-linked role for you.

There are a number of scenarios in which you might want to automatically stop or terminate your
instance. For example, you might have instances dedicated to batch payroll processing jobs or scientific
computing tasks that run for a period of time and then complete their work. Rather than letting those
instances sit idle (and accrue charges), you can stop or terminate them, which can help you to save
money. The main difference between using the stop and the terminate alarm actions is that you can
easily start a stopped instance if you need to run it again later, and you can keep the same instance

784
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
ID and root volume. However, you cannot start a terminated instance. Instead, you must launch a new
instance.

You can add the stop, terminate, reboot, or recover actions to any alarm that is set on an Amazon EC2
per-instance metric, including basic and detailed monitoring metrics provided by Amazon CloudWatch
(in the AWS/EC2 namespace), as well as any custom metrics that include the InstanceId dimension, as
long as its value refers to a valid running Amazon EC2 instance.

Console support

You can create alarms using the Amazon EC2 console or the CloudWatch console. The procedures in
this documentation use the Amazon EC2 console. For procedures that use the CloudWatch console, see
Create Alarms That Stop, Terminate, Reboot, or Recover an Instance in the Amazon CloudWatch User
Guide.

Permissions

If you are an AWS Identity and Access Management (IAM) user, you must have the
iam:CreateServiceLinkedRole to create or modify an alarm that performs EC2 alarm actions.

Contents
• Add stop actions to Amazon CloudWatch alarms (p. 785)
• Add terminate actions to Amazon CloudWatch alarms (p. 787)
• Add reboot actions to Amazon CloudWatch alarms (p. 788)
• Add recover actions to Amazon CloudWatch alarms (p. 789)
• Use the Amazon CloudWatch console to view alarm and action history (p. 792)
• Amazon CloudWatch alarm action scenarios (p. 792)

Add stop actions to Amazon CloudWatch alarms

You can create an alarm that stops an Amazon EC2 instance when a certain threshold has been met. For
example, you may run development or test instances and occasionally forget to shut them off. You can
create an alarm that is triggered when the average CPU utilization percentage has been lower than 10
percent for 24 hours, signaling that it is idle and no longer in use. You can adjust the threshold, duration,
and period to suit your needs, plus you can add an Amazon Simple Notification Service (Amazon SNS)
notification so that you receive an email when the alarm is triggered.

Instances that use an Amazon EBS volume as the root device can be stopped or terminated, whereas
instances that use the instance store as the root device can only be terminated.

New console

To create an alarm to stop an idle instance (Amazon EC2 console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, Monitor and troubleshoot, Manage CloudWatch
alarms.

Alternatively, you can choose the plus sign ( ) in the Alarm status column.
4. On the Manage CloudWatch alarms page, do the following:

a. Choose Create a new alarm.

785
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
b. To receive an email when the alarm is triggered, for Alarm notification, choose an existing
Amazon SNS topic. You first need to create an Amazon SNS topic using the Amazon SNS
console. For more information, see Using Amazon SNS for application-to-person (A2P)
messaging in the Amazon Simple Notification Service Developer Guide.
c. Toggle on Alarm action, and choose Stop.
d. For Group samples by and Type of data to sample, choose a statistic and a metric. In this
example, choose Average and CPU Utilization.
e. For Alarm When and Percent, specify the metric threshold. In this example, specify >= and
10 percent.
f. For Consecutive Period and Period, specify the evaluation period for the alarm. In this
example, specify 1 consecutive period of 5 Minutes.
g. Amazon CloudWatch automatically creates an alarm name for you. To change the name, for
Alarm name, enter a new name. Alarm names must contain only ASCII characters.
Note
You can adjust the alarm configuration based on your own requirements before
creating the alarm, or you can edit them later. This includes the metric, threshold,
duration, action, and notification settings. However, after you create an alarm, you
cannot edit its name later.
h. Choose Create.

Old console

To create an alarm to stop an idle instance (Amazon EC2 console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance. On the Monitoring tab, choose Create Alarm.
4. In the Create Alarm dialog box, do the following:

a. To receive an email when the alarm is triggered, for Send a notification to, choose an
existing Amazon SNS topic, or choose create topic to create a new one.

To create a new topic, for Send a notification to, enter a name for the topic, and then for
With these recipients, enter the email addresses of the recipients (separated by commas).
After you create the alarm, you will receive a subscription confirmation email that you must
accept before you can get notifications for this topic.
b. Choose Take the action, Stop this instance.
c. For Whenever, choose the statistic you want to use and then choose the metric. In this
example, choose Average and CPU Utilization.
d. For Is, specify the metric threshold. In this example, enter 10 percent.
e. For For at least, specify the evaluation period for the alarm. In this example, enter 24
consecutive period(s) of 1 Hour.
f. To change the name of the alarm, for Name of alarm, enter a new name. Alarm names
must contain only ASCII characters.

If you don't enter a name for the alarm, Amazon CloudWatch automatically creates one for
you.
Note
You can adjust the alarm configuration based on your own requirements before
creating the alarm, or you can edit them later. This includes the metric, threshold,
duration, action, and notification settings. However, after you create an alarm, you
cannot edit its name later.
786
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
g. Choose Create Alarm.

Add terminate actions to Amazon CloudWatch alarms

You can create an alarm that terminates an EC2 instance automatically when a certain threshold has
been met (as long as termination protection is not enabled for the instance). For example, you might
want to terminate an instance when it has completed its work, and you don’t need the instance again.
If you might want to use the instance later, you should stop the instance instead of terminating it. For
information on enabling and disabling termination protection for an instance, see Enable termination
protection (p. 645).

New console

To create an alarm to terminate an idle instance (Amazon EC2 console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, Monitor and troubleshoot, Manage CloudWatch
alarms.

Alternatively, you can choose the plus sign ( ) in the Alarm status column.
4. On the Manage CloudWatch alarms page, do the following:

a. Choose Create a new alarm.

b. To receive an email when the alarm is triggered, for Alarm notification, choose an existing
Amazon SNS topic. You first need to create an Amazon SNS topic using the Amazon SNS
console. For more information, see Using Amazon SNS for application-to-person (A2P)
messaging in the Amazon Simple Notification Service Developer Guide.
c. Toggle on Alarm action, and choose Terminate.
d. For Group samples by and Type of data to sample, choose a statistic and a metric. In this
example, choose Average and CPU Utilization.
e. For Alarm When and Percent, specify the metric threshold. In this example, specify => and
10 percent.
f. For Consecutive Period and Period, specify the evaluation period for the alarm. In this
example, specify 24 consecutive period(s) of 1 Hour.
g. Amazon CloudWatch automatically creates an alarm name for you. To change the name, for
Alarm name, enter a new name. Alarm names must contain only ASCII characters.
Note
You can adjust the alarm configuration based on your own requirements before
creating the alarm, or you can edit them later. This includes the metric, threshold,
duration, action, and notification settings. However, after you create an alarm, you
cannot edit its name later.
h. Choose Create.

Old console

To create an alarm to terminate an idle instance (Amazon EC2 console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance. On the Monitoring tab, choose Create Alarm.
4. In the Create Alarm dialog box, do the following:

787
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
a. To receive an email when the alarm is triggered, for Send a notification to, choose an
existing Amazon SNS topic, or choose create topic to create a new one.

To create a new topic, for Send a notification to, enter a name for the topic, and then for
With these recipients, enter the email addresses of the recipients (separated by commas).
After you create the alarm, you will receive a subscription confirmation email that you must
accept before you can get notifications for this topic.
b. Choose Take the action, Terminate this instance.
c. For Whenever, choose a statistic and then choose the metric. In this example, choose
Average and CPU Utilization.
d. For Is, specify the metric threshold. In this example, enter 10 percent.
e. For For at least, specify the evaluation period for the alarm. In this example, enter 24
consecutive period(s) of 1 Hour.
f. To change the name of the alarm, for Name of alarm, enter a new name. Alarm names
must contain only ASCII characters.

If you don't enter a name for the alarm, Amazon CloudWatch automatically creates one for
you.
Note
You can adjust the alarm configuration based on your own requirements before
creating the alarm, or you can edit them later. This includes the metric, threshold,
duration, action, and notification settings. However, after you create an alarm, you
cannot edit its name later.
g. Choose Create Alarm.

Add reboot actions to Amazon CloudWatch alarms

You can create an Amazon CloudWatch alarm that monitors an Amazon EC2 instance and automatically
reboots the instance. The reboot alarm action is recommended for Instance Health Check failures (as
opposed to the recover alarm action, which is suited for System Health Check failures). An instance
reboot is equivalent to an operating system reboot. In most cases, it takes only a few minutes to reboot
your instance. When you reboot an instance, it remains on the same physical host, so your instance keeps
its public DNS name, private IP address, and any data on its instance store volumes.

Rebooting an instance doesn't start a new instance billing period (with a minimum one-minute charge),
unlike stopping and restarting your instance. For more information, see Reboot Your Instance in the
Amazon EC2 User Guide for Linux Instances.
Important
To avoid a race condition between the reboot and recover actions, avoid setting the same
number of evaluation periods for a reboot alarm and a recover alarm. We recommend that you
set reboot alarms to three evaluation periods of one minute each. For more information, see
Evaluating an Alarm in the Amazon CloudWatch User Guide.

New console

To create an alarm to reboot an instance (Amazon EC2 console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, Monitor and troubleshoot, Manage CloudWatch
alarms.

Alternatively, you can choose the plus sign ( ) in the Alarm status column.

788
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
4. On the Manage CloudWatch alarms page, do the following:

a. To receive an email when the alarm is triggered, for Alarm notification, choose an existing
Amazon SNS topic. You first need to create an Amazon SNS topic using the Amazon SNS
console. For more information, see Using Amazon SNS for application-to-person (A2P)
messaging in the Amazon Simple Notification Service Developer Guide.
b. Toggle on Alarm action, and choose Reboot.
c. For Group samples by and Type of data to sample, choose a statistic and a metric. In this
example, choose Average and Status Check Failed: Instance.
d. For Consecutive Period and Period, specify the evaluation period for the alarm. In this
example, enter 3 consecutive period(s) of 5 Minutes.
e. Amazon CloudWatch automatically creates an alarm name for you. To change the name, for
Alarm name, enter a new name. Alarm names must contain only ASCII characters.
f. Choose Create.

Old console

To create an alarm to reboot an instance (Amazon EC2 console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance. On the Monitoring tab, choose Create Alarm.
4. In the Create Alarm dialog box, do the following:

a. To receive an email when the alarm is triggered, for Send a notification to, choose an
existing Amazon SNS topic, or choose create topic to create a new one.

To create a new topic, for Send a notification to, enter a name for the topic, and for With
these recipients, enter the email addresses of the recipients (separated by commas). After
you create the alarm, you will receive a subscription confirmation email that you must
accept before you can get notifications for this topic.
b. Select Take the action, Reboot this instance.
c. For Whenever, choose Status Check Failed (Instance).
d. For For at least, specify the evaluation period for the alarm. In this example, enter 3
consecutive period(s) of 5 Minutes.
e. To change the name of the alarm, for Name of alarm, enter a new name. Alarm names
must contain only ASCII characters.

If you don't enter a name for the alarm, Amazon CloudWatch automatically creates one for
you.
f. Choose Create Alarm.

Add recover actions to Amazon CloudWatch alarms

You can create an Amazon CloudWatch alarm that monitors an Amazon EC2 instance. If the instance
becomes impaired due to an underlying hardware failure or a problem that requires AWS involvement
to repair, you can automatically recover the instance. Terminated instances cannot be recovered. A
recovered instance is identical to the original instance, including the instance ID, private IP addresses,
Elastic IP addresses, and all instance metadata.

CloudWatch prevents you from adding a recovery action to an alarm that is on an instance which does
not support recovery actions.
789
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
When the StatusCheckFailed_System alarm is triggered, and the recover action is initiated, you are
notified by the Amazon SNS topic that you chose when you created the alarm and associated the recover
action. During instance recovery, the instance is migrated during an instance reboot, and any data that
is in-memory is lost. When the process is complete, information is published to the SNS topic you've
configured for the alarm. Anyone who is subscribed to this SNS topic receives an email notification that
includes the status of the recovery attempt and any further instructions. You notice an instance reboot
on the recovered instance.

The recover action can be used only with StatusCheckFailed_System, not with
StatusCheckFailed_Instance.

The following problems can cause system status checks to fail:

• Loss of network connectivity

• Loss of system power
• Software issues on the physical host
• Hardware issues on the physical host that impact network reachability

The recover action is supported only on instances with the following characteristics:

• Use one of the following instance types: A1, C3, C4, C5, C5a, C5n, C6g, C6gn, Inf1, M3, M4, M5, M5a,
M5n, M5zn, M6g, P3, P4, R3, R4, R5, R5a, R5b, R5n, R6g, T2, T3, T3a, T4g, X1, or X1e
• Use default or dedicated instance tenancy
• Use EBS volumes only (do not configure instance store volumes). For more information, see 'Recover
this instance' is disabled.

If your instance has a public IP address, it retains the public IP address after recovery.
Important
To avoid a race condition between the reboot and recover actions, avoid setting the same
number of evaluation periods for a reboot alarm and a recover alarm. We recommend that you
set recover alarms to two evaluation periods of one minute each. For more information, see
Evaluating an Alarm in the Amazon CloudWatch User Guide.

New console

To create an alarm to recover an instance (Amazon EC2 console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance and choose Actions, Monitor and troubleshoot, Manage CloudWatch
alarms.

Alternatively, you can choose the plus sign ( ) in the Alarm status column.
4. On the Manage CloudWatch alarms page, do the following:

a. Choose Create a new alarm.

b. To receive an email when the alarm is triggered, for Alarm notification, choose an existing
Amazon SNS topic. You first need to create an Amazon SNS topic using the Amazon SNS
console. For more information, see Using Amazon SNS for application-to-person (A2P)
messaging in the Amazon Simple Notification Service Developer Guide.
Note

• Users must subscribe to the specified SNS topic to receive email notifications
when the alarm is triggered.

790
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
• The AWS account root user always receives email notifications when automatic
instance recovery actions occur, even if an SNS topic is not specified.
• The AWS account root user always receives email notifications when automatic
instance recovery actions occur, even if it is not subscribed to the specified SNS
topic.
c. Toggle on Alarm action, and choose Recover.
d. For Group samples by and Type of data to sample, choose a statistic and a metric. In this
example, choose Average and Status Check Failed: System.
e. For Consecutive Period and Period, specify the evaluation period for the alarm. In this
example, enter 2 consecutive period(s) of 5 Minutes.
f. Amazon CloudWatch automatically creates an alarm name for you. To change the name, for
Alarm name, enter a new name. Alarm names must contain only ASCII characters.
g. Choose Create.

Old console

To create an alarm to recover an instance (Amazon EC2 console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance. On the Monitoring tab, choose Create Alarm.
4. In the Create Alarm dialog box, do the following:

a. To receive an email when the alarm is triggered, for Send a notification to, choose an
existing Amazon SNS topic, or choose create topic to create a new one.

To create a new topic, for Send a notification to, enter a name for the topic, and for With
these recipients, enter the email addresses of the recipients (separated by commas). After
you create the alarm, you will receive a subscription confirmation email that you must
accept before you can get email for this topic.
Note

• Users must subscribe to the specified SNS topic to receive email notifications
when the alarm is triggered.
• The AWS account root user always receives email notifications when automatic
instance recovery actions occur, even if an SNS topic is not specified.
• The AWS account root user always receives email notifications when automatic
instance recovery actions occur, even if it is not subscribed to the specified SNS
topic.
b. Select Take the action, Recover this instance.
c. For Whenever, choose Status Check Failed (System).
d. For For at least, specify the evaluation period for the alarm. In this example, enter 2
consecutive period(s) of 5 Minutes.
e. To change the name of the alarm, for Name of alarm, enter a new name. Alarm names
must contain only ASCII characters.

If you don't enter a name for the alarm, Amazon CloudWatch automatically creates one for
you.
f. Choose Create Alarm.

791
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
Use the Amazon CloudWatch console to view alarm and action
history
You can view alarm and action history in the Amazon CloudWatch console. Amazon CloudWatch keeps
the last two weeks' worth of alarm and action history.

To view the history of triggered alarms and actions (CloudWatch console)

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. In the navigation pane, choose Alarms.
3. Select an alarm.
4. The Details tab shows the most recent state transition along with the time and metric values.
5. Choose the History tab to view the most recent history entries.

Amazon CloudWatch alarm action scenarios

You can use the Amazon EC2 console to create alarm actions that stop or terminate an Amazon EC2
instance when certain conditions are met. In the following screen capture of the console page where you
set the alarm actions, we've numbered the settings. We've also numbered the settings in the scenarios
that follow, to help you create the appropriate actions.

792
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
New console

793
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
Old console

Scenario 1: Stop idle development and test instances

Create an alarm that stops an instance used for software development or testing when it has been idle
for at least an hour.

Setting Value

1 Stop

2 Maximum

3 CPU Utilization

4 <=

5 10%

6 1

7 1 Hour

794
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
Scenario 2: Stop idle instances
Create an alarm that stops an instance and sends an email when the instance has been idle for 24 hours.

Setting Value

1 Stop and email

2 Average

3 CPU Utilization

4 <=

5 5%

6 24

7 1 Hour

Scenario 3: Send email about web servers with unusually high traffic
Create an alarm that sends email when an instance exceeds 10 GB of outbound network traffic per day.

Setting Value

1 Email

2 Sum

3 Network Out

4 >

5 10 GB

6 24

7 1 Hour

Scenario 4: Stop web servers with unusually high traffic

Create an alarm that stops an instance and send a text message (SMS) if outbound traffic exceeds 1 GB
per hour.

Setting Value

1 Stop and send SMS

2 Sum

3 Network Out

4 >

5 1 GB

795
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create alarms that stop, terminate,
reboot, or recover an instance
Setting Value

6 1

7 1 Hour

Scenario 5: Stop an instance experiencing a memory leak

Create an alarm that stops an instance when memory utilization reaches or exceeds 90%, so that
application logs can be retrieved for troubleshooting.
Note
The MemoryUtilization metric is a custom metric. In order to use the MemoryUtilization metric,
you must install the Perl scripts for Linux instances. For more information, see Monitoring
Memory and Disk Metrics for Amazon EC2 Linux Instances.

Setting Value

1 Stop

2 Maximum

3 MemoryUtilization

4 >=

5 90%

6 1

7 5 Minutes

Scenario 6: Stop an impaired instance

Create an alarm that stops an instance that fails three consecutive status checks (performed at 5-minute
intervals).

Setting Value

1 Stop

2 Average

3 Status Check Failed: System

4 -

5 -

6 1

7 15 Minutes

Scenario 7: Terminate instances when batch processing jobs are complete

Create an alarm that terminates an instance that runs batch jobs when it is no longer sending results
data.

796
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Automate Amazon EC2 with EventBridge

Setting Value

1 Terminate

2 Maximum

3 Network Out

4 <=

5 100,000 bytes

6 1

7 5 Minutes

Automate Amazon EC2 with EventBridge

Amazon EventBridge enables you to automate your AWS services and respond automatically to system
events such as application availability issues or resource changes. Events from AWS services are delivered
to EventBridge in near real time. You can write simple rules to indicate which events are of interest
to you, and the automated actions to take when an event matches a rule. The actions that can be
automatically triggered include the following:

• Invoking an AWS Lambda function

• Invoking Amazon EC2 Run Command
• Relaying the event to Amazon Kinesis Data Streams
• Activating an AWS Step Functions state machine
• Notifying an Amazon SNS topic or an Amazon SQS queue

Some examples of using EventBridge with Amazon EC2 include:

• Activating a Lambda function whenever a new Amazon EC2 instance starts.

• Notifying an Amazon SNS topic when an Amazon EBS volume is created or modified.
• Sending a command to one or more Amazon EC2 instances using Amazon EC2 Run Command
whenever a certain event in another AWS service occurs.

For more information, see the Amazon EventBridge User Guide.

Monitor memory and disk metrics for Amazon EC2

Linux instances
You can use Amazon CloudWatch to collect metrics and logs from the operating systems for your EC2
instances.
Important
The CloudWatch monitoring scripts are deprecated. We recommend that you use the
CloudWatch agent to collect metrics and logs.

797
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Collect metrics using the CloudWatch agent

If you are still migrating from the deprecated monitoring scripts to the agent, and require
information about the monitoring scripts, see Deprecated: Collect metrics using the CloudWatch
monitoring scripts (p. 798).

Collect metrics using the CloudWatch agent

You can use the CloudWatch agent to collect both system metrics and log files from Amazon EC2
instances and on-premises servers. The agent supports both Windows Server and Linux, and enables
you to select the metrics to be collected, including sub-resource metrics such as per-CPU core. We
recommend that you use the agent to collect metrics and logs instead of using the deprecated
monitoring scripts. For more information, see Collect Metrics from Amazon EC2 Instances and On-
Premises Servers with the CloudWatch Agent in the Amazon CloudWatch User Guide.

Deprecated: Collect metrics using the CloudWatch

monitoring scripts
Important
The CloudWatch monitoring scripts are deprecated. We provide information about the
monitoring scripts for customers who have not yet migrated from the deprecated monitoring
scripts to the CloudWatch agent.
We recommend that you use the CloudWatch agent to collect metrics and logs. For more
information, see Collect Metrics from Amazon EC2 Instances and On-Premises Servers with the
CloudWatch Agent in the Amazon CloudWatch User Guide.

The monitoring scripts demonstrate how to produce and consume custom metrics for Amazon
CloudWatch. These sample Perl scripts comprise a fully functional example that reports memory, swap,
and disk space utilization metrics for a Linux instance.

Standard Amazon CloudWatch usage charges for custom metrics apply to your use of these scripts. For
more information, see the Amazon CloudWatch pricing page.

Contents
• Supported systems (p. 798)
• Required permissions (p. 799)
• Install required packages (p. 799)
• Install monitoring scripts (p. 800)
• mon-put-instance-data.pl (p. 801)
• mon-get-instance-stats.pl (p. 804)
• View your custom metrics in the console (p. 805)
• Troubleshoot (p. 805)

Supported systems
The monitoring scripts were tested on instances using the following systems. Using the monitoring
scripts on any other operating system is unsupported.

• Amazon Linux 2
• Amazon Linux AMI 2014.09.2 and later
• Red Hat Enterprise Linux 6.9 and 7.4
• SUSE Linux Enterprise Server 12

798
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Deprecated: Collect metrics using
the CloudWatch monitoring scripts
• Ubuntu Server 14.04 and 16.04

Required permissions
Ensure that the scripts have permission to call the following actions by associating an IAM role with your
instance:

• cloudwatch:PutMetricData
• cloudwatch:GetMetricStatistics
• cloudwatch:ListMetrics
• ec2:DescribeTags

For more information, see Work with IAM roles (p. 1034).

Install required packages

With some versions of Linux, you must install additional Perl modules before you can use the monitoring
scripts.

To install the required packages on Amazon Linux 2 and Amazon Linux AMI

1. Log on to your instance. For more information, see Connect to your Linux instance (p. 598).
2. At a command prompt, install packages as follows:

sudo yum install -y perl-Switch perl-DateTime perl-Sys-Syslog perl-LWP-Protocol-https

perl-Digest-SHA.x86_64

To install the required packages on Ubuntu

1. Log on to your instance. For more information, see Connect to your Linux instance (p. 598).
2. At a command prompt, install packages as follows:

sudo apt-get update

sudo apt-get install unzip
sudo apt-get install libwww-perl libdatetime-perl

To install the required packages on Red Hat Enterprise Linux 7

1. Log on to your instance. For more information, see Connect to your Linux instance (p. 598).
2. At a command prompt, install packages as follows:

sudo yum install perl-Switch perl-DateTime perl-Sys-Syslog perl-LWP-Protocol-https

perl-Digest-SHA --enablerepo="rhui-REGION-rhel-server-optional" -y
sudo yum install zip unzip

To install the required packages on Red Hat Enterprise Linux 6.9

1. Log on to your instance. For more information, see Connect to your Linux instance (p. 598).
2. At a command prompt, install packages as follows:

799
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Deprecated: Collect metrics using
the CloudWatch monitoring scripts
sudo yum install perl-DateTime perl-CPAN perl-Net-SSLeay perl-IO-Socket-SSL perl-
Digest-SHA gcc -y
sudo yum install zip unzip

3. Run CPAN as an elevated user:

sudo cpan

Press ENTER through the prompts until you see the following prompt:

cpan[1]>

4. At the CPAN prompt, run each of the below commands: run one command and it installs, and when
you return to the CPAN prompt, run the next command. Press ENTER like before when prompted to
continue through the process:

cpan[1]> install YAML

cpan[2]> install LWP::Protocol::https
cpan[3]> install Sys::Syslog
cpan[4]> install Switch

To install the required packages on SUSE

1. Log on to your instance. For more information, see Connect to your Linux instance (p. 598).
2. On servers running SUSE Linux Enterprise Server 12, you might need to download the perl-
Switch package. You can download and install this package using the following commands:

wget http://download.opensuse.org/repositories/devel:/languages:/perl/SLE_12_SP3/
noarch/perl-Switch-2.17-32.1.noarch.rpm
sudo rpm -i perl-Switch-2.17-32.1.noarch.rpm

3. Install the required packages as follows:

sudo zypper install perl-Switch perl-DateTime

sudo zypper install –y "perl(LWP::Protocol::https)"

Install monitoring scripts

The following steps show you how to download, uncompress, and configure the CloudWatch Monitoring
Scripts on an EC2 Linux instance.

To download, install, and configure the monitoring scripts

1. At a command prompt, move to a folder where you want to store the monitoring scripts and run the
following command to download them:

curl https://aws-cloudwatch.s3.amazonaws.com/downloads/
CloudWatchMonitoringScripts-1.2.2.zip -O

2. Run the following commands to install the monitoring scripts you downloaded:

unzip CloudWatchMonitoringScripts-1.2.2.zip && \

rm CloudWatchMonitoringScripts-1.2.2.zip && \

800
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Deprecated: Collect metrics using
the CloudWatch monitoring scripts
cd aws-scripts-mon

The package for the monitoring scripts contains the following files:

• CloudWatchClient.pm – Shared Perl module that simplifies calling Amazon CloudWatch from other
scripts.
• mon-put-instance-data.pl – Collects system metrics on an Amazon EC2 instance (memory, swap, disk
space utilization) and sends them to Amazon CloudWatch.
• mon-get-instance-stats.pl – Queries Amazon CloudWatch and displays the most recent utilization
statistics for the EC2 instance on which this script is run.
• awscreds.template – File template for AWS credentials that stores your access key ID and secret access
key.
• LICENSE.txt – Text file containing the Apache 2.0 license.
• NOTICE.txt – Copyright notice.

mon-put-instance-data.pl
This script collects memory, swap, and disk space utilization data on the current system. It then makes a
remote call to Amazon CloudWatch to report the collected data as custom metrics.

Options

Name Description

--mem-util Collects and sends the MemoryUtilization metrics in percentages.

This metric counts memory allocated by applications and the
operating system as used, and also includes cache and buffer
memory as used if you specify the --mem-used-incl-cache-
buff option.

--mem-used Collects and sends the MemoryUsed metrics, reported in

megabytes. This metric counts memory allocated by applications
and the operating system as used, and also includes cache and
buffer memory as used if you specify the --mem-used-incl-
cache-buff option.

--mem-used-incl-cache- If you include this option, memory currently used for cache and
buff buffers is counted as "used" when the metrics are reported for --
mem-util, --mem-used, and --mem-avail.

--mem-avail Collects and sends the MemoryAvailable metrics, reported in

megabytes. This metric counts memory allocated by applications
and the operating system as used, and also includes cache and
buffer memory as used if you specify the --mem-used-incl-
cache-buff option.

--swap-util Collects and sends SwapUtilization metrics, reported in

percentages.

--swap-used Collects and sends SwapUsed metrics, reported in megabytes.

--disk-path=PATH Selects the disk on which to report.

PATH can specify a mount point or any file located on a mount

point for the filesystem that needs to be reported. For selecting
multiple disks, specify a --disk-path=PATH for each one of them.

801
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Deprecated: Collect metrics using
the CloudWatch monitoring scripts
Name Description
To select a disk for the filesystems mounted on / and /home, use
the following parameters:

--disk-path=/ --disk-path=/home

--disk-space-util Collects and sends the DiskSpaceUtilization metric for the selected
disks. The metric is reported in percentages.

Note that the disk utilization metrics calculated by this script differ
from the values calculated by the df -k -l command. If you find the
values from df -k -l more useful, you can change the calculations in
the script.

--disk-space-used Collects and sends the DiskSpaceUsed metric for the selected disks.
The metric is reported by default in gigabytes.

Due to reserved disk space in Linux operating systems, disk space

used and disk space available might not accurately add up to the
amount of total disk space.

--disk-space-avail Collects and sends the DiskSpaceAvailable metric for the selected
disks. The metric is reported in gigabytes.

Due to reserved disk space in the Linux operating systems, disk

space used and disk space available might not accurately add up to
the amount of total disk space.

--memory-units=UNITS Specifies units in which to report memory usage. If not specified,

memory is reported in megabytes. UNITS may be one of the
following: bytes, kilobytes, megabytes, gigabytes.

--disk-space- Specifies units in which to report disk space usage. If not specified,
units=UNITS disk space is reported in gigabytes. UNITS may be one of the
following: bytes, kilobytes, megabytes, gigabytes.

--aws-credential- Provides the location of the file containing AWS credentials.

file=PATH
This parameter cannot be used with the --aws-access-key-id
and --aws-secret-key parameters.

--aws-access-key- Specifies the AWS access key ID to use to identify the caller. Must be
id=VALUE used together with the --aws-secret-key option. Do not use this
option with the --aws-credential-file parameter.

--aws-secret-key=VALUE Specifies the AWS secret access key to use to sign the request to
CloudWatch. Must be used together with the --aws-access-key-
id option. Do not use this option with --aws-credential-file
parameter.

--aws-iam-role=VALUE Specifies the IAM role used to provide AWS credentials. The value
=VALUE is required. If no credentials are specified, the default IAM
role associated with the EC2 instance is applied. Only one IAM role
can be used. If no IAM roles are found, or if more than one IAM role
is found, the script will return an error.

Do not use this option with the --aws-credential-file, --

aws-access-key-id, or --aws-secret-key parameters.

802
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Deprecated: Collect metrics using
the CloudWatch monitoring scripts
Name Description

--aggregated[=only] Adds aggregated metrics for instance type, AMI ID, and overall for
the Region. The value =only is optional; if specified, the script
reports only aggregated metrics.

--auto-scaling[=only] Adds aggregated metrics for the Auto Scaling group. The value
=only is optional; if specified, the script reports only Auto Scaling
metrics. The IAM policy associated with the IAM account or role
using the scripts need to have permissions to call the EC2 action
DescribeTags.

--verify Performs a test run of the script that collects the metrics, prepares
a complete HTTP request, but does not actually call CloudWatch
to report the data. This option also checks that credentials are
provided. When run in verbose mode, this option outputs the
metrics that will be sent to CloudWatch.

--from-cron Use this option when calling the script from cron. When this option
is used, all diagnostic output is suppressed, but error messages are
sent to the local system log of the user account.

--verbose Displays detailed information about what the script is doing.

--help Displays usage information.

--version Displays the version number of the script.

Examples

The following examples assume that you provided an IAM role or awscreds.conf file. Otherwise, you
must provide credentials using the --aws-access-key-id and --aws-secret-key parameters for
these commands.

The following example performs a simple test run without posting data to CloudWatch.

./mon-put-instance-data.pl --mem-util --verify --verbose

The following example collects all available memory metrics and sends them to CloudWatch, counting
cache and buffer memory as used

./mon-put-instance-data.pl --mem-used-incl-cache-buff --mem-util --mem-used --mem-avail

The following example collects aggregated metrics for an Auto Scaling group and sends them to Amazon
CloudWatch without reporting individual instance metrics.

./mon-put-instance-data.pl --mem-util --mem-used --mem-avail --auto-scaling=only

The following example collects aggregated metrics for instance type, AMI ID and region, and sends them
to Amazon CloudWatch without reporting individual instance metrics

./mon-put-instance-data.pl --mem-util --mem-used --mem-avail --aggregated=only

To set a cron schedule for metrics reported to CloudWatch, start editing the crontab using the crontab
-e command. Add the following command to report memory and disk space utilization to CloudWatch
every five minutes:

803
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Deprecated: Collect metrics using
the CloudWatch monitoring scripts
*/5 * * * * ~/aws-scripts-mon/mon-put-instance-data.pl --mem-used-incl-cache-buff --mem-
util --disk-space-util --disk-path=/ --from-cron

If the script encounters an error, it writes the error message in the system log.

mon-get-instance-stats.pl
This script queries CloudWatch for statistics on memory, swap, and disk space metrics within the time
interval provided using the number of most recent hours. This data is provided for the Amazon EC2
instance on which this script is run.

Options

Name Description

--recent-hours=N Specifies the number of recent hours to report on, as represented

by N where N is an integer.

--aws-credential- Provides the location of the file containing AWS credentials.

file=PATH

--aws-access-key- Specifies the AWS access key ID to use to identify the caller. Must be
id=VALUE used together with the --aws-secret-key option. Do not use this
option with the --aws-credential-file option.

--aws-secret-key=VALUE Specifies the AWS secret access key to use to sign the request to
CloudWatch. Must be used together with the --aws-access-key-
id option. Do not use this option with --aws-credential-file
option.

--aws-iam-role=VALUE Specifies the IAM role used to provide AWS credentials. The value
=VALUE is required. If no credentials are specified, the default IAM
role associated with the EC2 instance is applied. Only one IAM role
can be used. If no IAM roles are found, or if more than one IAM role
is found, the script will return an error.

Do not use this option with the --aws-credential-file, --

aws-access-key-id, or --aws-secret-key parameters.

--verify Performs a test run of the script. This option also checks that
credentials are provided.

--verbose Displays detailed information about what the script is doing.

--help Displays usage information.

--version Displays the version number of the script.

Example

To get utilization statistics for the last 12 hours, run the following command:

./mon-get-instance-stats.pl --recent-hours=12

The following is an example response:

Instance metric statistics for the last 12 hours.

804
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Logging API calls with AWS CloudTrail

CPU Utilization
Average: 1.06%, Minimum: 0.00%, Maximum: 15.22%

Memory Utilization
Average: 6.84%, Minimum: 6.82%, Maximum: 6.89%

Swap Utilization
Average: N/A, Minimum: N/A, Maximum: N/A

Disk Space Utilization on /dev/xvda1 mounted as /

Average: 9.69%, Minimum: 9.69%, Maximum: 9.69%

View your custom metrics in the console

After you successfully run the mon-put-instance-data.pl script, you can view your custom metrics
in the Amazon CloudWatch console.

To view custom metrics

1. Run mon-put-instance-data.pl as described previously.

2. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.
3. Choose View Metrics.
4. For Viewing, your custom metrics posted by the script are displayed with the prefix System/Linux.

Troubleshoot
The CloudWatchClient.pm module caches instance metadata locally. If you create an AMI from an
instance where you have run the monitoring scripts, any instances launched from the AMI within the
cache TTL (default: six hours, 24 hours for Auto Scaling groups) emit metrics using the instance ID of
the original instance. After the cache TTL time period passes, the script retrieves fresh data and the
monitoring scripts use the instance ID of the current instance. To immediately correct this, remove the
cached data using the following command:

rm /var/tmp/aws-mon/instance-id

Log Amazon EC2 and Amazon EBS API calls with

AWS CloudTrail
Amazon EC2 and Amazon EBS are integrated with AWS CloudTrail, a service that provides a record of
actions taken by a user, role, or an AWS service in Amazon EC2 and Amazon EBS. CloudTrail captures all
API calls for Amazon EC2 and Amazon EBS as events, including calls from the console and from code
calls to the APIs. If you create a trail, you can enable continuous delivery of CloudTrail events to an
Amazon S3 bucket, including events for Amazon EC2 and Amazon EBS. If you don't configure a trail, you
can still view the most recent events in the CloudTrail console in Event history. Using the information
collected by CloudTrail, you can determine the request that was made to Amazon EC2 and Amazon
EBS, the IP address from which the request was made, who made the request, when it was made, and
additional details.

To learn more about CloudTrail, see the AWS CloudTrail User Guide.

805
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Amazon EC2 and Amazon EBS information in CloudTrail

Amazon EC2 and Amazon EBS information in

CloudTrail
CloudTrail is enabled on your AWS account when you create the account. When activity occurs in Amazon
EC2 and Amazon EBS, that activity is recorded in a CloudTrail event along with other AWS service events
in Event history. You can view, search, and download recent events in your AWS account. For more
information, see Viewing Events with CloudTrail Event History.

For an ongoing record of events in your AWS account, including events for Amazon EC2 and Amazon
EBS, create a trail. A trail enables CloudTrail to deliver log files to an Amazon S3 bucket. By default, when
you create a trail in the console, the trail applies to all Regions. The trail logs events from all Regions in
the AWS partition and delivers the log files to the Amazon S3 bucket that you specify. Additionally, you
can configure other AWS services to further analyze and act upon the event data collected in CloudTrail
logs. For more information, see:

• Overview for Creating a Trail

• CloudTrail Supported Services and Integrations
• Configuring Amazon SNS Notifications for CloudTrail
• Receiving CloudTrail Log Files from Multiple Regions and Receiving CloudTrail Log Files from Multiple
Accounts

All Amazon EC2 actions, and Amazon EBS management actions, are logged by CloudTrail and
are documented in the Amazon EC2 API Reference. For example, calls to the RunInstances,
DescribeInstances, or CreateImage actions generate entries in the CloudTrail log files.

Every event or log entry contains information about who generated the request. The identity
information helps you determine the following:

• Whether the request was made with root or IAM user credentials.
• Whether the request was made with temporary security credentials for a role or federated user.
• Whether the request was made by another AWS service.

For more information, see the CloudTrail userIdentity Element.

Understand Amazon EC2 and Amazon EBS log file

entries
A trail is a configuration that enables delivery of events as log files to an Amazon S3 bucket that you
specify. CloudTrail log files contain one or more log entries. An event represents a single request from
any source and includes information about the requested action, the date and time of the action, request
parameters, and so on. CloudTrail log files are not an ordered stack trace of the public API calls, so they
do not appear in any specific order.

The following log file record shows that a user terminated an instance.

{
"Records":[
{
"eventVersion":"1.03",
"userIdentity":{
"type":"Root",
"principalId":"123456789012",
"arn":"arn:aws:iam::123456789012:root",

806
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Auditing users that connect via EC2 Instance Connect

"accountId":"123456789012",
"accessKeyId":"AKIAIOSFODNN7EXAMPLE",
"userName":"user"
},
"eventTime":"2016-05-20T08:27:45Z",
"eventSource":"ec2.amazonaws.com",
"eventName":"TerminateInstances",
"awsRegion":"us-west-2",
"sourceIPAddress":"198.51.100.1",
"userAgent":"aws-cli/1.10.10 Python/2.7.9 Windows/7botocore/1.4.1",
"requestParameters":{
"instancesSet":{
"items":[{
"instanceId":"i-1a2b3c4d"
}]
}
},
"responseElements":{
"instancesSet":{
"items":[{
"instanceId":"i-1a2b3c4d",
"currentState":{
"code":32,
"name":"shutting-down"
},
"previousState":{
"code":16,
"name":"running"
}
}]
}
},
"requestID":"be112233-1ba5-4ae0-8e2b-1c302EXAMPLE",
"eventID":"6e12345-2a4e-417c-aa78-7594fEXAMPLE",
"eventType":"AwsApiCall",
"recipientAccountId":"123456789012"
}
]
}

Use AWS CloudTrail to audit users that connect via

EC2 Instance Connect
Use AWS CloudTrail to audit the users that connect to your instances via EC2 Instance Connect.

To audit SSH activity via EC2 Instance Connect using the AWS CloudTrail console

1. Open the AWS CloudTrail console at https://console.aws.amazon.com/cloudtrail/.

2. Verify that you are in the correct Region.
3. In the navigation pane, choose Event history.
4. For Filter, choose Event source, ec2-instance-connect.amazonaws.com.
5. (Optional) For Time range, select a time range.
6. Choose the Refresh events icon.
7. The page displays the events that correspond to the SendSSHPublicKey API calls. Expand an event
using the arrow to view additional details, such as the user name and AWS access key that was used
to make the SSH connection, and the source IP address.
8. To display the full event information in JSON format, choose View event. The requestParameters
field contains the destination instance ID, OS user name, and public key that were used to make the
SSH connection.

807
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Auditing users that connect via EC2 Instance Connect

{
"eventVersion": "1.05",
"userIdentity": {
"type": "IAMUser",
"principalId": "ABCDEFGONGNOMOOCB6XYTQEXAMPLE",
"arn": "arn:aws:iam::1234567890120:user/IAM-friendly-name",
"accountId": "123456789012",
"accessKeyId": "ABCDEFGUKZHNAW4OSN2AEXAMPLE",
"userName": "IAM-friendly-name",
"sessionContext": {
"attributes": {
"mfaAuthenticated": "false",
"creationDate": "2018-09-21T21:37:58Z"}
}
},
"eventTime": "2018-09-21T21:38:00Z",
"eventSource": "ec2-instance-connect.amazonaws.com",
"eventName": "SendSSHPublicKey ",
"awsRegion": "us-west-2",
"sourceIPAddress": "123.456.789.012",
"userAgent": "aws-cli/1.15.61 Python/2.7.10 Darwin/16.7.0 botocore/1.10.60",
"requestParameters": {
"instanceId": "i-0123456789EXAMPLE",
"osUser": "ec2-user",
"SSHKey": {
"publicKey": "ssh-rsa ABCDEFGHIJKLMNO01234567890EXAMPLE"
}
},
"responseElements": null,
"requestID": "1a2s3d4f-bde6-11e8-a892-f7ec64543add",
"eventID": "1a2w3d4r5-a88f-4e28-b3bf-30161f75be34",
"eventType": "AwsApiCall",
"recipientAccountId": "0987654321"
}

If you have configured your AWS account to collect CloudTrail events in an S3 bucket, you can
download and audit the information programmatically. For more information, see Getting and
Viewing Your CloudTrail Log Files in the AWS CloudTrail User Guide.

808
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance IP addressing

Networking in Amazon EC2

Amazon EC2 provides the following networking features.

Features
• Amazon EC2 instance IP addressing (p. 809)
• Bring your own IP addresses (BYOIP) in Amazon EC2 (p. 825)
• Elastic IP addresses (p. 831)
• Elastic network interfaces (p. 839)
• Enhanced networking on Linux (p. 864)
• Elastic Fabric Adapter (p. 895)
• Placement groups (p. 927)
• Network maximum transmission unit (MTU) for your EC2 instance (p. 939)
• Virtual private clouds (p. 942)
• EC2-Classic (p. 943)

Amazon EC2 instance IP addressing

Amazon EC2 and Amazon VPC support both the IPv4 and IPv6 addressing protocols. By default, Amazon
EC2 and Amazon VPC use the IPv4 addressing protocol; you can't disable this behavior. When you create
a VPC, you must specify an IPv4 CIDR block (a range of private IPv4 addresses). You can optionally assign
an IPv6 CIDR block to your VPC and subnets, and assign IPv6 addresses from that block to instances in
your subnet. IPv6 addresses are reachable over the Internet. For more information about IPv6, see IP
Addressing in Your VPC in the Amazon VPC User Guide.

Contents
• Private IPv4 addresses and internal DNS hostnames (p. 809)
• Public IPv4 addresses and external DNS hostnames (p. 810)
• Elastic IP addresses (IPv4) (p. 811)
• Amazon DNS server (p. 811)
• IPv6 addresses (p. 811)
• Work with the IPv4 addresses for your instances (p. 812)
• Work with the IPv6 addresses for your instances (p. 815)
• Multiple IP addresses (p. 817)

Private IPv4 addresses and internal DNS hostnames

A private IPv4 address is an IP address that's not reachable over the Internet. You can use private IPv4
addresses for communication between instances in the same VPC. For more information about the
standards and specifications of private IPv4 addresses, see RFC 1918. We allocate private IPv4 addresses
to instances using DHCP.
Note
You can create a VPC with a publicly routable CIDR block that falls outside of the private IPv4
address ranges specified in RFC 1918. However, for the purposes of this documentation, we refer

809
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Public IPv4 addresses and external DNS hostnames

to private IPv4 addresses (or 'private IP addresses') as the IP addresses that are within the IPv4
CIDR range of your VPC.

When you launch an instance, we allocate a primary private IPv4 address for the instance. Each instance
is also given an internal DNS hostname that resolves to the primary private IPv4 address; for example,
ip-10-251-50-12.ec2.internal. You can use the internal DNS hostname for communication
between instances in the same VPC, but we can't resolve the internal DNS hostname outside of the VPC.

An instance receives a primary private IP address from the IPv4 address range of the subnet. For more
information, see VPC and subnet sizing in the Amazon VPC User Guide. If you don't specify a primary
private IP address when you launch the instance, we select an available IP address in the subnet's IPv4
range for you. Each instance has a default network interface (eth0) that is assigned the primary private
IPv4 address. You can also specify additional private IPv4 addresses, known as secondary private IPv4
addresses. Unlike primary private IP addresses, secondary private IP addresses can be reassigned from
one instance to another. For more information, see Multiple IP addresses (p. 817).

A private IPv4 address, regardless of whether it is a primary or secondary address, remains associated
with the network interface when the instance is stopped and started, or hibernated and started, and is
released when the instance is terminated.

Public IPv4 addresses and external DNS hostnames

A public IP address is an IPv4 address that's reachable from the Internet. You can use public addresses for
communication between your instances and the Internet.

Each instance that receives a public IP address is also given an external DNS hostname; for example,
ec2-203-0-113-25.compute-1.amazonaws.com. We resolve an external DNS hostname to the
public IP address of the instance from outside its VPC, and to the private IPv4 address of the instance
from inside its VPC. The public IP address is mapped to the primary private IP address through network
address translation (NAT). For more information, see RFC 1631: The IP Network Address Translator (NAT).

When you launch an instance in a default VPC, we assign it a public IP address by default. When you
launch an instance into a nondefault VPC, the subnet has an attribute that determines whether instances
launched into that subnet receive a public IP address from the public IPv4 address pool. By default, we
don't assign a public IP address to instances launched in a nondefault subnet.

You can control whether your instance receives a public IP address as follows:

• Modifying the public IP addressing attribute of your subnet. For more information, see Modifying the
public IPv4 addressing attribute for your subnet in the Amazon VPC User Guide.
• Enabling or disabling the public IP addressing feature during launch, which overrides the subnet's
public IP addressing attribute. For more information, see Assign a public IPv4 address during instance
launch (p. 814).

A public IP address is assigned to your instance from Amazon's pool of public IPv4 addresses, and is not
associated with your AWS account. When a public IP address is disassociated from your instance, it is
released back into the public IPv4 address pool, and you cannot reuse it.

You cannot manually associate or disassociate a public IP address from your instance. Instead, in certain
cases, we release the public IP address from your instance, or assign it a new one:

• We release your instance's public IP address when it is stopped, hibernated, or terminated. Your
stopped or hibernated instance receives a new public IP address when it is started.
• We release your instance's public IP address when you associate an Elastic IP address with it. When you
disassociate the Elastic IP address from your instance, it receives a new public IP address.
• If the public IP address of your instance in a VPC has been released, it will not receive a new one if
there is more than one network interface attached to your instance.

810
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Elastic IP addresses (IPv4)

• If your instance's public IP address is released while it has a secondary private IP address that is
associated with an Elastic IP address, the instance does not receive a new public IP address.

If you require a persistent public IP address that can be associated to and from instances as you require,
use an Elastic IP address instead.

If you use dynamic DNS to map an existing DNS name to a new instance's public IP address, it might take
up to 24 hours for the IP address to propagate through the Internet. As a result, new instances might
not receive traffic while terminated instances continue to receive requests. To solve this problem, use an
Elastic IP address. You can allocate your own Elastic IP address, and associate it with your instance. For
more information, see Elastic IP addresses (p. 831).

If you assign an Elastic IP address to an instance, it receives an IPv4 DNS hostname if DNS hostnames are
enabled. For more information, see Using DNS with your VPC in the Amazon VPC User Guide.
Note
Instances that access other instances through their public NAT IP address are charged for
regional or Internet data transfer, depending on whether the instances are in the same Region.

Elastic IP addresses (IPv4)

An Elastic IP address is a public IPv4 address that you can allocate to your account. You can associate
it to and disassociate it from instances as you require. It's allocated to your account until you choose
to release it. For more information about Elastic IP addresses and how to use them, see Elastic IP
addresses (p. 831).

We do not support Elastic IP addresses for IPv6.

Amazon DNS server

Amazon provides a DNS server that resolves Amazon-provided IPv4 DNS hostnames to IPv4 addresses.
The Amazon DNS server is located at the base of your VPC network range plus two. For more
information, see Amazon DNS server in the Amazon VPC User Guide.

IPv6 addresses
You can optionally associate an IPv6 CIDR block with your VPC, and associate IPv6 CIDR blocks with
your subnets. The IPv6 CIDR block for your VPC is automatically assigned from Amazon's pool of IPv6
addresses; you cannot choose the range yourself. For more information, see the following topics in the
Amazon VPC User Guide:

• VPC and subnet sizing for IPv6

• Associating an IPv6 CIDR block with your VPC
• Associating an IPv6 CIDR block with your subnet

IPv6 addresses are globally unique, and therefore reachable over the Internet. Your instance receives an
IPv6 address if an IPv6 CIDR block is associated with your VPC and subnet, and if one of the following is
true:

• Your subnet is configured to automatically assign an IPv6 address to an instance during launch. For
more information, see Modifying the IPv6 addressing attribute for your subnet.
• You assign an IPv6 address to your instance during launch.
• You assign an IPv6 address to the primary network interface of your instance after launch.
• You assign an IPv6 address to a network interface in the same subnet, and attach the network
interface to your instance after launch.

811
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with the IPv4 addresses for your instances

When your instance receives an IPv6 address during launch, the address is associated with the primary
network interface (eth0) of the instance. You can disassociate the IPv6 address from the network
interface. We do not support IPv6 DNS hostnames for your instance.

An IPv6 address persists when you stop and start, or hibernate and start, your instance, and is released
when you terminate your instance. You cannot reassign an IPv6 address while it's assigned to another
network interface—you must first unassign it.

You can assign additional IPv6 addresses to your instance by assigning them to a network interface
attached to your instance. The number of IPv6 addresses you can assign to a network interface and
the number of network interfaces you can attach to an instance varies per instance type. For more
information, see IP addresses per network interface per instance type (p. 841).

Work with the IPv4 addresses for your instances

You can assign a public IPv4 address to your instance when you launch it. You can view the IPv4
addresses for your in the console through either the Instances page or the Network Interfaces page.

Contents
• View the IPv4 addresses (p. 812)
• Assign a public IPv4 address during instance launch (p. 814)

View the IPv4 addresses

You can use the Amazon EC2 console to view the private IPv4 addresses, public IPv4 addresses, and
Elastic IP addresses of your instances. You can also determine the public IPv4 and private IPv4 addresses
of your instance from within your instance by using instance metadata. For more information, see
Instance metadata and user data (p. 704).

The public IPv4 address is displayed as a property of the network interface in the console, but it's
mapped to the primary private IPv4 address through NAT. Therefore, if you inspect the properties of your
network interface on your instance, for example, through ifconfig (Linux) or ipconfig (Windows), the
public IPv4 address is not displayed. To determine your instance's public IPv4 address from an instance,
use instance metadata.

New console

To view the IPv4 addresses for an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and select your instance.
3. The following information is available on the Networking tab:

• Public IPv4 address — The public IPv4 address. If you associated an Elastic IP address with
the instance or the primary network interface, this is the Elastic IP address.
• Public IPv4 DNS — The external DNS hostname.
• Private IPv4 addresses — The private IPv4 address.
• Private IPv4 DNS — The internal DNS hostname.
• Secondary private IPv4 addresses — Any secondary private IPv4 addresses.
• Elastic IP addresses — Any associated Elastic IP addresses.
4. Alternatively, under Network interfaces on the Networking tab, choose the interface ID for the
primary network interface (for example, eni-123abc456def78901). The following information is
available:

• Private DNS (IPv4) — The internal DNS hostname.

812
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with the IPv4 addresses for your instances

• Primary private IPv4 IP — The primary private IPv4 address.

• Secondary private IPv4 IPs — Any secondary private IPv4 addresses.
• Public DNS — The external DNS hostname.
• IPv4 Public IP — The public IPv4 address. If you associated an Elastic IP address with the
instance or the primary network interface, this is the Elastic IP address.
• Elastic IPs — Any associated Elastic IP addresses.

Old console

To view the IPv4 addresses for an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and select your instance.
3. The following information is available on the Description tab:

• Private DNS — The internal DNS hostname.

• Private IPs — The private IPv4 address.
• Secondary private IPs — Any secondary private IPv4 addresses.
• Public DNS — The external DNS hostname.
• IPv4 Public IP — The public IPv4 address. If you associated an Elastic IP address with the
instance or the primary network interface, this is the Elastic IP address.
• Elastic IPs — Any associated Elastic IP addresses.
4. Alternatively, you can view the IPv4 addresses for the instance using the primary network
interface. Under Network interfaces on the Description tab, choose eth0, and then choose the
interface ID (for example, eni-123abc456def78901). The following information is available:

• Private DNS (IPv4) — The internal DNS hostname.

• Primary private IPv4 IP — The primary private IPv4 address.
• Secondary private IPv4 IPs — Any secondary private IPv4 addresses.
• Public DNS — The external DNS hostname.
• IPv4 Public IP — The public IPv4 address. If you associated an Elastic IP address with the
instance or the primary network interface, this is the Elastic IP address.
• Elastic IPs — Any associated Elastic IP addresses.

To view the IPv4 addresses for an instance using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• describe-instances (AWS CLI)

• Get-EC2Instance (AWS Tools for Windows PowerShell).

To determine your instance's IPv4 addresses using instance metadata

1. Connect to your instance. For more information, see Connect to your Linux instance (p. 598).
2. Use the following command to access the private IP address:

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-

aws-ec2-metadata-token-ttl-seconds: 21600"` \

813
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with the IPv4 addresses for your instances

&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/

meta-data/local-ipv4

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/local-ipv4

3. Use the following command to access the public IP address:

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-

aws-ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/
meta-data/public-ipv4

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/public-ipv4

If an Elastic IP address is associated with the instance, the value returned is that of the Elastic IP
address.

Assign a public IPv4 address during instance launch

Each subnet has an attribute that determines whether instances launched into that subnet are assigned
a public IP address. By default, nondefault subnets have this attribute set to false, and default subnets
have this attribute set to true. When you launch an instance, a public IPv4 addressing feature is also
available for you to control whether your instance is assigned a public IPv4 address; you can override
the default behavior of the subnet's IP addressing attribute. The public IPv4 address is assigned from
Amazon's pool of public IPv4 addresses, and is assigned to the network interface with the device index of
eth0. This feature depends on certain conditions at the time you launch your instance.

Considerations

• You can't manually disassociate the public IP address from your instance after launch. Instead, it's
automatically released in certain cases, after which you cannot reuse it. For more information, see
Public IPv4 addresses and external DNS hostnames (p. 810). If you require a persistent public IP
address that you can associate or disassociate at will, assign an Elastic IP address to the instance after
launch instead. For more information, see Elastic IP addresses (p. 831).
• You cannot auto-assign a public IP address if you specify more than one network interface.
Additionally, you cannot override the subnet setting using the auto-assign public IP feature if you
specify an existing network interface for eth0.
• The public IP addressing feature is only available during launch. However, whether you assign a public
IP address to your instance during launch or not, you can associate an Elastic IP address with your
instance after it's launched. For more information, see Elastic IP addresses (p. 831). You can also
modify your subnet's public IPv4 addressing behavior. For more information, see Modifying the public
IPv4 addressing attribute for your subnet.

To enable or disable the public IP addressing feature using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Launch Instance.

814
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with the IPv6 addresses for your instances

3. Select an AMI and an instance type, and then choose Next: Configure Instance Details.
4. On the Configure Instance Details page, for Network, select a VPC. The Auto-assign Public IP list is
displayed. Choose Enable or Disable to override the default setting for the subnet.
5. Follow the steps on the next pages of the wizard to complete your instance's setup. For more
information about the wizard configuration options, see Launch an instance using the Launch
Instance Wizard (p. 526). On the final Review Instance Launch page, review your settings, and then
choose Launch to choose a key pair and launch your instance.
6. On the Instances page, select your new instance and view its public IP address in IPv4 Public IP field
in the details pane.

To enable or disable the public IP addressing feature using the command line
You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• Use the --associate-public-ip-address or the --no-associate-public-ip-address

option with the run-instances command (AWS CLI)
• Use the -AssociatePublicIp parameter with the New-EC2Instance command (AWS Tools for
Windows PowerShell)

Work with the IPv6 addresses for your instances

You can view the IPv6 addresses assigned to your instance, assign a public IPv6 address to your instance,
or unassign an IPv6 address from your instance. You can view these addresses in the console through
either the Instances page or the Network Interfaces page.

Contents
• View the IPv6 addresses (p. 815)
• Assign an IPv6 address to an instance (p. 816)
• Unassign an IPv6 address from an instance (p. 817)

View the IPv6 addresses

You can use the Amazon EC2 console, AWS CLI, and instance metadata to view the IPv6 addresses for
your instances.

New console

To view the IPv4 addresses for an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance.
4. On the Networking tab, locate IPv6 addresses.
5. Alternatively, under Network interfaces on the Networking tab, choose the interface ID for the
network interface (for example, eni-123abc456def78901). Locate IPv6 IPs.

Old console

To view the IPv4 addresses for an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

815
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with the IPv6 addresses for your instances

2. In the navigation pane, choose Instances.

3. Select the instance.
4. On the Networking tab, locate IPv6 IPs.
5. Alternatively, under Network interfaces on the Description tab, choose eth0, and then choose
the interface ID (for example, eni-123abc456def78901). Locate IPv6 IPs.

To view the IPv6 addresses for an instance using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• describe-instances (AWS CLI)

• Get-EC2Instance (AWS Tools for Windows PowerShell).

To view the IPv6 addresses for an instance using instance metadata

1. Connect to your instance. For more information, see Connect to your Linux instance (p. 598).
2. Use the following command to view the IPv6 address (you can get the MAC address from
http://169.254.169.254/latest/meta-data/network/interfaces/macs/).

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-

aws-ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/
meta-data/network/interfaces/macs/mac-address/ipv6s

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/network/interfaces/

macs/mac-address/ipv6s

Assign an IPv6 address to an instance

If your VPC and subnet have IPv6 CIDR blocks associated with them, you can assign an IPv6 address to
your instance during or after launch. The IPv6 address is assigned from the IPv6 address range of the
subnet, and is assigned to the network interface with the device index of eth0.

To assign an IPv6 address to an instance during launch

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Select an AMI and an instance type that supports IPv6, and choose Next: Configure Instance
Details.
3. On the Configure Instance Details page, for Network, select a VPC and for Subnet, select a subnet.
For Auto-assign IPv6 IP, choose Enable.
4. Follow the remaining steps in the wizard to launch your instance.

To assign an IPv6 address to an instance after launch

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.

816
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Multiple IP addresses

3. Select your instance, and choose Actions, Networking, Manage IP addresses.

4. Expand the network interface. Under IPv6 addresses, choose Assign new IP address. Enter an IPv6
address from the range of the subnet or leave the field blank to let Amazon choose an IPv6 address
for you.
5. Choose Save.

To assign an IPv6 address using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• Use the --ipv6-addresses option with the run-instances command (AWS CLI)
• Use the Ipv6Addresses property for -NetworkInterface in the New-EC2Instance command (AWS
Tools for Windows PowerShell)
• assign-ipv6-addresses (AWS CLI)
• Register-EC2Ipv6AddressList (AWS Tools for Windows PowerShell)

Unassign an IPv6 address from an instance

You can unassign an IPv6 address from an instance at any time.

To unassign an IPv6 address from an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select your instance, and choose Actions, Networking, Manage IP addresses.
4. Expand the network interface. Under IPv6 addresses, choose Unassign next to the IPv6 address.
5. Choose Save.

To unassign an IPv6 address from an instance using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• unassign-ipv6-addresses (AWS CLI)

• Unregister-EC2Ipv6AddressList (AWS Tools for Windows PowerShell).

Multiple IP addresses
You can specify multiple private IPv4 and IPv6 addresses for your instances. The number of network
interfaces and private IPv4 and IPv6 addresses that you can specify for an instance depends on the
instance type. For more information, see IP addresses per network interface per instance type (p. 841).

It can be useful to assign multiple IP addresses to an instance in your VPC to do the following:

• Host multiple websites on a single server by using multiple SSL certificates on a single server and
associating each certificate with a specific IP address.
• Operate network appliances, such as firewalls or load balancers, that have multiple IP addresses for
each network interface.
• Redirect internal traffic to a standby instance in case your instance fails, by reassigning the secondary
IP address to the standby instance.

817
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Multiple IP addresses

Contents
• How multiple IP addresses work (p. 818)
• Work with multiple IPv4 addresses (p. 818)
• Work with multiple IPv6 addresses (p. 822)

How multiple IP addresses work

The following list explains how multiple IP addresses work with network interfaces:

• You can assign a secondary private IPv4 address to any network interface. The network interface need
not be attached to the instance.
• You can assign multiple IPv6 addresses to a network interface that's in a subnet that has an associated
IPv6 CIDR block.
• You must choose a secondary IPv4 address from the IPv4 CIDR block range of the subnet for the
network interface.
• You must choose IPv6 addresses from the IPv6 CIDR block range of the subnet for the network
interface.
• You associate security groups with network interfaces, not individual IP addresses. Therefore, each IP
address you specify in a network interface is subject to the security group of its network interface.
• Multiple IP addresses can be assigned and unassigned to network interfaces attached to running or
stopped instances.
• Secondary private IPv4 addresses that are assigned to a network interface can be reassigned to
another one if you explicitly allow it.
• An IPv6 address cannot be reassigned to another network interface; you must first unassign the IPv6
address from the existing network interface.
• When assigning multiple IP addresses to a network interface using the command line tools or API, the
entire operation fails if one of the IP addresses can't be assigned.
• Primary private IPv4 addresses, secondary private IPv4 addresses, Elastic IP addresses, and IPv6
addresses remain with a secondary network interface when it is detached from an instance or attached
to an instance.
• Although you can't detach the primary network interface from an instance, you can reassign the
secondary private IPv4 address of the primary network interface to another network interface.

The following list explains how multiple IP addresses work with Elastic IP addresses (IPv4 only):

• Each private IPv4 address can be associated with a single Elastic IP address, and vice versa.
• When a secondary private IPv4 address is reassigned to another interface, the secondary private IPv4
address retains its association with an Elastic IP address.
• When a secondary private IPv4 address is unassigned from an interface, an associated Elastic IP
address is automatically disassociated from the secondary private IPv4 address.

Work with multiple IPv4 addresses

You can assign a secondary private IPv4 address to an instance, associate an Elastic IPv4 address with a
secondary private IPv4 address, and unassign a secondary private IPv4 address.

Contents
• Assign a secondary private IPv4 address (p. 819)

818
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Multiple IP addresses

• Configure the operating system on your instance to recognize secondary private IPv4
addresses (p. 820)
• Associate an Elastic IP address with the secondary private IPv4 address (p. 821)
• View your secondary private IPv4 addresses (p. 821)
• Unassign a secondary private IPv4 address (p. 821)

Assign a secondary private IPv4 address

You can assign the secondary private IPv4 address to the network interface for an instance as you launch
the instance, or after the instance is running. This section includes the following procedures.

• To assign a secondary private IPv4 address when launching an instance (p. 819)
• To assign a secondary IPv4 address during launch using the command line (p. 820)
• To assign a secondary private IPv4 address to a network interface (p. 820)
• To assign a secondary private IPv4 to an existing instance using the command line (p. 820)

To assign a secondary private IPv4 address when launching an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Launch Instance.
3. Select an AMI, then choose an instance type and choose Next: Configure Instance Details.
4. On the Configure Instance Details page, for Network, select a VPC and for Subnet, select a subnet.
5. In the Network Interfaces section, do the following, and then choose Next: Add Storage:

• To add another network interface, choose Add Device. The console enables you to specify up
to two network interfaces when you launch an instance. After you launch the instance, choose
Network Interfaces in the navigation pane to add additional network interfaces. The total
number of network interfaces that you can attach varies by instance type. For more information,
see IP addresses per network interface per instance type (p. 841).
Important
When you add a second network interface, the system can no longer auto-assign a public
IPv4 address. You will not be able to connect to the instance over IPv4 unless you assign
an Elastic IP address to the primary network interface (eth0). You can assign the Elastic
IP address after you complete the Launch wizard. For more information, see Work with
Elastic IP addresses (p. 832).
• For each network interface, under Secondary IP addresses, choose Add IP, and then enter a
private IP address from the subnet range, or accept the default Auto-assign value to let Amazon
select an address.
6. On the next Add Storage page, you can specify volumes to attach to the instance besides the
volumes specified by the AMI (such as the root device volume), and then choose Next: Add Tags.
7. On the Add Tags page, specify tags for the instance, such as a user-friendly name, and then choose
Next: Configure Security Group.
8. On the Configure Security Group page, select an existing security group or create a new one.
Choose Review and Launch.
9. On the Review Instance Launch page, review your settings, and then choose Launch to choose a key
pair and launch your instance. If you're new to Amazon EC2 and haven't created any key pairs, the
wizard prompts you to create one.

Important
After you have added a secondary private IP address to a network interface, you must connect
to the instance and configure the secondary private IP address on the instance itself. For more

819
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Multiple IP addresses

information, see Configure the operating system on your instance to recognize secondary
private IPv4 addresses (p. 820).

To assign a secondary IPv4 address during launch using the command line

• You can use one of the following commands. For more information about these command line
interfaces, see Access Amazon EC2 (p. 3).

• The --secondary-private-ip-addresses option with the run-instances command (AWS CLI)

• Define -NetworkInterface and specify the PrivateIpAddresses parameter with the New-
EC2Instance command (AWS Tools for Windows PowerShell).

To assign a secondary private IPv4 address to a network interface

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces, and then select the network interface attached
to the instance.
3. Choose Actions, Manage IP Addresses.
4. Under IPv4 Addresses, choose Assign new IP.
5. Enter a specific IPv4 address that's within the subnet range for the instance, or leave the field blank
to let Amazon select an IP address for you.
6. (Optional) Choose Allow reassignment to allow the secondary private IP address to be reassigned if
it is already assigned to another network interface.
7. Choose Yes, Update.

Alternatively, you can assign a secondary private IPv4 address to an instance. Choose Instances in the
navigation pane, select the instance, and then choose Actions, Networking, Manage IP Addresses. You
can configure the same information as you did in the steps above. The IP address is assigned to the
primary network interface (eth0) for the instance.

To assign a secondary private IPv4 to an existing instance using the command line

• You can use one of the following commands. For more information about these command line
interfaces, see Access Amazon EC2 (p. 3).

• assign-private-ip-addresses (AWS CLI)

• Register-EC2PrivateIpAddress (AWS Tools for Windows PowerShell)

Configure the operating system on your instance to recognize secondary private

IPv4 addresses
After you assign a secondary private IPv4 address to your instance, you need to configure the operating
system on your instance to recognize the secondary private IP address.

• If you are using Amazon Linux, the ec2-net-utils package can take care of this step for you. It
configures additional network interfaces that you attach while the instance is running, refreshes
secondary IPv4 addresses during DHCP lease renewal, and updates the related routing rules. You can
immediately refresh the list of interfaces by using the command sudo service network restart
and then view the up-to-date list using ip addr li. If you require manual control over your network
configuration, you can remove the ec2-net-utils package. For more information, see Configure your
network interface using ec2-net-utils (p. 862).
• If you are using another Linux distribution, see the documentation for your Linux distribution. Search
for information about configuring additional network interfaces and secondary IPv4 addresses. If the

820
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Multiple IP addresses

instance has two or more interfaces on the same subnet, search for information about using routing
rules to work around asymmetric routing.

For information about configuring a Windows instance, see Configuring a secondary private IP address
for your Windows instance in a VPC in the Amazon EC2 User Guide for Windows Instances.

Associate an Elastic IP address with the secondary private IPv4 address

To associate an Elastic IP address with a secondary private IPv4 address

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Choose Actions, and then select Associate address.
4. For Network interface, select the network interface, and then select the secondary IP address from
the Private IP list.
5. Choose Associate.

To associate an Elastic IP address with a secondary private IPv4 address using the command
line

• You can use one of the following commands. For more information about these command line
interfaces, see Access Amazon EC2 (p. 3).

• associate-address (AWS CLI)

• Register-EC2Address (AWS Tools for Windows PowerShell)

View your secondary private IPv4 addresses

To view the private IPv4 addresses assigned to a network interface

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface with private IP addresses to view.
4. On the Details tab in the details pane, check the Primary private IPv4 IP and Secondary private
IPv4 IPs fields for the primary private IPv4 address and any secondary private IPv4 addresses
assigned to the network interface.

To view the private IPv4 addresses assigned to an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance with private IPv4 addresses to view.
4. On the Description tab in the details pane, check the Private IPs and Secondary private IPs fields
for the primary private IPv4 address and any secondary private IPv4 addresses assigned to the
instance through its network interface.

Unassign a secondary private IPv4 address

If you no longer require a secondary private IPv4 address, you can unassign it from the instance or the
network interface. When a secondary private IPv4 address is unassigned from a network interface, the
Elastic IP address (if it exists) is also disassociated.

821
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Multiple IP addresses

To unassign a secondary private IPv4 address from an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select an instance, choose Actions, Networking, Manage IP Addresses.
4. Under IPv4 Addresses, choose Unassign for the IPv4 address to unassign.
5. Choose Yes, Update.

To unassign a secondary private IPv4 address from a network interface

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface, choose Actions, Manage IP Addresses.
4. Under IPv4 Addresses, choose Unassign for the IPv4 address to unassign.
5. Choose Yes, Update.

To unassign a secondary private IPv4 address using the command line

• You can use one of the following commands. For more information about these command line
interfaces, see Access Amazon EC2 (p. 3).

• unassign-private-ip-addresses (AWS CLI)

• Unregister-EC2PrivateIpAddress (AWS Tools for Windows PowerShell)

Work with multiple IPv6 addresses

You can assign multiple IPv6 addresses to your instance, view the IPv6 addresses assigned to your
instance, and unassign IPv6 addresses from your instance.

Contents
• Assign multiple IPv6 addresses (p. 822)
• View your IPv6 addresses (p. 824)
• Unassign an IPv6 address (p. 824)

Assign multiple IPv6 addresses

You can assign one or more IPv6 addresses to your instance during launch or after launch. To assign
an IPv6 address to an instance, the VPC and subnet in which you launch the instance must have an
associated IPv6 CIDR block. For more information, see VPCs and Subnets in the Amazon VPC User Guide.

To assign multiple IPv6 addresses during launch

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the dashboard, choose Launch Instance.
3. Select an AMI, choose an instance type, and choose Next: Configure Instance Details. Ensure that
you choose an instance type that support IPv6. For more information, see Instance types (p. 203).
4. On the Configure Instance Details page, select a VPC from the Network list, and a subnet from the
Subnet list.
5. In the Network Interfaces section, do the following, and then choose Next: Add Storage:

822
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Multiple IP addresses

• To assign a single IPv6 address to the primary network interface (eth0), under IPv6 IPs, choose
Add IP. To add a secondary IPv6 address, choose Add IP again. You can enter an IPv6 address from
the range of the subnet, or leave the default Auto-assign value to let Amazon choose an IPv6
address from the subnet for you.
• Choose Add Device to add another network interface and repeat the steps above to add one
or more IPv6 addresses to the network interface. The console enables you to specify up to two
network interfaces when you launch an instance. After you launch the instance, choose Network
Interfaces in the navigation pane to add additional network interfaces. The total number of
network interfaces that you can attach varies by instance type. For more information, see IP
addresses per network interface per instance type (p. 841).
6. Follow the next steps in the wizard to attach volumes and tag your instance.
7. On the Configure Security Group page, select an existing security group or create a new one. If
you want your instance to be reachable over IPv6, ensure that your security group has rules that
allow access from IPv6 addresses. For more information, see Security group rules for different use
cases (p. 1072). Choose Review and Launch.
8. On the Review Instance Launch page, review your settings, and then choose Launch to choose a key
pair and launch your instance. If you're new to Amazon EC2 and haven't created any key pairs, the
wizard prompts you to create one.

You can use the Instances screen Amazon EC2 console to assign multiple IPv6 addresses to an existing
instance. This assigns the IPv6 addresses to the primary network interface (eth0) for the instance. To
assign a specific IPv6 address to the instance, ensure that the IPv6 address is not already assigned to
another instance or network interface.

To assign multiple IPv6 addresses to an existing instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select your instance, choose Actions, Networking, Manage IP Addresses.
4. Under IPv6 Addresses, choose Assign new IP for each IPv6 address you want to add. You can specify
an IPv6 address from the range of the subnet, or leave the Auto-assign value to let Amazon choose
an IPv6 address for you.
5. Choose Yes, Update.

Alternatively, you can assign multiple IPv6 addresses to an existing network interface. The network
interface must have been created in a subnet that has an associated IPv6 CIDR block. To assign a specific
IPv6 address to the network interface, ensure that the IPv6 address is not already assigned to another
network interface.

To assign multiple IPv6 addresses to a network interface

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select your network interface, choose Actions, Manage IP Addresses.
4. Under IPv6 Addresses, choose Assign new IP for each IPv6 address you want to add. You can specify
an IPv6 address from the range of the subnet, or leave the Auto-assign value to let Amazon choose
an IPv6 address for you.
5. Choose Yes, Update.

823
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Multiple IP addresses

CLI overview

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• Assign an IPv6 address during launch:

• Use the --ipv6-addresses or --ipv6-address-count options with the run-instances command
(AWS CLI)
• Define -NetworkInterface and specify the Ipv6Addresses or Ipv6AddressCount parameters
with the New-EC2Instance command (AWS Tools for Windows PowerShell).
• Assign an IPv6 address to a network interface:
• assign-ipv6-addresses (AWS CLI)
• Register-EC2Ipv6AddressList (AWS Tools for Windows PowerShell)

View your IPv6 addresses

You can view the IPv6 addresses for an instance or for a network interface.

To view the IPv6 addresses assigned to an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select your instance. In the details pane, review the IPv6 IPs field.

To view the IPv6 addresses assigned to a network interface

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select your network interface. In the details pane, review the IPv6 IPs field.

CLI overview

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• View the IPv6 addresses for an instance:

• describe-instances (AWS CLI)
• Get-EC2Instance (AWS Tools for Windows PowerShell).
• View the IPv6 addresses for a network interface:
• describe-network-interfaces (AWS CLI)
• Get-EC2NetworkInterface (AWS Tools for Windows PowerShell)

Unassign an IPv6 address

You can unassign an IPv6 address from the primary network interface of an instance, or you can unassign
an IPv6 address from a network interface.

To unassign an IPv6 address from an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.

824
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Bring your own IP addresses

3. Select your instance, choose Actions, Networking, Manage IP Addresses.

4. Under IPv6 Addresses, choose Unassign for the IPv6 address to unassign.
5. Choose Yes, Update.

To unassign an IPv6 address from a network interface

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select your network interface, choose Actions, Manage IP Addresses.
4. Under IPv6 Addresses, choose Unassign for the IPv6 address to unassign.
5. Choose Save.

CLI overview

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• unassign-ipv6-addresses (AWS CLI)

• Unregister-EC2Ipv6AddressList (AWS Tools for Windows PowerShell).

Bring your own IP addresses (BYOIP) in Amazon

EC2
You can bring part or all of your public IPv4 address range or IPv6 address range from your on-premises
network to your AWS account. You continue to own the address range, but AWS advertises it on the
internet by default. After you bring the address range to AWS, it appears in your account as an address
pool.

BYOIP is not available in all Regions and for all resources. For a list of supported Regions and resources,
see the FAQ for Bring Your Own IP.
Note
The following steps describe how to bring your own IP address range for use in Amazon EC2
only. For steps to bring your own IP address range for use in AWS Global Accelerator, see Bring
your own IP addresses (BYOIP) in the AWS Global Accelerator Developer Guide.

Topics
• Requirements (p. 825)
• Prepare to bring your address range to your AWS account (p. 826)
• Provision the address range for use with AWS (p. 828)
• Advertise the address range through AWS (p. 829)
• Work with your address range (p. 829)
• Deprovision the address range (p. 830)

Requirements
• The address range must be registered with your Regional internet registry (RIR), such as the American
Registry for Internet Numbers (ARIN), Réseaux IP Européens Network Coordination Centre (RIPE), or

825
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Prepare to bring your address range to your AWS account

Asia-Pacific Network Information Centre (APNIC). It must be registered to a business or institutional

entity and cannot be registered to an individual person.
• The most specific IPv4 address range that you can bring is /24.
• The most specific IPv6 address range that you can bring is /48 for CIDRs that are publicly advertised,
and /56 for CIDRs that are not publicly advertised (p. 828).
• You can bring each address range to one Region at a time.
• You can bring a total of five IPv4 and IPv6 address ranges per Region to your AWS account.
• You cannot share your IP address range with other accounts using AWS Resource Access Manager (AWS
RAM).
• The addresses in the IP address range must have a clean history. We might investigate the reputation
of the IP address range and reserve the right to reject an IP address range if it contains an IP address
that has a poor reputation or is associated with malicious behavior.
• You must own the IP address that you use. This means that only the following are supported:
• ARIN - "Direct Allocation" and "Direct Assignment" network types
• RIPE - "ALLOCATED PA", "LEGACY", "ASSIGNED PI", and "ALLOCATED-BY-RIR" allocation statuses
• APNIC – "ALLOCATED PORTABLE" and "ASSIGNED PORTABLE" allocation statuses

Prepare to bring your address range to your AWS

account
To ensure that only you can bring your address range to your AWS account, you must authorize Amazon
to advertise the address range. You must also provide proof that you own the address range through a
signed authorization message.

A Route Origin Authorization (ROA) is a cryptographic statement about your route announcements that
you can create through your RIR. It contains the address range, the Autonomous System numbers (ASN)
that are allowed to advertise the address range, and an expiration date. An ROA authorizes Amazon to
advertise an address range under a specific AS number. However, it does not authorize your AWS account
to bring the address range to AWS. To authorize your AWS account to bring an address range to AWS,
you must publish a self-signed X509 certificate in the Registry Data Access Protocol (RDAP) remarks for
the address range. The certificate contains a public key, which AWS uses to verify the authorization-
context signature that you provide. Keep your private key secure and use it to sign the authorization-
context message.

The commands in these tasks are supported on Linux. On Windows, you can use the Windows Subsystem
for Linux to run Linux commands.

Tasks
• Create a ROA object (p. 826)
• Create a self-signed X509 certificate (p. 827)
• Create a signed authorization message (p. 827)

Create a ROA object

Create a ROA object to authorize Amazon ASNs 16509 and 14618 to advertise your address range, plus
the ASNs that are currently authorized to advertise the address range. You must set the maximum length
to the size of the smallest prefix that you want to bring (for example, /24). It might take up to 24 hours
for the ROA to become available to Amazon. For more information, see the following:

• ARIN — ROA Requests

• RIPE — Managing ROAs

826
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Prepare to bring your address range to your AWS account

• APNIC — Route Management

Create a self-signed X509 certificate

Use the following procedure to create a self-signed X509 certificate and add it to the RDAP record for
your RIR. The openssl commands require OpenSSL version 1.0.2 or later.

Copy the commands below and replace only the placeholder values (in colored italic text).

To create a self-signed X509 certificate and add it to the RDAP record

1. Generate an RSA 2048-bit key pair as shown in the following.

openssl genrsa -out private.key 2048

2. Create a public X509 certificate from the key pair using the following command. In this example, the
certificate expires in 365 days, after which time it cannot be trusted. Be sure to set the expiration
appropriately. When prompted for information, you can accept the default values.

openssl req -new -x509 -key private.key -days 365 | tr -d "\n" > publickey.cer

3. Update the RDAP record for your RIR with the X509 certificate. Be sure to copy the -----BEGIN
CERTIFICATE----- and -----END CERTIFICATE----- from the certificate. Be sure that you
have removed newline characters, if you haven't already done so using the tr -d "\n" commands in
the previous steps. To view your certificate, run the following command.

cat publickey.cer

For ARIN, add the certificate in the "Public Comments" section for your address range. Do not add it
to the comments section for your organization.

For RIPE, add the certificate as a new "descr" field for your address range. Do not add it to the
comments section for your organization.

For APNIC, email the public key to [email protected] to manually add it to the "remarks" field for
your address range. Send the email using the APNIC authorized contact for the IP addresses.

Create a signed authorization message

The format of the signed authorization message is as follows, where the date is the expiry date of the
message.

1|aws|account|cidr|YYYYMMDD|SHA256|RSAPSS

To create a signed authorization message

1. Create a plaintext authorization message and store it in a variable named text_message as shown
in the following example. Copy the following example and replace only the example account
number, address range, and expiry date with your own values.

text_message="1|aws|123456789012|198.51.100.0/24|20191201|SHA256|RSAPSS"

2. Sign the authorization message in text_message using the key pair that you created, and store it in
a variable named signed_message.

827
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Provision the address range for use with AWS

signed_message=$(echo $text_message | tr -d "\n" | openssl dgst -sha256 -sigopt

rsa_padding_mode:pss -sigopt rsa_pss_saltlen:-1 -sign private.key -keyform PEM |
openssl base64 | tr -- '+=/' '-_~' | tr -d "\n")

Important
We recommend that you copy and paste this command. Do not modify or replace any of the
values.

Provision the address range for use with AWS

When you provision an address range for use with AWS, you are confirming that you own the address
range and are authorizing Amazon to advertise it. We also verify that you own the address range through
a signed authorization message. This message is signed with the self-signed X509 key pair that you used
when updating the RDAP record with the X509 certificate.

To provision the address range

Use the provision-byoip-cidr command. Replace the example address range with your own address
range. The --cidr-authorization-context option uses the variables that you created previously,
not the ROA message.

aws ec2 provision-byoip-cidr --cidr address-range --cidr-authorization-context

Message="$text_message",Signature="$signed_message"

Provisioning an address range is an asynchronous operation, so the call returns immediately, but the
address range is not ready to use until its status changes from pending-provision to provisioned.
Note
It can take up to three weeks to complete the provisioning process.

To monitor the status of the address ranges that you've provisioned

Use the describe-byoip-cidrs command.

aws ec2 describe-byoip-cidrs --max-results 5

If there are issues during provisioning and the status goes to failed-provision, you must run the
provision-byoip-cidr command again after the issues have been resolved.

Provision an IPv6 address range that's not publicly advertised

By default, an address range is provisioned to be publicly advertised to the internet. You can provision
an IPv6 address range that will not be publicly advertised. When you associate an IPv6 CIDR block from a
non-public address range with a VPC, the IPv6 CIDR can only be accessed through an AWS Direct Connect
connection.

An ROA is not required to provision a non-public address range.

To provision an IPv6 address range that will not be publicly advertised, use the following provision-
byoip-cidr command.

aws ec2 provision-byoip-cidr --cidr address-range --cidr-authorization-context

Message="$text_message",Signature="$signed_message" --no-publicly-advertisable

828
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Advertise the address range through AWS

Important
You can only set the publicly-advertisable or no-publicly-advertisable flag during
provisioning. You cannot change the advertisable status of an address range later.

Advertise the address range through AWS

After the address range is provisioned, it is ready to be advertised. You must advertise the exact address
range that you provisioned. You can't advertise only a portion of the provisioned address range.

If you provisioned an IPv6 address range that will not be publicly advertised, you do not need to
complete this step.

We recommend that you stop advertising the address range from other locations before you advertise
it through AWS. If you keep advertising your IP address range from other locations, we can't reliably
support it or troubleshoot issues. Specifically, we can't guarantee that traffic to the address range will
enter our network.

To minimize down time, you can configure your AWS resources to use an address from your address
pool before it is advertised, and then simultaneously stop advertising it from the current location and
start advertising it through AWS. For more information about allocating an Elastic IP address from your
address pool, see Allocate an Elastic IP address (p. 832).

To advertise the address range, use the following advertise-byoip-cidr command.

aws ec2 advertise-byoip-cidr --cidr address-range

Important
You can run the advertise-byoip-cidr command at most once every 10 seconds, even if you
specify different address ranges each time.

To stop advertising the address range, use the following withdraw-byoip-cidr command.

aws ec2 withdraw-byoip-cidr --cidr address-range

Important
You can run the withdraw-byoip-cidr command at most once every 10 seconds, even if you
specify different address ranges each time.

Work with your address range

You can view and work with the IPv4 and IPv6 address ranges that you've provisioned in your account.

IPv4 address ranges

You can create an Elastic IP address from your IPv4 address pool and use it with your AWS resources,
such as EC2 instances, NAT gateways, and Network Load Balancers.

To view information about the IPv4 address pools that you've provisioned in your account, use the
following describe-public-ipv4-pools command.

aws ec2 describe-public-ipv4-pools

To create an Elastic IP address from your IPv4 address pool, use the allocate-address command. You can
use the --public-ipv4-pool option to specify the ID of the address pool returned by describe-
byoip-cidrs. Or you can use the --address option to specify an address from the address range that
you provisioned.

829
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Deprovision the address range

IPv6 address ranges

To view information about the IPv6 address pools that you've provisioned in your account, use the
following describe-ipv6-pools command.

aws ec2 describe-ipv6-pools

To create a VPC and specify an IPv6 CIDR from your IPv6 address pool, use the following create-vpc
command. To let Amazon choose the IPv6 CIDR from your IPv6 address pool, omit the --ipv6-cidr-
block option.

aws ec2 create-vpc --cidr-block 10.0.0.0/16 --ipv6-cidr-block ipv6-cidr --ipv6-pool pool-id

To associate an IPv6 CIDR block from your IPv6 address pool with a VPC, use the following associate-
vpc-cidr-block command. To let Amazon choose the IPv6 CIDR from your IPv6 address pool, omit the --
ipv6-cidr-block option.

aws ec2 associate-vpc-cidr-block --vpc-id vpc-123456789abc123ab --ipv6-cidr-block ipv6-cidr

--ipv6-pool pool-id

To view your VPCs and the associated IPv6 address pool information, use the describe-vpcs command. To
view information about associated IPv6 CIDR blocks from a specific IPv6 address pool, use the following
get-associated-ipv6-pool-cidrs command.

aws ec2 get-associated-ipv6-pool-cidrs --pool-id pool-id

If you disassociate the IPv6 CIDR block from your VPC, it's released back into your IPv6 address pool.

For more information about working with IPv6 CIDR blocks in the VPC console, see Working with VPCs
and Subnets in the Amazon VPC User Guide.

Deprovision the address range

To stop using your address range with AWS, first release any Elastic IP addresses and disassociate any
IPv6 CIDR blocks that are still allocated from the address pool. Then stop advertising the address range,
and finally, deprovision the address range.

You cannot deprovision a portion of the address range. If you want to use a more specific address range
with AWS, deprovision the entire address range and provision a more specific address range.

(IPv4) To release each Elastic IP address, use the following release-address command.

aws ec2 release-address --allocation-id eipalloc-12345678abcabcabc

(IPv6) To disassociate an IPv6 CIDR block, use the following disassociate-vpc-cidr-block command.

aws ec2 disassociate-vpc-cidr-block --association-id vpc-cidr-assoc-12345abcd1234abc1

To stop advertising the address range, use the following withdraw-byoip-cidr command.

aws ec2 withdraw-byoip-cidr --cidr address-range

To deprovision the address range, use the following deprovision-byoip-cidr command.

aws ec2 deprovision-byoip-cidr --cidr address-range

830
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Elastic IP addresses

It can take up to a day to deprovision an address range.

Elastic IP addresses
An Elastic IP address is a static IPv4 address designed for dynamic cloud computing. By using an Elastic IP
address, you can mask the failure of an instance or software by rapidly remapping the address to another
instance in your account. An Elastic IP address is allocated to your AWS account, and is yours until you
release it.

An Elastic IP address is a public IPv4 address, which is reachable from the internet. If your instance does
not have a public IPv4 address, you can associate an Elastic IP address with your instance to enable
communication with the internet. For example, this allows you to connect to your instance from your
local computer.

We currently do not support Elastic IP addresses for IPv6.

To ensure efficient use of Elastic IP addresses, we impose a small hourly charge if an Elastic IP address
is not associated with a running instance, or if it is associated with a stopped instance or an unattached
network interface. While your instance is running, you are not charged for one Elastic IP address
associated with the instance, but you are charged for any additional Elastic IP addresses associated with
the instance. For more information, see Elastic IP Addresses on the Amazon EC2 Pricing, On-Demand
Pricing page.

Contents
• Elastic IP address basics (p. 831)
• Work with Elastic IP addresses (p. 832)
• Use reverse DNS for email applications (p. 838)
• Elastic IP address limit (p. 838)

Elastic IP address basics

The following are the basic characteristics of an Elastic IP address:

• An Elastic IP address is static; it does not change over time.

• To use an Elastic IP address, you first allocate one to your account, and then associate it with your
instance or a network interface.
• When you associate an Elastic IP address with an instance, it is also associated with the instance's
primary network interface. When you associate an Elastic IP address with a network interface that is
attached to an instance, it is also associated with the instance.
• When you associate an Elastic IP address with an instance or its primary network interface, the
instance's public IPv4 address (if it had one) is released back into Amazon's pool of public IPv4
addresses. You cannot reuse a public IPv4 address, and you cannot convert a public IPv4 address
to an Elastic IP address. For more information, see Public IPv4 addresses and external DNS
hostnames (p. 810).
• You can disassociate an Elastic IP address from a resource, and then associate it with a different
resource. To avoid unexpected behavior, ensure that all active connections to the resource named in
the existing association are closed before you make the change. After you have associated your Elastic
IP address to a different resource, you can reopen your connections to the newly associated resource.
• A disassociated Elastic IP address remains allocated to your account until you explicitly release it.
• When you associate an Elastic IP address with an instance that previously had a public IPv4 address,
the public DNS host name of the instance changes to match the Elastic IP address.

831
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with Elastic IP addresses

• We resolve a public DNS host name to the public IPv4 address or the Elastic IP address of the instance
outside the network of the instance, and to the private IPv4 address of the instance from within the
network of the instance.
• An Elastic IP address comes from Amazon's pool of IPv4 addresses, or from a custom IP address pool
that you have brought to your AWS account.
• When you allocate an Elastic IP address from an IP address pool that you have brought to your AWS
account, it does not count toward your Elastic IP address limits. For more information, see Elastic IP
address limit (p. 838).
• When you allocate the Elastic IP addresses, you can associate the Elastic IP addresses with a network
border group. This is the location from which we advertise the CIDR block. Setting the network border
group limits the CIDR block to this group. If you do not specify the network border group, we set the
border group containing all of the Availability Zones in the Region (for example, us-west-2).
• An Elastic IP address is for use in a specific network border group only.
• An Elastic IP address is for use in a specific Region only, and cannot be moved to a different Region.

Work with Elastic IP addresses

The following sections describe how you can work with Elastic IP addresses.

Tasks
• Allocate an Elastic IP address (p. 832)
• Describe your Elastic IP addresses (p. 833)
• Tage an Elastic IP address (p. 834)
• Associate an Elastic IP address with a running instance or network interface (p. 835)
• Disassociate an Elastic IP address (p. 836)
• Release an Elastic IP address (p. 837)
• Recover an Elastic IP address (p. 837)

Allocate an Elastic IP address

You can allocate an Elastic IP address from Amazon's pool of public IPv4 addresses, or from a custom IP
address pool that you have brought to your AWS account. For more information about bringing your own
IP address range to your AWS account, see Bring your own IP addresses (BYOIP) in Amazon EC2 (p. 825).

You can allocate an Elastic IP address using one of the following methods.

New console

To allocate an Elastic IP address

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Choose Allocate Elastic IP address.
4. For Scope, choose either VPC or EC2-Classic, depending on the scope in which it will be used.
5. (VPC scope only) For Public IPv4 address pool, choose one of the following:

• Amazon's pool of IPv4 addresses—If you want an IPv4 address to be allocated from
Amazon's pool of IPv4 addresses.
• My pool of public IPv4 addresses—If you want to allocate an IPv4 address from an IP address
pool that you have brought to your AWS account. This option is disabled if you do not have
any IP address pools.

832
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with Elastic IP addresses

• Customer owned pool of IPv4 addresses—If you want to allocate an IPv4 address from a
pool created from your on-premises network for use with an AWS Outpost. This option is
disabled if you do not have an AWS Outpost.
6. (Optional) Add or remove a tag.

[Add a tag] Choose Add new tag and do the following:

• For Key, enter the key name.

• For Value, enter the key value.

[Remove a tag] Choose Remove to the right of the tag’s Key and Value.
7. Choose Allocate.

Old console

To allocate an Elastic IP address

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Choose Allocate new address.
4. For IPv4 address pool, choose Amazon pool.
5. Choose Allocate, and close the confirmation screen.

AWS CLI

To allocate an Elastic IP address

Use the allocate-address AWS CLI command.

PowerShell

To allocate an Elastic IP address

Use the New-EC2Address AWS Tools for Windows PowerShell command.

Describe your Elastic IP addresses

You can describe an Elastic IP address using one of the following methods.

New console

To describe your Elastic IP addresses

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select the Elastic IP address to view and choose Actions, View details.

Old console

To describe your Elastic IP addresses

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.

833
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with Elastic IP addresses

3. Select a filter from the Resource Attribute list to begin searching. You can use multiple filters in
a single search.

AWS CLI

To describe your Elastic IP addresses

Use the describe-addresses AWS CLI command.

PowerShell

To describe your Elastic IP addresses

Use the Get-EC2Address AWS Tools for Windows PowerShell command.

Tage an Elastic IP address

You can assign custom tags to your Elastic IP addresses to categorize them in different ways, for
example, by purpose, owner, or environment. This helps you to quickly find a specific Elastic IP address
based on the custom tags that you assigned to it.

You can only tag Elastic IP addresses that are in the VPC scope.
Note
Cost allocation tracking using Elastic IP address tags is not supported.

You can tag an Elastic IP address using one of the following methods.

New console

To tag an Elastic IP address

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select the Elastic IP address to tag and choose Actions, View details.
4. In the Tags section, choose Manage tags.
5. Specify a tag key and value pair.
6. (Optional) Choose Add tag to add additional tags.
7. Choose Save.

Old console

To tag an Elastic IP address

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select the Elastic IP address to tag and choose Tags.
4. Choose Add/Edit Tags.
5. In the Add/Edit Tags dialog box, choose Create Tag, and then specify the key and value for the
tag.
6. (Optional) Choose Create Tag to add additional tags to the Elastic IP address.
7. Choose Save.

AWS CLI

To tag an Elastic IP address

834
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with Elastic IP addresses

Use the create-tags AWS CLI command.

aws ec2 create-tags --resources eipalloc-12345678 --tags Key=Owner,Value=TeamA

PowerShell

To tag an Elastic IP address

Use the New-EC2Tag AWS Tools for Windows PowerShell command.

The New-EC2Tag command needs a Tag parameter, which specifies the key and value pair to be
used for the Elastic IP address tag. The following commands create the Tag parameter.

PS C:\> $tag = New-Object Amazon.EC2.Model.Tag

PS C:\> $tag.Key = "Owner"
PS C:\> $tag.Value = "TeamA"

PS C:\> New-EC2Tag -Resource eipalloc-12345678 -Tag $tag

Associate an Elastic IP address with a running instance or

network interface
If you're associating an Elastic IP address with your instance to enable communication with the internet,
you must also ensure that your instance is in a public subnet. For more information, see Internet
Gateways in the Amazon VPC User Guide.

You can associate an Elastic IP address with an instance or network interface using one of the following
methods.

New console

To associate an Elastic IP address with an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select the Elastic IP address to associate and choose Actions, Associate Elastic IP address.
4. For Resource type, choose Instance.
5. For instance, choose the instance with which to associate the Elastic IP address. You can also
enter text to search for a specific instance.
6. (Optional) For Private IP address, specify a private IP address with which to associate the Elastic
IP address.
7. Choose Associate.

To associate an Elastic IP address with a network interface

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select the Elastic IP address to associate and choose Actions, Associate Elastic IP address.
4. For Resource type, choose Network interface.
5. For Network interface, choose the network interface with which to associate the Elastic IP
address. You can also enter text to search for a specific network interface.

835
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with Elastic IP addresses

6. (Optional) For Private IP address, specify a private IP address with which to associate the Elastic
IP address.
7. Choose Associate.

Old console

To associate an Elastic IP address with an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select an Elastic IP address and choose Actions, Associate address.
4. Select the instance from Instance and then choose Associate.

AWS CLI

To associate an Elastic IP address

Use the associate-address AWS CLI command.

PowerShell

To associate an Elastic IP address

Use the Register-EC2Address AWS Tools for Windows PowerShell command.

Disassociate an Elastic IP address

You can disassociate an Elastic IP address from an instance or network interface at any time. After you
disassociate the Elastic IP address, you can reassociate it with another resource.

You can disassociate an Elastic IP address using one of the following methods.

New console

To disassociate and reassociate an Elastic IP address

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select the Elastic IP address to disassociate, choose Actions, Disassociate Elastic IP address.
4. Choose Disassociate.

Old console

To disassociate and reassociate an Elastic IP address

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select the Elastic IP address, choose Actions, and then select Disassociate address.
4. Choose Disassociate address.

AWS CLI

To disassociate an Elastic IP address

836
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with Elastic IP addresses

Use the disassociate-address AWS CLI command.

PowerShell

To disassociate an Elastic IP address

Use the Unregister-EC2Address AWS Tools for Windows PowerShell command.

Release an Elastic IP address

If you no longer need an Elastic IP address, we recommend that you release it using one of the following
methods. The address to release must not be currently associated with an AWS resource, such as an EC2
instance, NAT gateway, or Network Load Balancer.

New console

To release an Elastic IP address

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select the Elastic IP address to release and choose Actions, Release Elastic IP addresses.
4. Choose Release.

Old console

To release an Elastic IP address

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select the Elastic IP address, choose Actions, and then select Release addresses. Choose
Release when prompted.

AWS CLI

To release an Elastic IP address

Use the release-address AWS CLI command.

PowerShell

To release an Elastic IP address

Use the Remove-EC2Address AWS Tools for Windows PowerShell command.

Recover an Elastic IP address

If you have released your Elastic IP address, you might be able to recover it. The following rules apply:

• You cannot recover an Elastic IP address if it has been allocated to another AWS account, or if it will
result in your exceeding your Elastic IP address limit.
• You cannot recover tags associated with an Elastic IP address.
• You can recover an Elastic IP address using the Amazon EC2 API or a command line tool only.

AWS CLI

To recover an Elastic IP address

837
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Use reverse DNS for email applications

Use the allocate-address AWS CLI command and specify the IP address using the --address
parameter as follows.

aws ec2 allocate-address --domain vpc --address 203.0.113.3

PowerShell

To recover an Elastic IP address

Use the New-EC2Address AWS Tools for Windows PowerShell command and specify the IP address
using the -Address parameter as follows.

PS C:\> New-EC2Address -Address 203.0.113.3 -Domain vpc -Region us-east-1

Use reverse DNS for email applications

If you intend to send email to third parties from an instance, we suggest that you provision one or
more Elastic IP addresses and provide them to AWS. AWS works with ISPs and internet anti-spam
organizations to reduce the chance that your email sent from these addresses will be flagged as spam.

In addition, assigning a static reverse DNS record to your Elastic IP address that is used to send email can
help avoid having email flagged as spam by some anti-spam organizations. Note that a corresponding
forward DNS record (record type A) pointing to your Elastic IP address must exist before we can create
your reverse DNS record.

If a reverse DNS record is associated with an Elastic IP address, the Elastic IP address is locked to your
account and cannot be released from your account until the record is removed.

To remove email sending limits, or to provide us with your Elastic IP addresses and reverse DNS records,
go to the Request to Remove Email Sending Limitations page.

Elastic IP address limit

By default, all AWS accounts are limited to five (5) Elastic IP addresses per Region, because public (IPv4)
internet addresses are a scarce public resource. We strongly encourage you to use an Elastic IP address
primarily for the ability to remap the address to another instance in the case of instance failure, and to
use DNS hostnames for all other inter-node communication.

To verify how many Elastic IP addresses are in use

Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/ and choose Elastic IPs from the
navigation pane.

To verify your current account limit for Elastic IP addresses

You can verify your limit in either the Amazon EC2 console or the Service Quotas console. Do one of the
following:

• Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

Choose Limits from the navigation pane, and then enter IP in the search field. The limit is EC2-VPC
Elastic IPs. If you have access to EC2-Classic, there is an additional limit, EC2-Classic Elastic IPs.
• Open the Service Quotas console at https://console.aws.amazon.com/servicequotas/.

On the Dashboard, choose Amazon Elastic Compute Cloud (Amazon EC2). If Amazon Elastic Compute
Cloud (Amazon EC2) is not listed on the Dashboard, choose AWS services, enter EC2 in the search
field, and then choose Amazon Elastic Compute Cloud (Amazon EC2).

838
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Network interfaces

On the Amazon EC2 service quotas page, enter IP in the search field. The limit is EC2-VPC Elastic
IPs. If you have access to EC2-Classic, there is an additional limit, EC2-Classic Elastic IPs. For more
information, choose the limit.

If you think your architecture warrants additional Elastic IP addresses, you can request a quota increase
directly from the Service Quotas console.

Elastic network interfaces

An elastic network interface is a logical networking component in a VPC that represents a virtual network
card. It can include the following attributes:

• A primary private IPv4 address from the IPv4 address range of your VPC
• One or more secondary private IPv4 addresses from the IPv4 address range of your VPC
• One Elastic IP address (IPv4) per private IPv4 address
• One public IPv4 address
• One or more IPv6 addresses
• One or more security groups
• A MAC address
• A source/destination check flag
• A description

You can create and configure network interfaces in your account and attach them to instances in your
VPC. Your account might also have requester-managed network interfaces, which are created and
managed by AWS services to enable you to use other resources and services. You cannot manage these
network interfaces yourself. For more information, see Requester-managed network interfaces (p. 863).

This AWS resource is referred to as a network interface in the AWS Management Console and the Amazon
EC2 API. Therefore, we use "network interface" in this documentation instead of "elastic network
interface". The term "network interface" in this documentation always means "elastic network interface".

Contents
• Network interface basics (p. 839)
• Network cards (p. 841)
• IP addresses per network interface per instance type (p. 841)
• Work with network interfaces (p. 854)
• Scenarios for network interfaces (p. 860)
• Best practices for configuring network interfaces (p. 862)
• Requester-managed network interfaces (p. 863)

Network interface basics

You can create a network interface, attach it to an instance, detach it from an instance, and attach it to
another instance. The attributes of a network interface follow it as it's attached or detached from an
instance and reattached to another instance. When you move a network interface from one instance to
another, network traffic is redirected to the new instance.

Primary network interface

839
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Network interface basics

Each instance has a default network interface, called the primary network interface. You cannot detach
a primary network interface from an instance. You can create and attach additional network interfaces.
The maximum number of network interfaces that you can use varies by instance type. For more
information, see IP addresses per network interface per instance type (p. 841).

Public IPv4 addresses for network interfaces

In a VPC, all subnets have a modifiable attribute that determines whether network interfaces created in
that subnet (and therefore instances launched into that subnet) are assigned a public IPv4 address. For
more information, see IP addressing behavior for your subnet in the Amazon VPC User Guide. The public
IPv4 address is assigned from Amazon's pool of public IPv4 addresses. When you launch an instance, the
IP address is assigned to the primary network interface that's created.

When you create a network interface, it inherits the public IPv4 addressing attribute from the subnet.
If you later modify the public IPv4 addressing attribute of the subnet, the network interface keeps the
setting that was in effect when it was created. If you launch an instance and specify an existing network
interface as the primary network interface, the public IPv4 address attribute is determined by this
network interface.

For more information, see Public IPv4 addresses and external DNS hostnames (p. 810).

Elastic IP addresses for network interface

If you have an Elastic IP address, you can associate it with one of the private IPv4 addresses for the
network interface. You can associate one Elastic IP address with each private IPv4 address.

If you disassociate an Elastic IP address from a network interface, you can release it back to the address
pool. This is the only way to associate an Elastic IP address with an instance in a different subnet or VPC,
as network interfaces are specific to subnets.

IPv6 addresses for network interfaces

If you associate IPv6 CIDR blocks with your VPC and subnet, you can assign one or more IPv6 addresses
from the subnet range to a network interface. Each IPv6 address can be assigned to one network
interface.

All subnets have a modifiable attribute that determines whether network interfaces created in that
subnet (and therefore instances launched into that subnet) are automatically assigned an IPv6 address
from the range of the subnet. For more information, see IP addressing behavior for your subnet in the
Amazon VPC User Guide. When you launch an instance, the IPv6 address is assigned to the primary
network interface that's created.

For more information, see IPv6 addresses (p. 811).

Termination behavior

You can set the termination behavior for a network interface that's attached to an instance. You can
specify whether the network interface should be automatically deleted when you terminate the instance
to which it's attached.

Source/destination checking

Disabling source/destination checking enables an instance to handle network traffic that isn't specifically
destined for the instance. For example, instances running services such as network address translation,
routing, or a firewall should disable the source/destination check attribute. This attribute is enabled by
default.

Monitoring IP traffic

You can enable a VPC flow log on your network interface to capture information about the IP traffic
going to and from a network interface. After you've created a flow log, you can view and retrieve its data
in Amazon CloudWatch Logs. For more information, see VPC Flow Logs in the Amazon VPC User Guide.

840
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Network cards

Network cards
Instances with multiple network cards provide higher network performance, including bandwidth
capabilities above 100 Gbps and improved packet rate performance. Each network interface is attached
to a network card. The primary network interface must be assigned to network card index 0.

If you enable Elastic Fabric Adapter (EFA) when you launch an instance that supports multiple network
cards, all network cards are available. You can assign up to one EFA per network card. An EFA counts as a
network interface.

The following instances support multiple network cards. All other instance types support one network
card.

Instance type Number of network cards

P4d 4

IP addresses per network interface per instance type

The following table lists the maximum number of network interfaces per instance type, and the
maximum number of private IPv4 addresses and IPv6 addresses per network interface. The limit for IPv6
addresses is separate from the limit for private IPv4 addresses per network interface. Not all instance
types support IPv6 addressing. Network interfaces, multiple private IPv4 addresses, and IPv6 addresses
are only available for instances running in a VPC. IPv6 addresses are public and reachable over the
Internet. For more information, see Multiple IP addresses (p. 817). For more information about IPv6 in
VPC, see IP Addressing in your VPC in the Amazon VPC User Guide.

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

a1.medium 2 4 4

a1.large 3 10 10

a1.xlarge 4 15 15

a1.2xlarge 4 15 15

a1.4xlarge 8 30 30

a1.metal 8 30 30

c1.medium 2 6 IPv6 not supported

c1.xlarge 4 15 IPv6 not supported

c3.large 3 10 10

c3.xlarge 4 15 15

c3.2xlarge 4 15 15

c3.4xlarge 8 30 30

c3.8xlarge 8 30 30

c4.large 3 10 10

841
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

c4.xlarge 4 15 15

c4.2xlarge 4 15 15

c4.4xlarge 8 30 30

c4.8xlarge 8 30 30

c5.large 3 10 10

c5.xlarge 4 15 15

c5.2xlarge 4 15 15

c5.4xlarge 8 30 30

c5.9xlarge 8 30 30

c5.12xlarge 8 30 30

c5.18xlarge 15 50 50

c5.24xlarge 15 50 50

c5.metal 15 50 50

c5a.large 3 10 10

c5a.xlarge 4 15 15

c5a.2xlarge 4 15 15

c5a.4xlarge 8 30 30

c5a.8xlarge 8 30 30

c5a.12xlarge 8 30 30

c5a.16xlarge 15 50 50

c5a.24xlarge 15 50 50

c5ad.large 3 10 10

c5ad.xlarge 4 15 15

c5ad.2xlarge 4 15 15

c5ad.4xlarge 8 30 30

c5ad.8xlarge 8 30 30

c5ad.12xlarge 8 30 30

c5ad.16xlarge 15 50 50

c5ad.24xlarge 15 50 50

c5d.large 3 10 10

842
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

c5d.xlarge 4 15 15

c5d.2xlarge 4 15 15

c5d.4xlarge 8 30 30

c5d.9xlarge 8 30 30

c5d.12xlarge 8 30 30

c5d.18xlarge 15 50 50

c5d.24xlarge 15 50 50

c5d.metal 15 50 50

c5n.large 3 10 10

c5n.xlarge 4 15 15

c5n.2xlarge 4 15 15

c5n.4xlarge 8 30 30

c5n.9xlarge 8 30 30

c5n.18xlarge 15 50 50

c5n.metal 15 50 50

c6g.medium 2 4 4

c6g.large 3 10 10

c6g.xlarge 4 15 15

c6g.2xlarge 4 15 15

c6g.4xlarge 8 30 30

c6g.8xlarge 8 30 30

c6g.12xlarge 8 30 30

c6g.16xlarge 15 50 50

c6g.metal 15 50 50

c6gd.medium 2 4 4

c6gd.large 3 10 10

c6gd.xlarge 4 15 15

c6gd.2xlarge 4 15 15

c6gd.4xlarge 8 30 30

c6gd.8xlarge 8 30 30

843
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

c6gd.12xlarge 8 30 30

c6gd.16xlarge 15 50 50

c6gd.metal 15 50 50

c6gn.medium 2 4 4

c6gn.large 3 10 10

c6gn.xlarge 4 15 15

c6gn.2xlarge 4 15 15

c6gn.4xlarge 8 30 30

c6gn.8xlarge 8 30 30

c6gn.12xlarge 8 30 30

c6gn.16xlarge 15 50 50

cc2.8xlarge 8 30 IPv6 not supported

cr1.8xlarge 8 30 IPv6 not supported

d2.xlarge 4 15 15

d2.2xlarge 4 15 15

d2.4xlarge 8 30 30

d2.8xlarge 8 30 30

d3.xlarge 4 3 3

d3.2xlarge 4 5 5

d3.4xlarge 4 10 10

d3.8xlarge 3 20 20

d3en.large 4 2 2

d3en.xlarge 4 3 3

d3en.2xlarge 4 5 5

d3en.4xlarge 4 10 10

d3en.6large 4 15 15

d3en.8xlarge 4 20 20

d3en.12xlarge 3 30 30

f1.2xlarge 4 15 15

f1.4xlarge 8 30 30

844
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

f1.16xlarge 8 50 50

g2.2xlarge 4 15 IPv6 not supported

g2.8xlarge 8 30 IPv6 not supported

g3s.xlarge 4 15 15

g3.4xlarge 8 30 30

g3.8xlarge 8 30 30

g3.16xlarge 15 50 50

g4ad.4xlarge 3 10 10

g4ad.8xlarge 4 15 15

g4ad.16xlarge 8 30 30

g4dn.xlarge 3 10 10

g4dn.2xlarge 3 10 10

g4dn.4xlarge 3 10 10

g4dn.8xlarge 4 15 15

g4dn.12xlarge 8 30 30

g4dn.16xlarge 4 15 15

g4dn.metal 15 50 50

h1.2xlarge 4 15 15

h1.4xlarge 8 30 30

h1.8xlarge 8 30 30

h1.16xlarge 15 50 50

hs1.8xlarge 8 30 IPv6 not supported

i2.xlarge 4 15 15

i2.2xlarge 4 15 15

i2.4xlarge 8 30 30

i2.8xlarge 8 30 30

i3.large 3 10 10

i3.xlarge 4 15 15

i3.2xlarge 4 15 15

i3.4xlarge 8 30 30

845
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

i3.8xlarge 8 30 30

i3.16xlarge 15 50 50

i3.metal 15 50 50

i3en.large 3 10 10

i3en.xlarge 4 15 15

i3en.2xlarge 4 15 15

i3en.3xlarge 4 15 15

i3en.6xlarge 8 30 30

i3en.12xlarge 8 30 30

i3en.24xlarge 15 50 50

i3en.metal 15 50 50

inf1.xlarge 4 10 10

inf1.2xlarge 4 10 10

inf1.6xlarge 8 30 30

inf1.24xlarge 15 30 30

m1.small 2 4 IPv6 not supported

m1.medium 2 6 IPv6 not supported

m1.large 3 10 IPv6 not supported

m1.xlarge 4 15 IPv6 not supported

m2.xlarge 4 15 IPv6 not supported

m2.2xlarge 4 30 IPv6 not supported

m2.4xlarge 8 30 IPv6 not supported

m3.medium 2 6 IPv6 not supported

m3.large 3 10 IPv6 not supported

m3.xlarge 4 15 IPv6 not supported

m3.2xlarge 4 30 IPv6 not supported

m4.large 2 10 10

m4.xlarge 4 15 15

m4.2xlarge 4 15 15

m4.4xlarge 8 30 30

846
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

m4.10xlarge 8 30 30

m4.16xlarge 8 30 30

m5.large 3 10 10

m5.xlarge 4 15 15

m5.2xlarge 4 15 15

m5.4xlarge 8 30 30

m5.8xlarge 8 30 30

m5.12xlarge 8 30 30

m5.16xlarge 15 50 50

m5.24xlarge 15 50 50

m5.metal 15 50 50

m5a.large 3 10 10

m5a.xlarge 4 15 15

m5a.2xlarge 4 15 15

m5a.4xlarge 8 30 30

m5a.8xlarge 8 30 30

m5a.12xlarge 8 30 30

m5a.16xlarge 15 50 50

m5a.24xlarge 15 50 50

m5ad.large 3 10 10

m5ad.xlarge 4 15 15

m5ad.2xlarge 4 15 15

m5ad.4xlarge 8 30 30

m5ad.8xlarge 8 30 30

m5ad.12xlarge 8 30 30

m5ad.16xlarge 15 50 50

m5ad.24xlarge 15 50 50

m5d.large 3 10 10

m5d.xlarge 4 15 15

m5d.2xlarge 4 15 15

847
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

m5d.4xlarge 8 30 30

m5d.8xlarge 8 30 30

m5d.12xlarge 8 30 30

m5d.16xlarge 15 50 50

m5d.24xlarge 15 50 50

m5d.metal 15 50 50

m5dn.large 3 10 10

m5dn.xlarge 4 15 15

m5dn.2xlarge 4 15 15

m5dn.4xlarge 8 30 30

m5dn.8xlarge 8 30 30

m5dn.12xlarge 8 30 30

m5dn.16xlarge 15 50 50

m5dn.24xlarge 15 50 50

m5n.large 3 10 10

m5n.xlarge 4 15 15

m5n.2xlarge 4 15 15

m5n.4xlarge 8 30 30

m5n.8xlarge 8 30 30

m5n.12xlarge 8 30 30

m5n.16xlarge 15 50 50

m5n.24xlarge 15 50 50

m5zn.large 3 10 10

m5zn.xlarge 4 15 15

m5zn.2xlarge 4 15 15

m5zn.3xlarge 8 30 30

m5zn.6xlarge 8 30 30

m5zn.12xlarge 15 50 50

m5zn.metal 15 50 50

m6g.medium 2 4 4

848
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

m6g.large 3 10 10

m6g.xlarge 4 15 15

m6g.2xlarge 4 15 15

m6g.4xlarge 8 30 30

m6g.8xlarge 8 30 30

m6g.12xlarge 8 30 30

m6g.16xlarge 15 50 50

m6g.metal 15 50 50

m6gd.medium 2 4 4

m6gd.large 3 10 10

m6gd.xlarge 4 15 15

m6gd.2xlarge 4 15 15

m6gd.4xlarge 8 30 30

m6gd.8xlarge 8 30 30

m6gd.12xlarge 8 30 30

m6gd.16xlarge 15 50 50

m6gd.metal 15 50 50

mac1.metal 8 30 30

p2.xlarge 4 15 15

p2.8xlarge 8 30 30

p2.16xlarge 8 30 30

p3.2xlarge 4 15 15

p3.8xlarge 8 30 30

p3.16xlarge 8 30 30

p3dn.24xlarge 15 50 50

p4d.24xlarge 4x15 50 50

r3.large 3 10 10

r3.xlarge 4 15 15

r3.2xlarge 4 15 15

r3.4xlarge 8 30 30

849
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

r3.8xlarge 8 30 30

r4.large 3 10 10

r4.xlarge 4 15 15

r4.2xlarge 4 15 15

r4.4xlarge 8 30 30

r4.8xlarge 8 30 30

r4.16xlarge 15 50 50

r5.large 3 10 10

r5.xlarge 4 15 15

r5.2xlarge 4 15 15

r5.4xlarge 8 30 30

r5.8xlarge 8 30 30

r5.12xlarge 8 30 30

r5.16xlarge 15 50 50

r5.24xlarge 15 50 50

r5.metal 15 50 50

r5a.large 3 10 10

r5a.xlarge 4 15 15

r5a.2xlarge 4 15 15

r5a.4xlarge 8 30 30

r5a.8xlarge 8 30 30

r5a.12xlarge 8 30 30

r5a.16xlarge 15 50 50

r5a.24xlarge 15 50 50

r5ad.large 3 10 10

r5ad.xlarge 4 15 15

r5ad.2xlarge 4 15 15

r5ad.4xlarge 8 30 30

r5ad.8xlarge 8 30 30

r5ad.12xlarge 8 30 30

850
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

r5ad.16xlarge 15 50 50

r5ad.24xlarge 15 50 50

r5b.large 3 10 10

r5b.xlarge 4 15 15

r5b.2xlarge 4 15 15

r5b.4xlarge 8 30 30

r5b.8xlarge 8 30 30

r5b.12xlarge 8 30 30

r5b.16xlarge 15 50 50

r5b.24xlarge 15 50 50

r5b.metal 15 50 50

r5d.large 3 10 10

r5d.xlarge 4 15 15

r5d.2xlarge 4 15 15

r5d.4xlarge 8 30 30

r5d.8xlarge 8 30 30

r5d.12xlarge 8 30 30

r5d.16xlarge 15 50 50

r5d.24xlarge 15 50 50

r5d.metal 15 50 50

r5dn.large 3 10 10

r5dn.xlarge 4 15 15

r5dn.2xlarge 4 15 15

r5dn.4xlarge 8 30 30

r5dn.8xlarge 8 30 30

r5dn.12xlarge 8 30 30

r5dn.16xlarge 15 50 50

r5dn.24xlarge 15 50 50

r5n.large 3 10 10

r5n.xlarge 4 15 15

851
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

r5n.2xlarge 4 15 15

r5n.4xlarge 8 30 30

r5n.8xlarge 8 30 30

r5n.12xlarge 8 30 30

r5n.16xlarge 15 50 50

r5n.24xlarge 15 50 50

r6g.medium 2 4 4

r6g.large 3 10 10

r6g.xlarge 4 15 15

r6g.2xlarge 4 15 15

r6g.4xlarge 8 30 30

r6g.8xlarge 8 30 30

r6g.12xlarge 8 30 30

r6g.16xlarge 15 50 50

r6g.metal 15 50 50

r6gd.medium 2 4 4

r6gd.large 3 10 10

r6gd.xlarge 4 15 15

r6gd.2xlarge 4 15 15

r6gd.4xlarge 8 30 30

r6gd.8xlarge 8 30 30

r6gd.12xlarge 8 30 30

r6gd.16xlarge 15 50 50

r6gd.metal 15 50 50

t1.micro 2 2 IPv6 not supported

t2.nano 2 2 2

t2.micro 2 2 2

t2.small 3 4 4

t2.medium 3 6 6

t2.large 3 12 12

852
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IP addresses per network interface per instance type

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

t2.xlarge 3 15 15

t2.2xlarge 3 15 15

t3.nano 2 2 2

t3.micro 2 2 2

t3.small 3 4 4

t3.medium 3 6 6

t3.large 3 12 12

t3.xlarge 4 15 15

t3.2xlarge 4 15 15

t3a.nano 2 2 2

t3a.micro 2 2 2

t3a.small 2 4 4

t3a.medium 3 6 6

t3a.large 3 12 12

t3a.xlarge 4 15 15

t3a.2xlarge 4 15 15

t4g.nano 2 2 2

t4g.micro 2 2 2

t4g.small 2 4 4

t4g.medium 3 6 6

t4g.large 3 12 12

t4g.xlarge 4 15 15

t4g.2xlarge 4 15 15

u-6tb1.metal 5 30 30

u-9tb1.metal 5 30 30

u-12tb1.metal 5 30 30

u-18tb1.metal 15 50 50

u-24tb1.metal 15 50 50

x1.16xlarge 8 30 30

x1.32xlarge 8 30 30

853
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with network interfaces

Instance type Maximum network Private IPv4 addresses IPv6 addresses per
interfaces per interface interface

x1e.xlarge 3 10 10

x1e.2xlarge 4 15 15

x1e.4xlarge 4 15 15

x1e.8xlarge 4 15 15

x1e.16xlarge 8 30 30

x1e.32xlarge 8 30 30

z1d.large 3 10 10

z1d.xlarge 4 15 15

z1d.2xlarge 4 15 15

z1d.3xlarge 8 30 30

z1d.6xlarge 8 30 30

z1d.12xlarge 15 50 50

z1d.metal 15 50 50

Work with network interfaces

You can work with network interfaces using the Amazon EC2 console or the command line.

Contents
• Create a network interface (p. 854)
• View details about a network interface (p. 855)
• Attach a network interface to an instance (p. 855)
• Detach a network interface from an instance (p. 856)
• Manage IP addresses (p. 857)
• Modify network interface attributes (p. 858)
• Add or edit tags (p. 859)
• Delete a network interface (p. 860)

Create a network interface

You can create a network interface in a subnet. You can't move the network interface to another subnet
after it's created, and you can only attach the network interface to instances in the same Availability
Zone.

To create a network interface using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Choose Create Network Interface.

854
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with network interfaces

4. For Description, enter a descriptive name.

5. For Subnet, select the subnet.
6. For Private IP (or IPv4 Private IP), enter the primary private IPv4 address. If you don't specify an
IPv4 address, we select an available private IPv4 address from within the selected subnet.
7. (IPv6 only) If you selected a subnet that has an associated IPv6 CIDR block, you can optionally
specify an IPv6 address in the IPv6 IP field.
8. To create an Elastic Fabric Adapter, select Elastic Fabric Adapter.
9. For Security groups, select one or more security groups.
10. (Optional) Choose Add Tag and enter a tag key and a tag value.
11. Choose Yes, Create.

To create a network interface using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• create-network-interface (AWS CLI)

• New-EC2NetworkInterface (AWS Tools for Windows PowerShell)

View details about a network interface

You can view all the network interfaces in your account.

To describe a network interface using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface.
4. To view the details, choose Details.

To describe a network interface using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• describe-network-interfaces (AWS CLI)

• Get-EC2NetworkInterface (AWS Tools for Windows PowerShell)

To describe a network interface attribute using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• describe-network-interface-attribute (AWS CLI)

• Get-EC2NetworkInterfaceAttribute (AWS Tools for Windows PowerShell)

Attach a network interface to an instance

You can attach a network interface to any of your stopped or running instances, using either
the Instances or Network Interfaces pages of the Amazon EC2 console. Alternatively, you can

855
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with network interfaces

specify an existing network interface or attach an additional network interface when you launch an
instance (p. 526).

If the public IPv4 address on your instance is released, it does not receive a new one if there is more than
one network interface attached to the instance. For more information about the behavior of public IPv4
addresses, see Public IPv4 addresses and external DNS hostnames (p. 810).

To attach a network interface to an instance using the Instances page

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance.
4. Choose Actions, Networking, Attach network interface.
5. Select a network interface. If the instance supports multiple network cards, you can choose a
network card.
6. Choose Attach.

To attach a network interface to an instance using the Network Interfaces page

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface and choose Attach.
4. Select an instance. If the instance supports multiple network cards, you can choose a network card.
5. Choose Attach.

To attach a network interface to an instance using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• attach-network-interface (AWS CLI)

• Add-EC2NetworkInterface (AWS Tools for Windows PowerShell)

Detach a network interface from an instance

You can detach a secondary network interface that is attached to an EC2 instance at any time, using
either the Instances or Network Interfaces page of the Amazon EC2 console.

To detach a network interface from an instance using the Instances page

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance.
4. Choose Actions, Networking, Detach network interface.
5. Select the network interface and choose Detach.

You can't use the Amazon EC2 console to detach a network interface that is attached to a resource from
another service, such as an Elastic Load Balancing load balancer, a Lambda function, a WorkSpace, or a
NAT gateway. The network interfaces for those resources are deleted when the resource is deleted.

856
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with network interfaces

To detach a network interface from an instance using the Network Interfaces page

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface and check the description to verify that the network interface is
attached to an instance, not another type of resource. If the resource is an EC2 instance, choose
Detach.

If the network interface is the primary network interface for the instance, the Detach button is
disabled.
4. When prompted for confirmation, choose Yes, Detach.
5. If the network interface fails to detach from the instance, choose Force detachment and then try
again. We recommend that you choose this option only as a last resort. Forcing a detachment can
prevent you from attaching a different network interface on the same index until you restart the
instance. It can also prevent the instance metadata from reflecting that the network interface was
detached until you restart the instance.

To detach a network interface using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• detach-network-interface (AWS CLI)

• Dismount-EC2NetworkInterface (AWS Tools for Windows PowerShell)

Manage IP addresses
You can manage the following IP addresses for your network interfaces:

• Elastic IP addresses (one per private IPv4 address)

• IPv4 addresses
• IPv6 addresses

To manage the Elastic IP addresses of a network interface using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface.
4. To associate an Elastic IP address, do the following:

a. Choose Actions, Associate Address.

b. For Address, select the Elastic IP address.
c. For Associate to private IP address, select the private IPv4 address to associate with the Elastic
IP address.
d. Choose Allow reassociation to allow the Elastic IP address to be associated with the specified
network interface if it's currently associated with another instance or network interface, and
then choose Associate Address.
5. To disassociate an Elastic IP address, do the following:

a. Choose Actions, Disassociate Address.

b. In the Disassociate IP Address dialog box, choose Yes, Disassociate.

857
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with network interfaces

To manage the IPv4 and IPv6 addresses of a network interface using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface.
4. Choose Actions, Manage IP Addresses.
5. For IPv4 Addresses, modify the IP addresses as needed. To assign an IPv4 address, choose Assign
new IP and then specify an IPv4 address from the subnet range or let AWS choose one for you. To
unassign an IPv4 address, choose Unassign next to the address.
6. For IPv6 Addresses, modify the IP addresses as needed. To assign an IPv6 address, choose Assign
new IP and then specify an IPv6 address from the subnet range or let AWS choose one for you. To
unassign an IPv6 address, choose Unassign next to the address.
7. Choose Yes, Update.

To manage the IP addresses of a network interface using the AWS CLI

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• assign-ipv6-addresses
• associate-address
• disassociate-address
• unassign-ipv6-addresses

To manage the IP addresses of a network interface using the Tools for Windows PowerShell

You can use one of the following commands.

• Register-EC2Address
• Register-EC2Ipv6AddressList
• Unregister-EC2Address
• Unregister-EC2Ipv6AddressList

Modify network interface attributes

You can change the following network interface attributes:

• Description (p. 858)

• Security groups (p. 859)
• Delete on termination (p. 859)
• Source/destination check (p. 859)

To change the description of a network interface using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface and choose Actions, Change Description.
4. For Change Description, enter a description for the network interface, and then choose Save.

858
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with network interfaces

To change the security groups of a network interface using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface and choose Actions, Change Security Groups.
4. For Change Security Groups, select the security groups to use, and then choose Save.

The security group and network interface must be created for the same VPC. To change the security
group for interfaces owned by other services, such as Elastic Load Balancing, do so through that
service.

To change the termination behavior of a network interface using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface and choose Actions, Change Termination Behavior.
4. In the Change Termination Behavior dialog box, select the Delete on termination check box if you
want the network interface to be deleted when you terminate an instance.

To change source/destination checking for a network interface using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface and choose Actions, Change Source/Dest Check.
4. In the dialog box, choose Enabled (if enabling) or Disabled (if disabling), and Save.

To modify network interface attributes using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• modify-network-interface-attribute (AWS CLI)

• Edit-EC2NetworkInterfaceAttribute (AWS Tools for Windows PowerShell)

Add or edit tags

Tags are metadata that you can add to a network interface. Tags are private and are only visible to your
account. Each tag consists of a key and an optional value. For more information about tags, see Tag your
Amazon EC2 resources (p. 1311).

To add or edit tags for a network interface using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface.
4. In the details pane, choose Tags, Add/Edit Tags.
5. In the Add/Edit Tags dialog box, choose Create Tag for each tag to create, and enter a key and
optional value. When you're done, choose Save.

859
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scenarios for network interfaces

To add or edit tags for a network interface using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• create-tags (AWS CLI)

• New-EC2Tag (AWS Tools for Windows PowerShell)

Delete a network interface

To delete an instance, you must first detach the network interface. Deleting a network interface releases
all attributes associated with the interface and releases any private IP addresses or Elastic IP addresses to
be used by another instance.

To delete a network interface using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select a network interface and choose Delete.
4. In the Delete Network Interface dialog box, choose Yes, Delete.

To delete a network interface using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• delete-network-interface (AWS CLI)

• Remove-EC2NetworkInterface (AWS Tools for Windows PowerShell)

Scenarios for network interfaces

Attaching multiple network interfaces to an instance is useful when you want to:

• Create a management network.

• Use network and security appliances in your VPC.
• Create dual-homed instances with workloads/roles on distinct subnets.
• Create a low-budget, high-availability solution.

Create a management network

You can create a management network using network interfaces. In this scenario, as illustrated in the
following image:

• The primary network interface (eth0) on the instance handles public traffic.
• The secondary network interface (eth1) handles backend management traffic, and is connected to a
separate subnet in your VPC that has more restrictive access controls.

The public interface, which may or may not be behind a load balancer, has an associated security
group that allows access to the server from the internet (for example, allow TCP port 80 and 443 from
0.0.0.0/0, or from the load balancer).

860
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Scenarios for network interfaces

The private facing interface has an associated security group allowing SSH access only from an allowed
range of IP addresses, either within the VPC, or from the internet, a private subnet within the VPC, or a
virtual private gateway.

To ensure failover capabilities, consider using a secondary private IPv4 for incoming traffic on a network
interface. In the event of an instance failure, you can move the interface and/or secondary private IPv4
address to a standby instance.

Use network and security appliances in your VPC

Some network and security appliances, such as load balancers, network address translation (NAT)
servers, and proxy servers prefer to be configured with multiple network interfaces. You can create and
attach secondary network interfaces to instances in a VPC that are running these types of applications
and configure the additional interfaces with their own public and private IP addresses, security groups,
and source/destination checking.

Creating dual-homed instances with workloads/roles on distinct

subnets
You can place a network interface on each of your web servers that connects to a mid-tier network where
an application server resides. The application server can also be dual-homed to a backend network
(subnet) where the database server resides. Instead of routing network packets through the dual-
homed instances, each dual-homed instance receives and processes requests on the front end, initiates a
connection to the backend, and then sends requests to the servers on the backend network.

861
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Best practices for configuring network interfaces

Create a low budget high availability solution

If one of your instances serving a particular function fails, its network interface can be attached to a
replacement or hot standby instance pre-configured for the same role in order to rapidly recover the
service. For example, you can use a network interface as your primary or secondary network interface to
a critical service such as a database instance or a NAT instance. If the instance fails, you (or more likely,
the code running on your behalf) can attach the network interface to a hot standby instance. Because
the interface maintains its private IP addresses, Elastic IP addresses, and MAC address, network traffic
begins flowing to the standby instance as soon as you attach the network interface to the replacement
instance. Users experience a brief loss of connectivity between the time the instance fails and the time
that the network interface is attached to the standby instance, but no changes to the VPC route table or
your DNS server are required.

Best practices for configuring network interfaces

• You can attach a network interface to an instance when it's running (hot attach), when it's stopped
(warm attach), or when the instance is being launched (cold attach).
• You can detach secondary network interfaces when the instance is running or stopped. However, you
can't detach the primary network interface.
• You can move a network interface from one instance to another, if the instances are in the same
Availability Zone and VPC but in different subnets.
• When launching an instance using the CLI, API, or an SDK, you can specify the primary network
interface and additional network interfaces.
• Launching an Amazon Linux or Windows Server instance with multiple network interfaces
automatically configures interfaces, private IPv4 addresses, and route tables on the operating system
of the instance.
• A warm or hot attach of an additional network interface may require you to manually bring up the
second interface, configure the private IPv4 address, and modify the route table accordingly. Instances
running Amazon Linux or Windows Server automatically recognize the warm or hot attach and
configure themselves.
• Attaching another network interface to an instance (for example, a NIC teaming configuration)
cannot be used as a method to increase or double the network bandwidth to or from the dual-homed
instance.
• If you attach two or more network interfaces from the same subnet to an instance, you might
encounter networking issues such as asymmetric routing. If possible, use a secondary private IPv4
address on the primary network interface instead.

Configure your network interface using ec2-net-utils

Amazon Linux AMIs may contain additional scripts installed by AWS, known as ec2-net-utils. These
scripts optionally automate the configuration of your network interfaces. These scripts are available for
Amazon Linux only.

Use the following command to install the package on Amazon Linux if it's not already installed, or
update it if it's installed and additional updates are available:

$ yum install ec2-net-utils

The following components are part of ec2-net-utils:

udev rules (/etc/udev/rules.d)

Identifies network interfaces when they are attached, detached, or reattached to a running instance,
and ensures that the hotplug script runs (53-ec2-network-interfaces.rules). Maps the

862
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Requester-managed network interfaces

MAC address to a device name (75-persistent-net-generator.rules, which generates 70-

persistent-net.rules).
hotplug script

Generates an interface configuration file suitable for use with DHCP (/etc/sysconfig/network-
scripts/ifcfg-ethN). Also generates a route configuration file (/etc/sysconfig/network-
scripts/route-ethN).
DHCP script

Whenever the network interface receives a new DHCP lease, this script queries the instance
metadata for Elastic IP addresses. For each Elastic IP address, it adds a rule to the routing policy
database to ensure that outbound traffic from that address uses the correct network interface. It
also adds each private IP address to the network interface as a secondary address.
ec2ifup ethN

Extends the functionality of the standard ifup. After this script rewrites the configuration files
ifcfg-ethN and route-ethN, it runs ifup.
ec2ifdown ethN

Extends the functionality of the standard ifdown. After this script removes any rules for the network
interface from the routing policy database, it runs ifdown.
ec2ifscan

Checks for network interfaces that have not been configured and configures them.

This script isn't available in the initial release of ec2-net-utils.

To list any configuration files that were generated by ec2-net-utils, use the following command:

$ ls -l /etc/sysconfig/network-scripts/*-eth?

To disable the automation on a per-instance basis, you can add EC2SYNC=no to the corresponding
ifcfg-ethN file. For example, use the following command to disable the automation for the eth1
interface:

$ sed -i -e 's/^EC2SYNC=yes/EC2SYNC=no/' /etc/sysconfig/network-scripts/ifcfg-eth1

To disable the automation completely, you can remove the package using the following command:

$ yum remove ec2-net-utils

Requester-managed network interfaces

A requester-managed network interface is a network interface that an AWS service creates in your VPC.
This network interface can represent an instance for another service, such as an Amazon RDS instance, or
it can enable you to access another service or resource, such as an AWS PrivateLink service, or an Amazon
ECS task.

You cannot modify or detach a requester-managed network interface. If you delete the resource that
the network interface represents, the AWS service detaches and deletes the network interface for
you. To change the security groups for a requester-managed network interface, you might have to
use the console or command line tools for that service. For more information, see the service-specific
documentation.

863
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking

You can tag a requester-managed network interface. For more information, see Add or edit
tags (p. 859).

You can view the requester-managed network interfaces that are in your account.

To view requester-managed network interfaces using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Select the network interface and view the following information on the details pane:

• Attachment owner: If you created the network interface, this field displays your AWS account ID.
Otherwise, it displays an alias or ID for the principal or service that created the network interface.
• Description: Provides information about the purpose of the network interface; for example, "VPC
Endpoint Interface".

To view requester-managed network interfaces using the command line

1. Use the describe-network-interfaces AWS CLI command to describe the network interfaces in your
account.

aws ec2 describe-network-interfaces

2. In the output, the RequesterManaged field displays true if the network interface is managed by
another AWS service.

{
"Status": "in-use",
...
"Description": "VPC Endpoint Interface vpce-089f2123488812123",
"NetworkInterfaceId": "eni-c8fbc27e",
"VpcId": "vpc-1a2b3c4d",
"PrivateIpAddresses": [
{
"PrivateDnsName": "ip-10-0-2-227.ec2.internal",
"Primary": true,
"PrivateIpAddress": "10.0.2.227"
}
],
"RequesterManaged": true,
...
}

Alternatively, use the Get-EC2NetworkInterface Tools for Windows PowerShell command.

Enhanced networking on Linux

Enhanced networking uses single root I/O virtualization (SR-IOV) to provide high-performance
networking capabilities on supported instance types (p. 865). SR-IOV is a method of device
virtualization that provides higher I/O performance and lower CPU utilization when compared to
traditional virtualized network interfaces. Enhanced networking provides higher bandwidth, higher
packet per second (PPS) performance, and consistently lower inter-instance latencies. There is no
additional charge for using enhanced networking.

For information about the supported network speed for each instance type, see Amazon EC2 Instance
Types.

864
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking support

Contents
• Enhanced networking support (p. 865)
• Enable enhanced networking on your instance (p. 865)
• Enable enhanced networking with the Elastic Network Adapter (ENA) on Linux instances (p. 865)
• Enable enhanced networking with the Intel 82599 VF interface on Linux instances (p. 875)
• Operating system optimizations (p. 881)
• Monitor network performance for your instance (p. 884)
• Troubleshoot the Elastic Network Adapter (ENA) (p. 887)

Enhanced networking support

All current generation (p. 204) instance types support enhanced networking, except for T2 instances.

You can enable enhanced networking using one of the following mechanisms:

Elastic Network Adapter (ENA)

The Elastic Network Adapter (ENA) supports network speeds of up to 100 Gbps for supported
instance types.

The current generation instances use ENA for enhanced networking, except for C4, D2, and M4
instances smaller than m4.16xlarge.
Intel 82599 Virtual Function (VF) interface

The Intel 82599 Virtual Function interface supports network speeds of up to 10 Gbps for supported
instance types.

The following instance types use the Intel 82599 VF interface for enhanced networking: C3, C4, D2,
I2, M4 (excluding m4.16xlarge), and R3.

For a summary of the enhanced networking mechanisms by instance type, see Summary of networking
and storage features (p. 210).

Enable enhanced networking on your instance

If your instance type supports the Elastic Network Adapter for enhanced networking, follow the
procedures in Enable enhanced networking with the Elastic Network Adapter (ENA) on Linux
instances (p. 865).

If your instance type supports the Intel 82599 VF interface for enhanced networking, follow
the procedures in Enable enhanced networking with the Intel 82599 VF interface on Linux
instances (p. 875).

Enable enhanced networking with the Elastic

Network Adapter (ENA) on Linux instances
Amazon EC2 provides enhanced networking capabilities through the Elastic Network Adapter (ENA). To
use enhanced networking, you must install the required ENA module and enable ENA support.

Contents
• Requirements (p. 866)

865
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: ENA

• Enhanced networking performance (p. 866)

• Test whether enhanced networking is enabled (p. 866)
• Enable enhanced networking on the Amazon Linux AMI (p. 868)
• Enable enhanced networking on Ubuntu (p. 870)
• Enable enhanced networking on Linux (p. 871)
• Enable enhanced networking on Ubuntu with DKMS (p. 873)
• Driver release notes (p. 875)
• Troubleshoot (p. 875)

Requirements
To prepare for enhanced networking using the ENA, set up your instance as follows:

• Launch the instance using a current generation (p. 204) instance type, other than C4, D2, M4 instances
smaller than m4.16xlarge, or T2.
• Launch the instance using a supported version of the Linux kernel and a supported distribution, so that
ENA enhanced networking is enabled for your instance automatically. For more information, see ENA
Linux Kernel Driver Release Notes.
• Ensure that the instance has internet connectivity.
• Install and configure the AWS CLI or the AWS Tools for Windows PowerShell on any computer you
choose, preferably your local desktop or laptop. For more information, see Access Amazon EC2 (p. 3).
Enhanced networking cannot be managed from the Amazon EC2 console.
• If you have important data on the instance that you want to preserve, you should back that data up
now by creating an AMI from your instance. Updating kernels and kernel modules, as well as enabling
the enaSupport attribute, might render incompatible instances or operating systems unreachable. If
you have a recent backup, your data will still be retained if this happens.

Enhanced networking performance

The following documentation provides a summary of the network performance for the instance types
that support ENA enhanced networking:

• Network Performance for Accelerated Computing Instances (p. 295)

• Network Performance for Compute Optimized Instances (p. 267)
• Network Performance for General Purpose Instances (p. 218)
• Network Performance for Memory Optimized Instances (p. 277)
• Network Performance for Storage Optimized Instances (p. 286)

Test whether enhanced networking is enabled

The following AMIs include the required ENA module and have ENA support enabled:

• Amazon Linux 2
• Amazon Linux AMI 2018.03
• Ubuntu 14.04 (with linux-aws kernel) or later
• Red Hat Enterprise Linux 7.4 or later
• SUSE Linux Enterprise Server 12 SP2 or later
• CentOS 7.4.1708 or later

866
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: ENA

• FreeBSD 11.1 or later

• Debian GNU/Linux 9 or later

To test whether enhanced networking is already enabled, verify that the ena module is installed on your
instance and that the enaSupport attribute is set. If your instance satisfies these two conditions, then
the ethtool -i ethn command should show that the module is in use on the network interface.

Kernel module (ena)

To verify that the ena module is installed, use the modinfo command as shown in the following
example.

[ec2-user ~]$ modinfo ena

filename: /lib/modules/4.14.33-59.37.amzn2.x86_64/kernel/drivers/amazon/net/ena/
ena.ko
version: 1.5.0g
license: GPL
description: Elastic Network Adapter (ENA)
author: Amazon.com, Inc. or its affiliates
srcversion: 692C7C68B8A9001CB3F31D0
alias: pci:v00001D0Fd0000EC21sv*sd*bc*sc*i*
alias: pci:v00001D0Fd0000EC20sv*sd*bc*sc*i*
alias: pci:v00001D0Fd00001EC2sv*sd*bc*sc*i*
alias: pci:v00001D0Fd00000EC2sv*sd*bc*sc*i*
depends:
retpoline: Y
intree: Y
name: ena
...

In the above Amazon Linux case, the ena module is installed.

ubuntu:~$ modinfo ena

ERROR: modinfo: could not find module ena

In the above Ubuntu instance, the module is not installed, so you must first install it. For more
information, see Enable enhanced networking on Ubuntu (p. 870).

Instance attribute (enaSupport)

To check whether an instance has the enhanced networking enaSupport attribute set, use one of the
following commands. If the attribute is set, the response is true.

• describe-instances (AWS CLI)

aws ec2 describe-instances --instance-ids instance_id --query

"Reservations[].Instances[].EnaSupport"

• Get-EC2Instance (Tools for Windows PowerShell)

(Get-EC2Instance -InstanceId instance-id).Instances.EnaSupport

Image attribute (enaSupport)

To check whether an AMI has the enhanced networking enaSupport attribute set, use one of the
following commands. If the attribute is set, the response is true.

867
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: ENA

• describe-images (AWS CLI)

aws ec2 describe-images --image-id ami_id --query "Images[].EnaSupport"

• Get-EC2Image (Tools for Windows PowerShell)

(Get-EC2Image -ImageId ami_id).EnaSupport

Network interface driver

Use the following command to verify that the ena module is being used on a particular interface,
substituting the interface name that you want to check. If you are using a single interface (default), it this
is eth0. If the operating system supports predictable network names (p. 871), this could be a name like
ens5.

In the following example, the ena module is not loaded, because the listed driver is vif.

[ec2-user ~]$ ethtool -i eth0

driver: vif
version:
firmware-version:
bus-info: vif-0
supports-statistics: yes
supports-test: no
supports-eeprom-access: no
supports-register-dump: no
supports-priv-flags: no

In this example, the ena module is loaded and at the minimum recommended version. This instance has
enhanced networking properly configured.

[ec2-user ~]$ ethtool -i eth0

driver: ena
version: 1.5.0g
firmware-version:
expansion-rom-version:
bus-info: 0000:00:05.0
supports-statistics: yes
supports-test: no
supports-eeprom-access: no
supports-register-dump: no
supports-priv-flags: no

Enable enhanced networking on the Amazon Linux AMI

Amazon Linux 2 and the latest versions of the Amazon Linux AMI include the module required for
enhanced networking with ENA installed and have ENA support enabled. Therefore, if you launch an
instance with an HVM version of Amazon Linux on a supported instance type, enhanced networking
is already enabled for your instance. For more information, see Test whether enhanced networking is
enabled (p. 866).

If you launched your instance using an older Amazon Linux AMI and it does not have enhanced
networking enabled already, use the following procedure to enable enhanced networking.

To enable enhanced networking on Amazon Linux AMI

1. Connect to your instance.

868
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: ENA

2. From the instance, run the following command to update your instance with the newest kernel and
kernel modules, including ena:

[ec2-user ~]$ sudo yum update

3. From your local computer, reboot your instance using the Amazon EC2 console or one of the
following commands: reboot-instances (AWS CLI), Restart-EC2Instance (AWS Tools for Windows
PowerShell).
4. Connect to your instance again and verify that the ena module is installed and at the minimum
recommended version using the modinfo ena command from Test whether enhanced networking is
enabled (p. 866).
5. [EBS-backed instance] From your local computer, stop the instance using the Amazon EC2 console
or one of the following commands: stop-instances (AWS CLI), Stop-EC2Instance (AWS Tools for
Windows PowerShell). If your instance is managed by AWS OpsWorks, you should stop the instance
in the AWS OpsWorks console so that the instance state remains in sync.

[Instance store-backed instance] You can't stop the instance to modify the attribute. Instead,
proceed to this procedure: To enable enhanced networking on Amazon Linux AMI (instance store-
backed instances) (p. 869).
6. From your local computer, enable the enhanced networking attribute using one of the following
commands:

• modify-instance-attribute (AWS CLI)

aws ec2 modify-instance-attribute --instance-id instance_id --ena-support

• Edit-EC2InstanceAttribute (Tools for Windows PowerShell)

Edit-EC2InstanceAttribute -InstanceId instance-id -EnaSupport $true

7. (Optional) Create an AMI from the instance, as described in Create an Amazon EBS-backed Linux
AMI (p. 125). The AMI inherits the enhanced networking enaSupport attribute from the instance.
Therefore, you can use this AMI to launch another instance with enhanced networking enabled by
default.
8. From your local computer, start the instance using the Amazon EC2 console or one of the following
commands: start-instances (AWS CLI), Start-EC2Instance (AWS Tools for Windows PowerShell). If
your instance is managed by AWS OpsWorks, you should start the instance in the AWS OpsWorks
console so that the instance state remains in sync.
9. Connect to your instance and verify that the ena module is installed and loaded on your network
interface using the ethtool -i ethn command from Test whether enhanced networking is
enabled (p. 866).

If you are unable to connect to your instance after enabling enhanced networking, see Troubleshoot
the Elastic Network Adapter (ENA) (p. 887).

To enable enhanced networking on Amazon Linux AMI (instance store-backed instances)

Follow the previous procedure until the step where you stop the instance. Create a new AMI as described
in Create an instance store-backed Linux AMI (p. 130), making sure to enable the enhanced networking
attribute when you register the AMI.

• register-image (AWS CLI)

aws ec2 register-image --ena-support ...

• Register-EC2Image (AWS Tools for Windows PowerShell)

869
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: ENA

Register-EC2Image -EnaSupport $true ...

Enable enhanced networking on Ubuntu

The latest Ubuntu HVM AMIs include the module required for enhanced networking with ENA installed
and have ENA support enabled. Therefore, if you launch an instance with the latest Ubuntu HVM AMI
on a supported instance type, enhanced networking is already enabled for your instance. For more
information, see Test whether enhanced networking is enabled (p. 866).

If you launched your instance using an older AMI and it does not have enhanced networking enabled
already, you can install the linux-aws kernel package to get the latest enhanced networking drivers
and update the required attribute.

To install the linux-aws kernel package (Ubuntu 16.04 or later)

Ubuntu 16.04 and 18.04 ship with the Ubuntu custom kernel (linux-aws kernel package). To use a
different kernel, contact AWS Support.

To install the linux-aws kernel package (Ubuntu Trusty 14.04)

1. Connect to your instance.

2. Update the package cache and packages.

ubuntu:~$ sudo apt-get update && sudo apt-get upgrade -y linux-aws

Important
If during the update process you are prompted to install grub, use /dev/xvda to install
grub onto, and then choose to keep the current version of /boot/grub/menu.lst.
3. [EBS-backed instance] From your local computer, stop the instance using the Amazon EC2 console
or one of the following commands: stop-instances (AWS CLI), Stop-EC2Instance (AWS Tools for
Windows PowerShell). If your instance is managed by AWS OpsWorks, you should stop the instance
in the AWS OpsWorks console so that the instance state remains in sync.

[Instance store-backed instance] You can't stop the instance to modify the attribute. Instead,
proceed to this procedure: To enable enhanced networking on Ubuntu (instance store-backed
instances) (p. 871).
4. From your local computer, enable the enhanced networking attribute using one of the following
commands:

• modify-instance-attribute (AWS CLI)

aws ec2 modify-instance-attribute --instance-id instance_id --ena-support

• Edit-EC2InstanceAttribute (Tools for Windows PowerShell)

Edit-EC2InstanceAttribute -InstanceId instance-id -EnaSupport $true

5. (Optional) Create an AMI from the instance, as described in Create an Amazon EBS-backed Linux
AMI (p. 125). The AMI inherits the enhanced networking enaSupport attribute from the instance.
Therefore, you can use this AMI to launch another instance with enhanced networking enabled by
default.
6. From your local computer, start the instance using the Amazon EC2 console or one of the following
commands: start-instances (AWS CLI), Start-EC2Instance (AWS Tools for Windows PowerShell). If

870
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: ENA

your instance is managed by AWS OpsWorks, you should start the instance in the AWS OpsWorks
console so that the instance state remains in sync.

To enable enhanced networking on Ubuntu (instance store-backed instances)

Follow the previous procedure until the step where you stop the instance. Create a new AMI as described
in Create an instance store-backed Linux AMI (p. 130), making sure to enable the enhanced networking
attribute when you register the AMI.

• register-image (AWS CLI)

aws ec2 register-image --ena-support ...

• Register-EC2Image (AWS Tools for Windows PowerShell)

Register-EC2Image -EnaSupport $true ...

Enable enhanced networking on Linux

The latest AMIs for Red Hat Enterprise Linux, SUSE Linux Enterprise Server, and CentOS include the
module required for enhanced networking with ENA and have ENA support enabled. Therefore, if
you launch an instance with the latest AMI on a supported instance type, enhanced networking is
already enabled for your instance. For more information, see Test whether enhanced networking is
enabled (p. 866).

The following procedure provides the general steps for enabling enhanced networking on a Linux
distribution other than Amazon Linux AMI or Ubuntu. For more information, such as detailed syntax
for commands, file locations, or package and tool support, see the documentation for your Linux
distribution.

To enable enhanced networking on Linux

1. Connect to your instance.

2. Clone the source code for the ena module on your instance from GitHub at https://github.com/
amzn/amzn-drivers. (SUSE Linux Enterprise Server 12 SP2 and later include ENA 2.02 by default, so
you are not required to download and compile the ENA driver. For SUSE Linux Enterprise Server 12
SP2 and later, you should file a request to add the driver version you want to the stock kernel).

git clone https://github.com/amzn/amzn-drivers

3. Compile and install the ena module on your instance. These steps depend on the Linux distribution.
For more information about compiling the module on Red Hat Enterprise Linux, see the AWS
Knowledge Center article.
4. Run the sudo depmod command to update module dependencies.
5. Update initramfs on your instance to ensure that the new module loads at boot time. For
example, if your distribution supports dracut, you can use the following command.

dracut -f -v

6. Determine if your system uses predictable network interface names by default. Systems that use
systemd or udev versions 197 or greater can rename Ethernet devices and they do not guarantee
that a single network interface will be named eth0. This behavior can cause problems connecting to
your instance. For more information and to see other configuration options, see Predictable Network
Interface Names on the freedesktop.org website.

871
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: ENA

a. You can check the systemd or udev versions on RPM-based systems with the following
command.

rpm -qa | grep -e '^systemd-[0-9]\+\|^udev-[0-9]\+'

systemd-208-11.el7_0.2.x86_64

In the above Red Hat Enterprise Linux 7 example, the systemd version is 208, so predictable
network interface names must be disabled.
b. Disable predictable network interface names by adding the net.ifnames=0 option to the
GRUB_CMDLINE_LINUX line in /etc/default/grub.

sudo sed -i '/^GRUB\_CMDLINE\_LINUX/s/\"$/\ net\.ifnames\=0\"/' /etc/default/grub

c. Rebuild the grub configuration file.

sudo grub2-mkconfig -o /boot/grub2/grub.cfg

7. [EBS-backed instance] From your local computer, stop the instance using the Amazon EC2 console
or one of the following commands: stop-instances (AWS CLI), Stop-EC2Instance (AWS Tools for
Windows PowerShell). If your instance is managed by AWS OpsWorks, you should stop the instance
in the AWS OpsWorks console so that the instance state remains in sync.

[Instance store-backed instance] You can't stop the instance to modify the attribute. Instead,
proceed to this procedure: To enable enhanced networking on Linux (instance store–backed
instances) (p. 873).
8. From your local computer, enable the enhanced networking enaSupport attribute using one of the
following commands:

• modify-instance-attribute (AWS CLI)

aws ec2 modify-instance-attribute --instance-id instance_id --ena-support

• Edit-EC2InstanceAttribute (Tools for Windows PowerShell)

Edit-EC2InstanceAttribute -InstanceId instance-id -EnaSupport $true

9. (Optional) Create an AMI from the instance, as described in Create an Amazon EBS-backed Linux
AMI (p. 125) . The AMI inherits the enhanced networking enaSupport attribute from the instance.
Therefore, you can use this AMI to launch another instance with enhanced networking enabled by
default.
Important
If your instance operating system contains an /etc/udev/rules.d/70-persistent-
net.rules file, you must delete it before creating the AMI. This file contains the MAC
address for the Ethernet adapter of the original instance. If another instance boots with
this file, the operating system will be unable to find the device and eth0 might fail, causing
boot issues. This file is regenerated at the next boot cycle, and any instances launched from
the AMI create their own version of the file.
10. From your local computer, start the instance using the Amazon EC2 console or one of the following
commands: start-instances (AWS CLI), Start-EC2Instance (AWS Tools for Windows PowerShell). If
your instance is managed by AWS OpsWorks, you should start the instance in the AWS OpsWorks
console so that the instance state remains in sync.
11. (Optional) Connect to your instance and verify that the module is installed.

872
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: ENA

If you are unable to connect to your instance after enabling enhanced networking, see Troubleshoot
the Elastic Network Adapter (ENA) (p. 887).

To enable enhanced networking on Linux (instance store–backed instances)

Follow the previous procedure until the step where you stop the instance. Create a new AMI as described
in Create an instance store-backed Linux AMI (p. 130), making sure to enable the enhanced networking
attribute when you register the AMI.

• register-image (AWS CLI)

aws ec2 register-image --ena-support ...

• Register-EC2Image (AWS Tools for Windows PowerShell)

Register-EC2Image -EnaSupport ...

Enable enhanced networking on Ubuntu with DKMS

This method is for testing and feedback purposes only. It is not intended for use with production
deployments. For production deployments, see Enable enhanced networking on Ubuntu (p. 870).
Important
Using DKMS voids the support agreement for your subscription. It should not be used for
production deployments.

To enable enhanced networking with ENA on Ubuntu (EBS-backed instances)

1. Follow steps 1 and 2 in Enable enhanced networking on Ubuntu (p. 870).

2. Install the build-essential packages to compile the kernel module and the dkms package so
that your ena module is rebuilt every time your kernel is updated.

ubuntu:~$ sudo apt-get install -y build-essential dkms

3. Clone the source for the ena module on your instance from GitHub at https://github.com/amzn/
amzn-drivers.

ubuntu:~$ git clone https://github.com/amzn/amzn-drivers

4. Move the amzn-drivers package to the /usr/src/ directory so DKMS can find it and build it for
each kernel update. Append the version number (you can find the current version number in the
release notes) of the source code to the directory name. For example, version 1.0.0 is shown in the
following example.

ubuntu:~$ sudo mv amzn-drivers /usr/src/amzn-drivers-1.0.0

5. Create the DKMS configuration file with the following values, substituting your version of ena.

Create the file.

ubuntu:~$ sudo touch /usr/src/amzn-drivers-1.0.0/dkms.conf

Edit the file and add the following values.

873
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: ENA

ubuntu:~$ sudo vim /usr/src/amzn-drivers-1.0.0/dkms.conf

PACKAGE_NAME="ena"
PACKAGE_VERSION="1.0.0"
CLEAN="make -C kernel/linux/ena clean"
MAKE="make -C kernel/linux/ena/ BUILD_KERNEL=${kernelver}"
BUILT_MODULE_NAME[0]="ena"
BUILT_MODULE_LOCATION="kernel/linux/ena"
DEST_MODULE_LOCATION[0]="/updates"
DEST_MODULE_NAME[0]="ena"
AUTOINSTALL="yes"

6. Add, build, and install the ena module on your instance using DKMS.

Add the module to DKMS.

ubuntu:~$ sudo dkms add -m amzn-drivers -v 1.0.0

Build the module using the dkms command.

ubuntu:~$ sudo dkms build -m amzn-drivers -v 1.0.0

Install the module using dkms.

ubuntu:~$ sudo dkms install -m amzn-drivers -v 1.0.0

7. Rebuild initramfs so the correct module is loaded at boot time.

ubuntu:~$ sudo update-initramfs -u -k all

8. Verify that the ena module is installed using the modinfo ena command from Test whether
enhanced networking is enabled (p. 866).

ubuntu:~$ modinfo ena

filename: /lib/modules/3.13.0-74-generic/updates/dkms/ena.ko
version: 1.0.0
license: GPL
description: Elastic Network Adapter (ENA)
author: Amazon.com, Inc. or its affiliates
srcversion: 9693C876C54CA64AE48F0CA
alias: pci:v00001D0Fd0000EC21sv*sd*bc*sc*i*
alias: pci:v00001D0Fd0000EC20sv*sd*bc*sc*i*
alias: pci:v00001D0Fd00001EC2sv*sd*bc*sc*i*
alias: pci:v00001D0Fd00000EC2sv*sd*bc*sc*i*
depends:
vermagic: 3.13.0-74-generic SMP mod_unload modversions
parm: debug:Debug level (0=none,...,16=all) (int)
parm: push_mode:Descriptor / header push mode
(0=automatic,1=disable,3=enable)
0 - Automatically choose according to device capability (default)
1 - Don't push anything to device memory
3 - Push descriptors and header buffer to device memory (int)
parm: enable_wd:Enable keepalive watchdog (0=disable,1=enable,default=1)
(int)
parm: enable_missing_tx_detection:Enable missing Tx completions. (default=1)
(int)
parm: numa_node_override_array:Numa node override map
(array of int)
parm: numa_node_override:Enable/Disable numa node override (0=disable)
(int)

874
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: Intel 82599 VF

9. Continue with Step 3 in Enable enhanced networking on Ubuntu (p. 870).

Driver release notes

For information about the versions of the Linux ENA driver, see the ENA Linux kernel driver release notes.

Troubleshoot
For troubleshooting information, see Troubleshoot the Elastic Network Adapter (ENA) (p. 887).

Enable enhanced networking with the Intel 82599 VF

interface on Linux instances
Amazon EC2 provides enhanced networking capabilities through the Intel 82599 VF interface, which uses
the Intel ixgbevf driver.

Contents
• Requirements (p. 875)
• Test whether enhanced networking is enabled (p. 876)
• Enable enhanced networking on Amazon Linux (p. 877)
• Enable enhanced networking on Ubuntu (p. 878)
• Enable enhanced networking on other Linux distributions (p. 879)
• Troubleshoot connectivity issues (p. 880)

Requirements
To prepare for enhanced networking using the Intel 82599 VF interface, set up your instance as follows:

• Select from the following supported instance types: C3, C4, D2, I2, M4 (excluding m4.16xlarge), and
R3.
• Launch the instance from an HVM AMI using Linux kernel version of 2.6.32 or later. The latest Amazon
Linux HVM AMIs have the modules required for enhanced networking installed and have the required
attributes set. Therefore, if you launch an Amazon EBS–backed, enhanced networking–supported
instance using a current Amazon Linux HVM AMI, enhanced networking is already enabled for your
instance.
Warning
Enhanced networking is supported only for HVM instances. Enabling enhanced networking
with a PV instance can make it unreachable. Setting this attribute without the proper module
or module version can also make your instance unreachable.
• Ensure that the instance has internet connectivity.
• Install and configure the AWS CLI or the AWS Tools for Windows PowerShell on any computer you
choose, preferably your local desktop or laptop. For more information, see Access Amazon EC2 (p. 3).
Enhanced networking cannot be managed from the Amazon EC2 console.
• If you have important data on the instance that you want to preserve, you should back that data
up now by creating an AMI from your instance. Updating kernels and kernel modules, as well as
enabling the sriovNetSupport attribute, might render incompatible instances or operating systems
unreachable. If you have a recent backup, your data will still be retained if this happens.

875
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: Intel 82599 VF

Test whether enhanced networking is enabled

Enhanced networking with the Intel 82599 VF interface is enabled if the ixgbevf module is installed on
your instance and the sriovNetSupport attribute is set.

Instance attribute (sriovNetSupport)

To check whether an instance has the enhanced networking sriovNetSupport attribute set, use one of
the following commands:

• describe-instance-attribute (AWS CLI)

aws ec2 describe-instance-attribute --instance-id instance_id --attribute sriovNetSupport

• Get-EC2InstanceAttribute (AWS Tools for Windows PowerShell)

Get-EC2InstanceAttribute -InstanceId instance-id -Attribute sriovNetSupport

If the attribute isn't set, SriovNetSupport is empty. If the attribute is set, the value is simple, as shown
in the following example output.

"SriovNetSupport": {
"Value": "simple"
},

Image attribute (sriovNetSupport)

To check whether an AMI already has the enhanced networking sriovNetSupport attribute set, use
one of the following commands:

• describe-images (AWS CLI)

aws ec2 describe-images --image-id ami_id --query "Images[].SriovNetSupport"

• Get-EC2Image (AWS Tools for Windows PowerShell)

(Get-EC2Image -ImageId ami-id).SriovNetSupport

If the attribute isn't set, SriovNetSupport is empty. If the attribute is set, the value is simple.

Network interface driver

Use the following command to verify that the module is being used on a particular interface,
substituting the interface name that you want to check. If you are using a single interface (default), this
is eth0. If the operating system supports predictable network names (p. 879), this could be a name like
ens5.

In the following example, the ixgbevf module is not loaded, because the listed driver is vif.

[ec2-user ~]$ ethtool -i eth0

driver: vif
version:
firmware-version:
bus-info: vif-0
supports-statistics: yes

876
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: Intel 82599 VF

supports-test: no
supports-eeprom-access: no
supports-register-dump: no
supports-priv-flags: no

In this example, the ixgbevf module is loaded. This instance has enhanced networking properly
configured.

[ec2-user ~]$ ethtool -i eth0

driver: ixgbevf
version: 4.0.3
firmware-version: N/A
bus-info: 0000:00:03.0
supports-statistics: yes
supports-test: yes
supports-eeprom-access: no
supports-register-dump: yes
supports-priv-flags: no

Enable enhanced networking on Amazon Linux

The latest Amazon Linux HVM AMIs have the ixgbevf module required for enhanced networking
installed and have the required sriovNetSupport attribute set. Therefore, if you launch an instance
type using a current Amazon Linux HVM AMI, enhanced networking is already enabled for your instance.
For more information, see Test whether enhanced networking is enabled (p. 876).

If you launched your instance using an older Amazon Linux AMI and it does not have enhanced
networking enabled already, use the following procedure to enable enhanced networking.
Warning
There is no way to disable the enhanced networking attribute after you've enabled it.

To enable enhanced networking

1. Connect to your instance.

2. From the instance, run the following command to update your instance with the newest kernel and
kernel modules, including ixgbevf:

[ec2-user ~]$ sudo yum update

3. From your local computer, reboot your instance using the Amazon EC2 console or one of the
following commands: reboot-instances (AWS CLI), Restart-EC2Instance (AWS Tools for Windows
PowerShell).
4. Connect to your instance again and verify that the ixgbevf module is installed and at the
minimum recommended version using the modinfo ixgbevf command from Test whether enhanced
networking is enabled (p. 876).
5. [EBS-backed instance] From your local computer, stop the instance using the Amazon EC2 console
or one of the following commands: stop-instances (AWS CLI), Stop-EC2Instance (AWS Tools for
Windows PowerShell). If your instance is managed by AWS OpsWorks, you should stop the instance
in the AWS OpsWorks console so that the instance state remains in sync.

[Instance store-backed instance] You can't stop the instance to modify the attribute.
Instead, proceed to this procedure: To enable enhanced networking (instance store-backed
instances) (p. 878).
6. From your local computer, enable the enhanced networking attribute using one of the following
commands:

• modify-instance-attribute (AWS CLI)

877
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: Intel 82599 VF

aws ec2 modify-instance-attribute --instance-id instance_id --sriov-net-support

simple

• Edit-EC2InstanceAttribute (AWS Tools for Windows PowerShell)

Edit-EC2InstanceAttribute -InstanceId instance_id -SriovNetSupport "simple"

7. (Optional) Create an AMI from the instance, as described in Create an Amazon EBS-backed Linux
AMI (p. 125) . The AMI inherits the enhanced networking attribute from the instance. Therefore, you
can use this AMI to launch another instance with enhanced networking enabled by default.
8. From your local computer, start the instance using the Amazon EC2 console or one of the following
commands: start-instances (AWS CLI), Start-EC2Instance (AWS Tools for Windows PowerShell). If
your instance is managed by AWS OpsWorks, you should start the instance in the AWS OpsWorks
console so that the instance state remains in sync.
9. Connect to your instance and verify that the ixgbevf module is installed and loaded on your
network interface using the ethtool -i ethn command from Test whether enhanced networking is
enabled (p. 876).

To enable enhanced networking (instance store-backed instances)

Follow the previous procedure until the step where you stop the instance. Create a new AMI as described
in Create an instance store-backed Linux AMI (p. 130), making sure to enable the enhanced networking
attribute when you register the AMI.

• register-image (AWS CLI)

aws ec2 register-image --sriov-net-support simple ...

• Register-EC2Image (AWS Tools for Windows PowerShell)

Register-EC2Image -SriovNetSupport "simple" ...

Enable enhanced networking on Ubuntu

Before you begin, check if enhanced networking is already enabled (p. 876) on your instance.

The Quick Start Ubuntu HVM AMIs include the necessary drivers for enhanced networking. If you have a
version of ixgbevf earlier than 2.16.4, you can install the linux-aws kernel package to get the latest
enhanced networking drivers.

The following procedure provides the general steps for compiling the ixgbevf module on an Ubuntu
instance.

To install the linux-aws kernel package

1. Connect to your instance.

2. Update the package cache and packages.

ubuntu:~$ sudo apt-get update && sudo apt-get upgrade -y linux-aws

Important
If during the update process, you are prompted to install grub, use /dev/xvda to install
grub, and then choose to keep the current version of /boot/grub/menu.lst.

878
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: Intel 82599 VF

Enable enhanced networking on other Linux distributions

Before you begin, check if enhanced networking is already enabled (p. 876) on your instance. The latest
Quick Start HVM AMIs include the necessary drivers for enhanced networking, therefore you do not need
to perform additional steps.

The following procedure provides the general steps if you need to enable enhanced networking with
the Intel 82599 VF interface on a Linux distribution other than Amazon Linux or Ubuntu. For more
information, such as detailed syntax for commands, file locations, or package and tool support, see the
specific documentation for your Linux distribution.

To enable enhanced networking on Linux

1. Connect to your instance.

2. Download the source for the ixgbevf module on your instance from Sourceforge at https://
sourceforge.net/projects/e1000/files/ixgbevf%20stable/.

Versions of ixgbevf earlier than 2.16.4, including version 2.14.2, do not build properly on some
Linux distributions, including certain versions of Ubuntu.
3. Compile and install the ixgbevf module on your instance.
Warning
If you compile the ixgbevf module for your current kernel and then upgrade your
kernel without rebuilding the driver for the new kernel, your system might revert to the
distribution-specific ixgbevf module at the next reboot. This could make your system
unreachable if the distribution-specific version is incompatible with enhanced networking.
4. Run the sudo depmod command to update module dependencies.
5. Update initramfs on your instance to ensure that the new module loads at boot time.
6. Determine if your system uses predictable network interface names by default. Systems that use
systemd or udev versions 197 or greater can rename Ethernet devices and they do not guarantee
that a single network interface will be named eth0. This behavior can cause problems connecting to
your instance. For more information and to see other configuration options, see Predictable Network
Interface Names on the freedesktop.org website.

a. You can check the systemd or udev versions on RPM-based systems with the following
command:

[ec2-user ~]$ rpm -qa | grep -e '^systemd-[0-9]\+\|^udev-[0-9]\+'

systemd-208-11.el7_0.2.x86_64

In the above Red Hat Enterprise Linux 7 example, the systemd version is 208, so predictable
network interface names must be disabled.
b. Disable predictable network interface names by adding the net.ifnames=0 option to the
GRUB_CMDLINE_LINUX line in /etc/default/grub.

[ec2-user ~]$ sudo sed -i '/^GRUB\_CMDLINE\_LINUX/s/\"$/\ net\.ifnames\=0\"/' /etc/

default/grub

c. Rebuild the grub configuration file.

[ec2-user ~]$ sudo grub2-mkconfig -o /boot/grub2/grub.cfg

7. [EBS-backed instance] From your local computer, stop the instance using the Amazon EC2 console
or one of the following commands: stop-instances (AWS CLI), Stop-EC2Instance (AWS Tools for
Windows PowerShell). If your instance is managed by AWS OpsWorks, you should stop the instance
in the AWS OpsWorks console so that the instance state remains in sync.

879
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Enhanced networking: Intel 82599 VF

[Instance store-backed instance] You can't stop the instance to modify the attribute.
Instead, proceed to this procedure: To enable enhanced networking (instance store–backed
instances) (p. 880).
8. From your local computer, enable the enhanced networking attribute using one of the following
commands:

• modify-instance-attribute (AWS CLI)

aws ec2 modify-instance-attribute --instance-id instance_id --sriov-net-support

simple

• Edit-EC2InstanceAttribute (AWS Tools for Windows PowerShell)

Edit-EC2InstanceAttribute -InstanceId instance_id -SriovNetSupport "simple"

9. (Optional) Create an AMI from the instance, as described in Create an Amazon EBS-backed Linux
AMI (p. 125) . The AMI inherits the enhanced networking attribute from the instance. Therefore, you
can use this AMI to launch another instance with enhanced networking enabled by default.
Important
If your instance operating system contains an /etc/udev/rules.d/70-persistent-
net.rules file, you must delete it before creating the AMI. This file contains the MAC
address for the Ethernet adapter of the original instance. If another instance boots with
this file, the operating system will be unable to find the device and eth0 might fail, causing
boot issues. This file is regenerated at the next boot cycle, and any instances launched from
the AMI create their own version of the file.
10. From your local computer, start the instance using the Amazon EC2 console or one of the following
commands: start-instances (AWS CLI), Start-EC2Instance (AWS Tools for Windows PowerShell). If
your instance is managed by AWS OpsWorks, you should start the instance in the AWS OpsWorks
console so that the instance state remains in sync.
11. (Optional) Connect to your instance and verify that the module is installed.

To enable enhanced networking (instance store–backed instances)

Follow the previous procedure until the step where you stop the instance. Create a new AMI as described
in Create an instance store-backed Linux AMI (p. 130), making sure to enable the enhanced networking
attribute when you register the AMI.

• register-image (AWS CLI)

aws ec2 register-image --sriov-net-support simple ...

• Register-EC2Image (AWS Tools for Windows PowerShell)

Register-EC2Image -SriovNetSupport "simple" ...

Troubleshoot connectivity issues

If you lose connectivity while enabling enhanced networking, the ixgbevf module might be
incompatible with the kernel. Try installing the version of the ixgbevf module included with the
distribution of Linux for your instance.

If you enable enhanced networking for a PV instance or AMI, this can make your instance unreachable.

For more information, see How do I enable and configure enhanced networking on my EC2 instances?.

880
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Operating system optimizations

Operating system optimizations

To achieve the maximum network performance on instances with enhanced networking, you may need
to modify the default operating system configuration. We recommend the following configuration
changes for applications that require high network performance.

In addition to these operating system optimizations, you should also consider the maximum transmission
unit (MTU) of your network traffic, and adjust according to your workload and network architecture. For
more information, see Network maximum transmission unit (MTU) for your EC2 instance (p. 939).

AWS regularly measures average round trip latencies between instances launched in a cluster placement
group of 50us and tail latencies of 200us at the 99.9 percentile. If your applications require consistently
low latencies, we recommend using the latest version of the ENA drivers on fixed performance instances
built on the Nitro System.

These procedures were written for Amazon Linux 2 and Amazon Linux AMI. However, they might also
work for other Linux distributions with kernel version 3.9 or newer. For more information, see your
system-specific documentation.

To optimize your Amazon Linux instance for enhanced networking

1. Check the clock source for your instance:

cat /sys/devices/system/clocksource/clocksource0/current_clocksource

2. If the clock source is xen, complete the following substeps. Otherwise, skip to Step 3 (p. 881).

a. Edit the GRUB configuration and add xen_nopvspin=1 and clocksource=tsc to the kernel
boot options.

• For Amazon Linux 2, edit the /etc/default/grub file and add these options to the
GRUB_CMDLINE_LINUX_DEFAULT line, as shown below:

GRUB_CMDLINE_LINUX_DEFAULT="console=tty0 console=ttyS0,115200n8 net.ifnames=0

biosdevname=0 nvme_core.io_timeout=4294967295 xen_nopvspin=1 clocksource=tsc"
GRUB_TIMEOUT=0

• For Amazon Linux AMI, edit the /boot/grub/grub.conf file and add these options to the
kernel line, as shown below:

kernel /boot/vmlinuz-4.14.62-65.117.amzn1.x86_64 root=LABEL=/ console=tty1

console=ttyS0 selinux=0 nvme_core.io_timeout=4294967295 xen_nopvspin=1
clocksource=tsc

b. (Amazon Linux 2 only) Rebuild your GRUB configuration file to pick up these changes:

sudo grub2-mkconfig -o /boot/grub2/grub.cfg

3. If your instance type is listed as supported on Processor state control for your EC2 instance (p. 658),
prevent the system from using deeper C-states to ensure low-latency system performance. For more
information, see High performance and low latency by limiting deeper C-states (p. 660).

a. Edit the GRUB configuration and add intel_idle.max_cstate=1 to the kernel boot options.

• For Amazon Linux 2, edit the /etc/default/grub file and add this option to the
GRUB_CMDLINE_LINUX_DEFAULT line, as shown below:

881
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Operating system optimizations

GRUB_CMDLINE_LINUX_DEFAULT="console=tty0 console=ttyS0,115200n8
net.ifnames=0 biosdevname=0 nvme_core.io_timeout=4294967295 xen_nopvspin=1
clocksource=tsc intel_idle.max_cstate=1"
GRUB_TIMEOUT=0

• For Amazon Linux AMI, edit the /boot/grub/grub.conf file and add this option to the
kernel line, as shown below:

kernel /boot/vmlinuz-4.14.62-65.117.amzn1.x86_64 root=LABEL=/ console=tty1

console=ttyS0 selinux=0 nvme_core.io_timeout=4294967295 xen_nopvspin=1
clocksource=tsc intel_idle.max_cstate=1

b. (Amazon Linux 2 only) Rebuild your GRUB configuration file to pick up these changes:

sudo grub2-mkconfig -o /boot/grub2/grub.cfg

4. Ensure that your reserved kernel memory is sufficient to sustain a high rate of packet buffer
allocations (the default value may be too small).

a. Open (as root or with sudo) the /etc/sysctl.conf file with the editor of your choice.
b. Add the vm.min_free_kbytes line to the file with the reserved kernel memory value (in
kilobytes) for your instance type. As a rule of thumb, you should set this value to between
1-3% of available system memory, and adjust this value up or down to meet the needs of your
application.

vm.min_free_kbytes = 1048576

c. Apply this configuration with the following command:

sudo sysctl -p

d. Verify that the setting was applied with the following command:

sudo sysctl -a 2>&1 | grep min_free_kbytes

5. Reboot your instance to load the new configuration:

sudo reboot

6. (Optional) Manually distribute packet receive interrupts so that they are associated with different
CPUs that all belong to the same NUMA node. Use this carefully, however, because irqbalancer is
disabled globally.
Note
The configuration change in this step does not survive a reboot.

a. Create a file called smp_affinity.sh and paste the following code block into it:

#/bin/sh
service irqbalance stop
affinity_values=(00000001 00000002 00000004 00000008 00000010 00000020 00000040
00000080)
irqs=($(grep eth /proc/interrupts|awk '{print $1}'|cut -d : -f 1))
irqLen=${#irqs[@]}
for (( i=0; i<${irqLen}; i++ ));
do
echo $(printf "0000,00000000,00000000,00000000,${affinity_values[$i]}") > /proc/
irq/${irqs[$i]}/smp_affinity;

882
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Operating system optimizations

echo "IRQ ${irqs[$i]} =" $(cat /proc/irq/${irqs[$i]}/smp_affinity);

done

b. Run the script with the following command:

sudo bash ./smp_affinity.sh

7. (Optional) If the vCPUs that handle receive IRQs are overloaded, or if your application network
processing is demanding on CPU, you can offload part of the network processing to other cores
with receive packet steering (RPS). Ensure that cores used for RPS belong to the same NUMA node
to avoid inter-NUMA node locks. For example, to use cores 8-15 for packet processing, use the
following command.
Note
The configuration change in this step does not survive a reboot.

for i in `seq 0 7`; do echo $(printf "0000,00000000,00000000,00000000,0000ff00") | sudo

tee /sys/class/net/eth0/queues/rx-$i/rps_cpus; done

8. (Optional) If possible, keep all processing on the same NUMA node.

a. Install numactl:

sudo yum install -y numactl

b. When you run your network processing program, bind it to a single NUMA node. For example,
the following command binds the shell script, run.sh, to NUMA node 0:

numactl --cpunodebind=0 --membind=0 run.sh

c. If you have hyperthreading enabled, you can configure your application to only use a single
hardware thread per CPU core.

• You can view which CPU cores map to a NUMA node with the lscpu command:

lscpu | grep NUMA

Output:

NUMA node(s): 2
NUMA node0 CPU(s): 0-15,32-47
NUMA node1 CPU(s): 16-31,48-63

• You can view which hardware threads belong to a physical CPU with the following command:

cat /sys/devices/system/cpu/cpu0/topology/thread_siblings_list

Output:

0,32

In this example, threads 0 and 32 map to CPU 0.

• To avoid running on threads 32-47 (which are actually hardware threads of the same CPUs as
0-15, use the following command:

numactl --physcpubind=+0-15 --membind=0 ./run.sh

883
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Monitor network performance

9. Use multiple elastic network interfaces for different classes of traffic. For example, if you are running
a web server that uses a backend database, use one elastic network interfaces for the web server
front end, and another for the database connection.

Monitor network performance for your instance

The Elastic Network Adapter (ENA) driver publishes network performance metrics from the instances
where they are enabled. You can use these metrics to troubleshoot instance performance issues, choose
the right instance size for a workload, plan scaling activities proactively, and benchmark applications to
determine whether they maximize the performance available on an instance.

Amazon EC2 defines network maximums at the instance level to ensure a high-quality networking
experience, including consistent network performance across instance sizes. AWS provides maximums for
the following for each instance:

• Bandwidth capability — Each EC2 instance has a maximum bandwidth for aggregate inbound
and outbound traffic, based on instance type and size. Some instances use a network I/O credit
mechanism to allocate network bandwidth based on average bandwidth utilization. Amazon EC2 also
has maximum bandwidth for traffic to AWS Direct Connect and the internet.
• Packet-per-second (PPS) performance — Each EC2 instance has a maximum PPS performance, based
on instance type and size.
• Connections tracked — The security group tracks each connection established to ensure that return
packets are delivered as expected. There is a maximum number of connections that can be tracked per
instance.
• Link-local service access — Amazon EC2 provides a maximum PPS per network interface for traffic to
services such as the DNS service, the Instance Metadata Service, and the Amazon Time Sync Service.

When the network traffic for an instance exceeds a maximum, AWS shapes the traffic that exceeds the
maximum by queueing and then dropping network packets. You can monitor when traffic exceeds a
maximum using the network performance metrics. These metrics inform you of impact to network
traffic, and possible network performance issues, in real time.

Contents
• Requirements (p. 884)
• Metrics for the ENA driver (p. 885)
• View the network performance metrics for your Linux instance (p. 885)
• Network performance metrics with the DPDK driver for ENA (p. 885)
• Metrics on instances running FreeBSD (p. 887)

Requirements
The following requirements apply to Linux instances.

• Install ENA driver version 2.2.10 or later. To verify the installed version, use the ethtool command. In
the following example, the version meets the minimum requirement.

[ec2-user ~]$ ethtool -i eth0 | grep version

version: 2.2.10

To upgrade your ENA driver, see Enhanced networking (p. 865).

• To import these metrics to Amazon CloudWatch, install the CloudWatch agent. For more information,
see Collect network performance metrics in the Amazon CloudWatch User Guide.

884
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Monitor network performance

Metrics for the ENA driver

The ENA driver delivers the following metrics to the instance in real time. They provide the cumulative
number of packets shaped on each network interface since the last driver reset.

The following metrics are available on Linux instances, FreeBSD instances, and DPDK environments.

Metric Description

bw_in_allowance_exceeded The number of packets shaped because the inbound

aggregate bandwidth exceeded the maximum for the
instance.

bw_out_allowance_exceeded The number of packets shaped because the outbound

aggregate bandwidth exceeded the maximum for the
instance.

conntrack_allowance_exceeded The number of packets shaped because connection

tracking exceeded the maximum for the instance and
new connections could not be established. This can
result in packet loss for traffic to or from the instance.

linklocal_allowance_exceeded The number of packets shaped because the PPS of the

traffic to local proxy services exceeded the maximum for
the network interface. This impacts traffic to the DNS
service, the Instance Metadata Service, and the Amazon
Time Sync Service.

pps_allowance_exceeded The number of packets shaped because the

bidirectional PPS exceeded the maximum for the
instance.

View the network performance metrics for your Linux instance

You can publish metrics to your favorite tools to visualize the metric data. For example, you can publish
the metrics to Amazon CloudWatch using the CloudWatch agent. The agent enables you to select
individual metrics and control publication.

You can also use the ethtool to retrieve the metrics for each network interface, such as eth0, as follows.

[ec2-user ~]$ ethtool -S eth0

bw_in_allowance_exceeded: 0
bw_out_allowance_exceeded: 0
pps_allowance_exceeded: 0
conntrack_allowance_exceeded: 0
linklocal_allowance_exceeded: 0

Network performance metrics with the DPDK driver for ENA

The ENA driver version 2.2.0 and later supports network metrics reporting. DPDK 20.11 includes the ENA
driver 2.2.0 and is the first DPDK version to support this feature.

You can use an example application to view DPDK statistics. To start an interactive version of the
example application, run the following command.

./app/dpdk-testpmd -- -i

885
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Monitor network performance

Within this interactive session, you can enter a command to retrieve extended statistics for a port. The
following example command retrieves the statistics for port 0.

show port xstats 0

The following is an example of an interactive session with the DPDK example application.

[[email protected] build]# ./app/dpdk-testpmd -- -i

EAL: Detected 4 lcore(s)
EAL: Detected 1 NUMA nodes
EAL: Multi-process socket /var/run/dpdk/rte/mp_socket
EAL: Selected IOVA mode 'PA'
EAL: Probing VFIO support...
EAL: Invalid NUMA socket, default to 0
EAL: Invalid NUMA socket, default to 0
EAL: Probe PCI driver: net_ena (1d0f:ec20) device: 0000:00:06.0
(socket 0)
EAL: No legacy callbacks, legacy socket not created
Interactive-mode selected

Port 0: link state change event

testpmd: create a new mbuf pool <mb_pool_0>: n=171456,
size=2176, socket=0
testpmd: preferred mempool ops selected: ring_mp_mc

Warning! port-topology=paired and odd forward ports number, the

last port will pair with itself.

Configuring Port 0 (socket 0)

Port 0: 02:C7:17:A2:60:B1
Checking link statuses...
Done
Error during enabling promiscuous mode for port 0: Operation
not supported - ignore
testpmd> show port xstats 0
###### NIC extended statistics for port 0
rx_good_packets: 0
tx_good_packets: 0
rx_good_bytes: 0
tx_good_bytes: 0
rx_missed_errors: 0
rx_errors: 0
tx_errors: 0
rx_mbuf_allocation_errors: 0
rx_q0_packets: 0
rx_q0_bytes: 0
rx_q0_errors: 0
tx_q0_packets: 0
tx_q0_bytes: 0
wd_expired: 0
dev_start: 1
dev_stop: 0
tx_drops: 0
bw_in_allowance_exceeded: 0
bw_out_allowance_exceeded: 0
pps_allowance_exceeded: 0
conntrack_allowance_exceeded: 0
linklocal_allowance_exceeded: 0
rx_q0_cnt: 0
rx_q0_bytes: 0
rx_q0_refill_partial: 0
rx_q0_bad_csum: 0
rx_q0_mbuf_alloc_fail: 0
rx_q0_bad_desc_num: 0

886
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot ENA

rx_q0_bad_req_id: 0
tx_q0_cnt: 0
tx_q0_bytes: 0
tx_q0_prepare_ctx_err: 0
tx_q0_linearize: 0
tx_q0_linearize_failed: 0
tx_q0_tx_poll: 0
tx_q0_doorbells: 0
tx_q0_bad_req_id: 0
tx_q0_available_desc: 1023
testpmd>

For more information about the example application and using it to retrieve extended statistics. see
Testpmd Application User Guide in the DPDK documentation.

Metrics on instances running FreeBSD

Starting with version 2.3.0, the ENA FreeBSD driver supports collecting network performance metrics on
instances running FreeBSD. To enable the collection of FreeBSD metrics, enter the following commmand
and set interval to a value between 1 and 3600. This specifies how often, in seconds, to collect
FreeBSD metrics.

sysctl dev.ena.network_interface.eni_metrics.sample_interval=interval

For example, the following command sets the driver to collect FreeBSD metrics on network interface 1
every 10 seconds:

sysctl dev.ena.1.eni_metrics.sample_interval=10

To turn off the collection of FreeBSD metrics, you can run the preceding command and specify 0 as the
interval.

Once you are collecting FreeBSD metrics, you can retrieve the latest set of collected metrics by running
the following command.

sysctl dev.ena.network_interface.en1_metrics

Troubleshoot the Elastic Network Adapter (ENA)

The Elastic Network Adapter (ENA) is designed to improve operating system health and reduce the
chances of long-term disruption because of unexpected hardware behavior and or failures. The ENA
architecture keeps device or driver failures as transparent to the system as possible. This topic provides
troubleshooting information for ENA.

If you are unable to connect to your instance, start with the Troubleshoot connectivity issues (p. 888)
section.

If you are able to connect to your instance, you can gather diagnostic information by using the failure
detection and recovery mechanisms that are covered in the later sections of this topic.

Contents
• Troubleshoot connectivity issues (p. 888)
• Keep-alive mechanism (p. 889)
• Register read timeout (p. 890)

887
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot ENA

• Statistics (p. 890)

• Driver error logs in syslog (p. 894)

Troubleshoot connectivity issues

If you lose connectivity while enabling enhanced networking, the ena module might be incompatible
with your instance's current running kernel. This can happen if you install the module for a specific kernel
version (without dkms, or with an improperly configured dkms.conf file) and then your instance kernel
is updated. If the instance kernel that is loaded at boot time does not have the ena module properly
installed, your instance will not recognize the network adapter and your instance becomes unreachable.

If you enable enhanced networking for a PV instance or AMI, this can also make your instance
unreachable.

If your instance becomes unreachable after enabling enhanced networking with ENA, you can disable the
enaSupport attribute for your instance and it will fall back to the stock network adapter.

To disable enhanced networking with ENA (EBS-backed instances)

1. From your local computer, stop the instance using the Amazon EC2 console or one of the following
commands: stop-instances (AWS CLI), Stop-EC2Instance (AWS Tools for Windows PowerShell). If your
instance is managed by AWS OpsWorks, you should stop the instance in the AWS OpsWorks console
so that the instance state remains in sync.
Important
If you are using an instance store-backed instance, you can't stop the instance.
Instead, proceed to To disable enhanced networking with ENA (instance store-backed
instances) (p. 888).
2. From your local computer, disable the enhanced networking attribute using the following command.

• modify-instance-attribute (AWS CLI)

$ aws ec2 modify-instance-attribute --instance-id instance_id --no-ena-support

3. From your local computer, start the instance using the Amazon EC2 console or one of the following
commands: start-instances (AWS CLI), Start-EC2Instance (AWS Tools for Windows PowerShell). If
your instance is managed by AWS OpsWorks, you should start the instance in the AWS OpsWorks
console so that the instance state remains in sync.
4. (Optional) Connect to your instance and try reinstalling the ena module with your current kernel
version by following the steps in Enable enhanced networking with the Elastic Network Adapter
(ENA) on Linux instances (p. 865).

To disable enhanced networking with ENA (instance store-backed instances)

If your instance is an instance store-backed instance, create a new AMI as described in Create an instance
store-backed Linux AMI (p. 130). Be sure to disable the enhanced networking enaSupport attribute
when you register the AMI.

• register-image (AWS CLI)

$ aws ec2 register-image --no-ena-support ...

• Register-EC2Image (AWS Tools for Windows PowerShell)

C:\> Register-EC2Image -EnaSupport $false ...

888
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot ENA

Keep-alive mechanism
The ENA device posts keep-alive events at a fixed rate (usually once every second). The ENA driver
implements a watchdog mechanism, which checks for the presence of these keep-alive messages. If a
message or messages are present, the watchdog is rearmed, otherwise the driver concludes that the
device experienced a failure and then does the following:

• Dumps its current statistics to syslog

• Resets the ENA device
• Resets the ENA driver state

The above reset procedure may result in some traffic loss for a short period of time (TCP connections
should be able to recover), but should not otherwise affect the user.

The ENA device may also indirectly request a device reset procedure, by not sending a keep-alive
notification, for example, if the ENA device reaches an unknown state after loading an irrecoverable
configuration.

Below is an example of the reset procedure:

[18509.800135] ena 0000:00:07.0 eth1: Keep alive watchdog timeout. // The watchdog process
initiates a reset
[18509.815244] ena 0000:00:07.0 eth1: Trigger reset is on
[18509.825589] ena 0000:00:07.0 eth1: tx_timeout: 0 // The driver logs the current
statistics
[18509.834253] ena 0000:00:07.0 eth1: io_suspend: 0
[18509.842674] ena 0000:00:07.0 eth1: io_resume: 0
[18509.850275] ena 0000:00:07.0 eth1: wd_expired: 1
[18509.857855] ena 0000:00:07.0 eth1: interface_up: 1
[18509.865415] ena 0000:00:07.0 eth1: interface_down: 0
[18509.873468] ena 0000:00:07.0 eth1: admin_q_pause: 0
[18509.881075] ena 0000:00:07.0 eth1: queue_0_tx_cnt: 0
[18509.888629] ena 0000:00:07.0 eth1: queue_0_tx_bytes: 0
[18509.895286] ena 0000:00:07.0 eth1: queue_0_tx_queue_stop: 0
.......
........
[18511.280972] ena 0000:00:07.0 eth1: free uncompleted tx skb qid 3 idx 0x7 // At the end
of the down process, the driver discards incomplete packets.
[18511.420112] [ENA_COM: ena_com_validate_version] ena device version: 0.10 //The driver
begins its up process
[18511.420119] [ENA_COM: ena_com_validate_version] ena controller version: 0.0.1
implementation version 1
[18511.420127] [ENA_COM: ena_com_admin_init] ena_defs : Version:[b9692e8] Build date [Wed
Apr 6 09:54:21 IDT 2016]
[18512.252108] ena 0000:00:07.0: Device watchdog is Enabled
[18512.674877] ena 0000:00:07.0: irq 46 for MSI/MSI-X
[18512.674933] ena 0000:00:07.0: irq 47 for MSI/MSI-X
[18512.674990] ena 0000:00:07.0: irq 48 for MSI/MSI-X
[18512.675037] ena 0000:00:07.0: irq 49 for MSI/MSI-X
[18512.675085] ena 0000:00:07.0: irq 50 for MSI/MSI-X
[18512.675141] ena 0000:00:07.0: irq 51 for MSI/MSI-X
[18512.675188] ena 0000:00:07.0: irq 52 for MSI/MSI-X
[18512.675233] ena 0000:00:07.0: irq 53 for MSI/MSI-X
[18512.675279] ena 0000:00:07.0: irq 54 for MSI/MSI-X
[18512.772641] [ENA_COM: ena_com_set_hash_function] Feature 10 isn't supported
[18512.772647] [ENA_COM: ena_com_set_hash_ctrl] Feature 18 isn't supported
[18512.775945] ena 0000:00:07.0: Device reset completed successfully // The reset process
is complete

889
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot ENA

Register read timeout

The ENA architecture suggests a limited usage of memory mapped I/O (MMIO) read operations. MMIO
registers are accessed by the ENA device driver only during its initialization procedure.

If the driver logs (available in dmesg output) indicate failures of read operations, this may be caused by
an incompatible or incorrectly compiled driver, a busy hardware device, or hardware failure.

Intermittent log entries that indicate failures on read operations should not be considered an issue; the
driver will retry them in this case. However, a sequence of log entries containing read failures indicate a
driver or hardware problem.

Below is an example of driver log entry indicating a read operation failure due to a timeout:

[ 47.113698] [ENA_COM: ena_com_reg_bar_read32] reading reg failed for timeout. expected:

req id[1] offset[88] actual: req id[57006] offset[0] 
[ 47.333715] [ENA_COM: ena_com_reg_bar_read32] reading reg failed for timeout. expected:
req id[2] offset[8] actual: req id[57007] offset[0] 
[ 47.346221] [ENA_COM: ena_com_dev_reset] Reg read32 timeout occurred

Statistics
If you experience insufficient network performance or latency issues, you should retrieve the device
statistics and examine them. These statistics can be obtained using ethtool as follows.

[ec2-user ~]$ ethtool –S ethN

NIC statistics:
tx_timeout: 0
suspend: 0
resume: 0
wd_expired: 0
interface_up: 1
interface_down: 0
admin_q_pause: 0
bw_in_allowance_exceeded: 0
bw_out_allowance_exceeded: 0
pps_allowance_exceeded: 0
conntrack_allowance_exceeded: 0
linklocal_allowance_exceeded: 0
queue_0_tx_cnt: 4329
queue_0_tx_bytes: 1075749
queue_0_tx_queue_stop: 0
...

The following command output parameters are described below:

tx_timeout: N

The number of times that the Netdev watchdog was activated.

suspend: N

The number of times the driver performed a suspend operation.

resume: N

The number of times the driver performed a resume operation.

wd_expired: N

The number of times that the driver did not receive the keep-alive event in the preceding three
seconds.

890
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot ENA

interface_up: N

The number of times that the ENA interface was brought up.
interface_down: N

The number of times that the ENA interface was brought down.
admin_q_pause: N

The number of times the admin queue was not found in a running state.
bw_in_allowance_exceeded: N

The number of rx packets dropped because the bandwidth allowance limit was exceeded.
bw_out_allowance_exceeded: N

The number of tx packets dropped because the bandwidth allowance limit was exceeded.
pps_allowance_exceeded: N

The number of packets dropped because the pps (packets per second) allowance limit was exceeded.
conntrack_allowance_exceeded: N

The number of packets dropped because the connection count allowance limit was exceeded.
linklocal_allowance_exceeded: N

The number of proxy packets dropped because the pps (packets per second) allowance limit was
exceeded.
queue_N_tx_cnt: N

The number of transmitted packets for this queue.

queue_N_tx_bytes: N

The number of transmitted bytes for this queue.

queue_N_tx_queue_stop: N

The number of times that queue N was full and stopped.

queue_N_tx_queue_wakeup: N

The number of times that queue N resumed after being stopped.

queue_N_tx_dma_mapping_err: N

Direct memory access error count. If this value is not 0, it indicates low system resources.
queue_N_tx_linearize: N

The number of times SKB linearization was attempted for this queue.
queue_N_tx_linearize_failed: N

The number of times SKB linearization failed for this queue.

queue_N_tx_napi_comp: N

The number of times the napi handler called napi_complete for this queue.
queue_N_tx_tx_poll: N

The number of times the napi handler was scheduled for this queue.
queue_N_tx_doorbells: N

The number of transmission doorbells for this queue.

891
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot ENA

queue_N_tx_prepare_ctx_err: N

The number of times ena_com_prepare_tx failed for this queue.

queue_N_tx_bad_req_id: N

Invalid req_id for this queue. The valid req_id is zero, minus the queue_size, minus 1.
queue_N_tx_llq_buffer_copy: N

The number of packets whose headers size are larger than llq entry for this queue.
queue_N_tx_missed_tx: N

The number of packets that were left uncompleted for this queue.
queue_N_tx_unmask_interrupt: N

The number of times the tx interrupt was unmasked for this queue.
queue_N_rx_cnt: N

The number of received packets for this queue.

queue_N_rx_bytes: N

The number of received bytes for this queue.

queue_N_rx_rx_copybreak_pkt: N

The number of times the rx queue received a packet that is less than the rx_copybreak packet size
for this queue.
queue_N_rx_csum_good: N

The number of times the rx queue received a packet where the checksum was checked and was
correct for this queue.
queue_N_rx_refil_partial: N

The number of times the driver did not succeed in refilling the empty portion of the rx queue with
the buffers for this queue. If this value is not zero, it indicates low memory resources.
queue_N_rx_bad_csum: N

The number of times the rx queue had a bad checksum for this queue (only if rx checksum offload is
supported).
queue_N_rx_page_alloc_fail: N

The number of time that page allocation failed for this queue. If this value is not zero, it indicates
low memory resources.
queue_N_rx_skb_alloc_fail: N

The number of time that SKB allocation failed for this queue. If this value is not zero, it indicates low
system resources.
queue_N_rx_dma_mapping_err: N

Direct memory access error count. If this value is not 0, it indicates low system resources.
queue_N_rx_bad_desc_num: N

Too many buffers per packet. If this value is not 0, it indicates the use of very small buffers.
queue_N_rx_bad_req_id: N

The req_id for this queue is not valid. The valid req_id is from [0, queue_size - 1 ].
queue_N_rx_empty_rx_ring: N

The number of times the rx queue was empty for this queue.

892
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot ENA

queue_N_rx_csum_unchecked: N

The number of times the rx queue received a packet whose checksum wasn't checked for this queue.
queue_N_rx_xdp_aborted: N

The number of times that an XDP packet was classified as XDP_ABORT.

queue_N_rx_xdp_drop: N

The number of times that an XDP packet was classified as XDP_DROP.

queue_N_rx_xdp_pass: N

The number of times that an XDP packet was classified as XDP_PASS.

queue_N_rx_xdp_tx: N

The number of times that an XDP packet was classified as XDP_TX.

queue_N_rx_xdp_invalid: N

The number of times that the XDP return code for the packet was not valid.
queue_N_rx_xdp_redirect: N

The number of times that an XDP packet was classified as XDP_REDIRECT.

queue_N_xdp_tx_cnt: N

The number of transmitted packets for this queue.

queue_N_xdp_tx_bytes: N

The number of transmitted bytes for this queue.

queue_N_xdp_tx_queue_stop: N

The number of times that this queue was full and stopped.
queue_N_xdp_tx_queue_wakeup: N

The number of times that this queue resumed after being stopped.
queue_N_xdp_tx_dma_mapping_err: N

Direct memory access error count. If this value is not 0, it indicates low system resources.
queue_N_xdp_tx_linearize: N

he number of times XDP buffer linearization was attempted for this queue.
queue_N_xdp_tx_linearize_failed: N

The number of times XDP buffer linearization failed for this queue.
queue_N_xdp_tx_napi_comp: N

The number of times the napi handler called napi_complete for this queue.
queue_N_xdp_tx_tx_poll: N

The number of times the napi handler was scheduled for this queue.
queue_N_xdp_tx_doorbells: N

The number of transmission doorbells for this queue.

queue_N_xdp_tx_prepare_ctx_err: N

The number of times ena_com_prepare_tx failed for this queue. This value should always be zero; if
not, see the driver logs.

893
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot ENA

queue_N_xdp_tx_bad_req_id: N

The req_id for this queue is not valid. The valid req_id is from [0, queue_size - 1 ].
queue_N_xdp_tx_llq_buffer_copy: N

The number of packets that had their headers copied using llq buffer copy for this queue.
queue_N_xdp_tx_missed_tx: N

The number of times a tx queue entry missed a completion timeout for this queue.
queue_N_xdp_tx_unmask_interrupt: N

The number of times the tx interrupt was unmasked for this queue.
ena_admin_q_aborted_cmd: N

The number of admin commands that were aborted. This usually happens during the auto-recovery
procedure.
ena_admin_q_submitted_cmd: N

The number of admin queue doorbells.

ena_admin_q_completed_cmd: N

The number of admin queue completions.

ena_admin_q_out_of_space: N

The number of times that the driver tried to submit new admin command, but the queue was full.
ena_admin_q_no_completion: N

The number of times that the driver did not get an admin completion for a command.

Driver error logs in syslog

The ENA driver writes log messages to syslog during system boot. You can examine these logs to look
for errors if you are experiencing issues. Below is an example of information logged by the ENA driver in
syslog during system boot, along with some annotations for select messages.

Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 478.416939] [ENA_COM: ena_com_validate_version]

ena device version: 0.10
Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 478.420915] [ENA_COM: ena_com_validate_version]
ena controller version: 0.0.1 implementation version 1
Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 479.256831] ena 0000:00:03.0: Device watchdog is
Enabled
Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 479.672947] ena 0000:00:03.0: creating 8 io
queues. queue size: 1024
Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 479.680885] [ENA_COM:
ena_com_init_interrupt_moderation] Feature 20 isn't supported // Interrupt moderation is
not supported by the device
Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 479.691609] [ENA_COM: ena_com_get_feature_ex]
Feature 10 isn't supported // RSS HASH function configuration is not supported by the
device
Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 479.694583] [ENA_COM: ena_com_get_feature_ex]
Feature 18 isn't supported //RSS HASH input source configuration is not supported by the
device
Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 479.697433] [ENA_COM:
ena_com_set_host_attributes] Set host attribute isn't supported
Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 479.701064] ena 0000:00:03.0 (unnamed
net_device) (uninitialized): Cannot set host attributes
Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 479.704917] ena 0000:00:03.0: Elastic Network
Adapter (ENA) found at mem f3000000, mac addr 02:8a:3c:1e:13:b5 Queues 8

894
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Elastic Fabric Adapter

Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 480.805037] EXT4-fs (xvda1): re-mounted. Opts:

(null)
Jun 3 22:37:46 ip-172-31-2-186 kernel: [ 481.025842] NET: Registered protocol family 10

Which errors can I ignore?

The following warnings that may appear in your system's error logs can be ignored for the Elastic
Network Adapter:

Set host attribute isn't supported

Host attributes are not supported for this device.

failed to alloc buffer for rx queue

This is a recoverable error, and it indicates that there may have been a memory pressure issue when
the error was thrown.
Feature X isn't supported

The referenced feature is not supported by the Elastic Network Adapter. Possible values for X
include:
• 10: RSS Hash function configuration is not supported for this device.
• 12: RSS Indirection table configuration is not supported for this device.
• 18: RSS Hash Input configuration is not supported for this device.
• 20: Interrupt moderation is not supported for this device.
• 27: The Elastic Network Adapter driver does not support polling the Ethernet capabilities from
snmpd.
Failed to config AENQ

The Elastic Network Adapter does not support AENQ configuration.

Trying to set unsupported AENQ events

This error indicates an attempt to set an AENQ events group that is not supported by the Elastic
Network Adapter.

Elastic Fabric Adapter

An Elastic Fabric Adapter (EFA) is a network device that you can attach to your Amazon EC2 instance to
accelerate High Performance Computing (HPC) and machine learning applications. EFA enables you to
achieve the application performance of an on-premises HPC cluster, with the scalability, flexibility, and
elasticity provided by the AWS Cloud.

EFA provides lower and more consistent latency and higher throughput than the TCP transport
traditionally used in cloud-based HPC systems. It enhances the performance of inter-instance
communication that is critical for scaling HPC and machine learning applications. It is optimized to work
on the existing AWS network infrastructure and it can scale depending on application requirements.

EFA integrates with Libfabric 1.11.1 and it supports Open MPI 4.0.5 and Intel MPI 2019 Update 7 for HPC
applications, and Nvidia Collective Communications Library (NCCL) for machine learning applications.
Note
The OS-bypass capabilities of EFAs are not supported on Windows instances. If you attach an
EFA to a Windows instance, the instance functions as an Elastic Network Adapter, without the
added EFA capabilities.

Contents
• EFA basics (p. 896)

895
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EFA basics

• Supported interfaces and libraries (p. 897)

• Supported instance types (p. 897)
• Supported AMIs (p. 898)
• EFA limitations (p. 898)
• Get started with EFA and MPI (p. 898)
• Get started with EFA and NCCL (p. 906)
• Work with EFA (p. 923)
• Monitor an EFA (p. 925)
• Verify the EFA installer using a checksum (p. 926)

EFA basics
An EFA is an Elastic Network Adapter (ENA) with added capabilities. It provides all of the functionality of
an ENA, with an additional OS-bypass functionality. OS-bypass is an access model that allows HPC and
machine learning applications to communicate directly with the network interface hardware to provide
low-latency, reliable transport functionality.

Traditionally, HPC applications use the Message Passing Interface (MPI) to interface with the system’s
network transport. In the AWS Cloud, this has meant that applications interface with MPI, which then

896
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Supported interfaces and libraries

uses the operating system's TCP/IP stack and the ENA device driver to enable network communication
between instances.

With an EFA, HPC applications use MPI or NCCL to interface with the Libfabric API. The Libfabric API
bypasses the operating system kernel and communicates directly with the EFA device to put packets on
the network. This reduces overhead and enables the HPC application to run more efficiently.
Note
Libfabric is a core component of the OpenFabrics Interfaces (OFI) framework, which defines and
exports the user-space API of OFI. For more information, see the Libfabric OpenFabrics website.

Differences between EFAs and ENAs

Elastic Network Adapters (ENAs) provide traditional IP networking features that are required to support
VPC networking. EFAs provide all of the same traditional IP networking features as ENAs, and they also
support OS-bypass capabilities. OS-bypass enables HPC and machine learning applications to bypass the
operating system kernel and to communicate directly with the EFA device.

Supported interfaces and libraries

EFA supports the following interfaces and libraries:

• Open MPI 4.0.5

• Intel MPI 2019 Update 7
• NVIDIA Collective Communications Library (NCCL) 2.4.2 and later

Supported instance types

The following instance types support EFAs:

• General purpose: m5dn.24xlarge | m5n.24xlarge | m5zn.12xlarge | m5zn.metal

• Compute optimized: c5n.18xlarge | c5n.metal | c6gn.16xlarge
• Memory optimized: r5dn.24xlarge | r5n.24xlarge
• Storage optimized: i3en.24xlarge | i3en.metal
• Accelerated computing: g4dn.metal | inf1.24xlarge | p3dn.24xlarge | p4d.24xlarge

The available instance types vary by Region. To see the available instance types that support EFA in
a Region, use the describe-instance-types command with the --region option and the appropriate
Region code.

aws ec2 describe-instance-types --region us-east-2 --filters Name=network-info.efa-

supported,Values=true --query "InstanceTypes[*].[InstanceType]" --output text

The following is example output.

g4dn.metal
i3en.24xlarge
r5n.24xlarge
c5n.18xlarge
m5n.24xlarge
inf1.24xlarge
m5dn.24xlarge
c5n.metal
p3dn.24xlarge

897
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Supported AMIs

i3en.metal
r5dn.24xlarge

Supported AMIs
The following AMIs support EFA with Intel x86-based instance types:

• Amazon Linux and Amazon Linux 2

• CentOS 7 and 8
• RHEL 7.6, 7.7, 7.8, 8.2, and 8.3
• Ubuntu 16.04, 18.04, and 20.04
• SUSE Linux Enterprise 15 SP2
• openSUSE Leap 15.2

The following AMIs support EFA with Arm-based (Graviton 2) instance types:

• Amazon Linux 2
• Ubuntu 18.04 and 20.04
• openSUSE Leap 15.2

EFA limitations
EFA has the following limitations:

• p4d.24xlarge instances support up to four EFAs. All other supported instance types support only
one EFA per instance.
• EFA OS-bypass traffic is limited to a single subnet. In other words, EFA traffic cannot be sent from one
subnet to another. Normal IP traffic from the EFA can be sent from one subnet to another.
• EFA OS-bypass traffic is not routable. Normal IP traffic from the EFA remains routable.
• The EFA must be a member of a security group that allows all inbound and outbound traffic to and
from the security group itself.
• EFA traffic between C6gn instances and other EFA-enabled instances is not supported.

Get started with EFA and MPI

This tutorial helps you to launch an EFA and MPI-enabled instance cluster for HPC workloads. In this
tutorial, you will perform the following steps:

Contents
• Step 1: Prepare an EFA-enabled security group (p. 899)
• Step 2: Launch a temporary instance (p. 899)
• Step 3: Install the EFA software (p. 900)
• Step 4: Disable ptrace protection (p. 902)
• Step 5: (Optional) Install Intel MPI (p. 903)
• Step 6: Install your HPC application (p. 904)
• Step 7: Create an EFA-enabled AMI (p. 904)
• Step 8: Launch EFA-enabled instances into a cluster placement group (p. 904)
• Step 9: Terminate the temporary instance (p. 905)

898
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and MPI

• Step 10: Enable passwordless SSH (p. 905)

Step 1: Prepare an EFA-enabled security group

An EFA requires a security group that allows all inbound and outbound traffic to and from the security
group itself. The following procedure allows all inbound and outbound traffic for testing purposes only.
For other scenarios, see Security group rules for different use cases (p. 1072).

To create an EFA-enabled security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups and then choose Create Security Group.
3. In the Create Security Group window, do the following:

a. For Security group name, enter a descriptive name for the security group, such as EFA-
enabled security group.
b. (Optional) For Description, enter a brief description of the security group.
c. For VPC, select the VPC into which you intend to launch your EFA-enabled instances.
d. Choose Create.
4. Select the security group that you created, and on the Description tab, copy the Group ID.
5. On the Inbound tab, do the following:

a. Choose Edit.
b. For Type, choose All traffic.
c. For Source, choose Custom and paste the security group ID that you copied into the field.
d. Choose Save.
6. On the Outbound tab, do the following:

a. Choose Edit.
b. For Type, choose All traffic.
c. For Destination, choose Custom and paste the security group ID that you copied into the field.
d. Choose Save.

Step 2: Launch a temporary instance

Launch a temporary instance that you can use to install and configure the EFA software components. You
use this instance to create an EFA-enabled AMI from which you can launch your EFA-enabled instances.

To launch a temporary instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Launch Instance.
3. On the Choose an AMI page, choose Select for one of the supported AMIs (p. 898).
4. On the Choose an Instance Type page, select one of the supported instance types (p. 897) and
then choose Next: Configure Instance Details.
5. On the Configure Instance Details page, do the following:

a. For Subnet, choose the subnet in which to launch the instance.

b. For Elastic Fabric Adapter, choose Enable.
c. In the Network Interfaces section, for device eth0, choose New network interface.
d. Choose Next: Add Storage.

899
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and MPI

6. On the Add Storage page, specify the volumes to attach to the instances in addition to the volumes
that are specified by the AMI (such as the root device volume). Then choose Next: Add Tags.
7. On the Add Tags page, specify a tag that you can use to identify the temporary instance, and then
choose Next: Configure Security Group.
8. On the Configure Security Group page, for Assign a security group, select Select an existing
security group, and then select the security group that you created in Step 1.
9. On the Review Instance Launch page, review the settings, and then choose Launch to choose a key
pair and to launch your instance.

Step 3: Install the EFA software

Install the EFA-enabled kernel, EFA drivers, Libfabric, and Open MPI stack that is required to support EFA
on your temporary instance.

The steps differ depending on whether you intend to use EFA with Open MPI, with Intel MPI, or with
Open MPI and Intel MPI.

To install the EFA software

1. Connect to the instance you launched. For more information, see Connect to your Linux
instance (p. 598).
2. To ensure that all of your software packages are up to date, perform a quick software update on
your instance. This process may take a few minutes.

• Amazon Linux, Amazon Linux 2, RHEL, and CentOS

$ sudo yum update -y --skip-broken

• Ubuntu

$ sudo apt-get update

$ sudo apt-get upgrade -y

• SUSE Linux Enterprise

$ sudo zypper update -y

3. Download the EFA software installation files. The software installation files are packaged into
a compressed tarball (.tar.gz) file. To download the latest stable version, use the following
command.

$ curl -O https://efa-installer.amazonaws.com/aws-efa-installer-1.11.1.tar.gz

You can also get the latest version by replacing the version number with latest in the preceding
command.
4. (Optional) Verify the authenticity and integrity of the EFA tarball (.tar.gz) file. We recommend
that you do this to verify the identity of the software publisher and to check that the file has not
been altered or corrupted since it was published. If you do not want to verify the tarball file, skip this
step.
Note
Alternatively, if you prefer to verify the tarball file by using an MD5 or SHA256 checksum
instead, see Verify the EFA installer using a checksum (p. 926).

900
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and MPI

a. Download the public GPG key and import it into your keyring.

$ wget https://efa-installer.amazonaws.com/aws-efa-installer.key

$ gpg --import aws-efa-installer.key

The command should return a key value. Make a note of the key value, because you need it in
the next step.
b. Verify the GPG key's fingerprint. Run the following command and specify the key value from the
previous step.

$ gpg --fingerprint key_value

The command should return a fingerprint that is identical to 4E90 91BC BB97 A96B 26B1
5E59 A054 80B1 DD2D 3CCC. If the fingerprint does not match, don't run the EFA installation
script, and contact AWS Support.
c. Download the signature file and verify the signature of the EFA tarball file.

$ wget https://efa-installer.amazonaws.com/aws-efa-installer-1.11.1.tar.gz.sig

$ gpg --verify ./aws-efa-installer-1.11.1.tar.gz.sig

The following shows example output.

gpg: Signature made Wed 29 Jul 2020 12:50:13 AM UTC using RSA key ID DD2D3CCC
gpg: Good signature from "Amazon EC2 EFA <[email protected]>"
gpg: WARNING: This key is not certified with a trusted signature!
gpg: There is no indication that the signature belongs to the owner.
Primary key fingerprint: 4E90 91BC BB97 A96B 26B1 5E59 A054 80B1 DD2D 3CCC

If the result includes Good signature, and the fingerprint matches the fingerprint returned
in the previous step, proceed to the next step. If not, don't run the EFA installation script, and
contact AWS Support.
5. Extract the files from the compressed .tar.gz file and navigate into the extracted directory.

$ tar -xf aws-efa-installer-1.11.1.tar.gz

$ cd aws-efa-installer

6. Install the EFA software. Do one of the following depending on your use case.

• Open MPI and Intel MPI

If you intend to use EFA with Open MPI and Intel MPI, you must install the EFA software with
Libfabric and Open MPI, and you must complete Step 5: (Optional) Install Intel MPI. To install the
EFA software with Libfabric and Open MPI, run the following command.

$ sudo ./efa_installer.sh -y

Libfabric is installed in the /opt/amazon/efa directory, while Open MPI is installed in the /opt/
amazon/openmpi directory.

901
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and MPI

• Open MPI only

If you intend to use EFA with Open MPI only, you must install the EFA software with Libfabric and
Open MPI, and you can skip Step 5: (Optional) Install Intel MPI. To install the EFA software with
Libfabric and Open MPI, run the following command.

$ sudo ./efa_installer.sh -y

Libfabric is installed in the /opt/amazon/efa directory, while Open MPI is installed in the /opt/
amazon/openmpi directory.
• Intel MPI only

If you intend to use EFA with Intel MPI only, you can install the EFA software without Libfabric and
Open MPI. In this case, Intel MPI uses its embedded Libfabric. If you choose to do this, you must
complete Step 5: (Optional) Install Intel MPI.

To install the EFA software without Libfabric and Open MPI, run the following command.

$ sudo ./efa_installer.sh -y --minimal

7. Reboot the instance and then log back in.

8. Confirm that the EFA software components were successfully installed.

$ fi_info -p efa -t FI_EP_RDM

The command should return information about the Libfabric EFA interfaces. The following example
shows the command output.

provider: efa
fabric: EFA-fe80::94:3dff:fe89:1b70
domain: efa_0-rdm
version: 2.0
type: FI_EP_RDM
protocol: FI_PROTO_EFA

Step 4: Disable ptrace protection

To improve your HPC application's performance, Libfabric uses the instance's local memory for
interprocess communications when the processes are running on the same instance.

The shared memory feature uses Cross Memory Attach (CMA), which is not supported with ptrace
protection. If you are using a Linux distribution that has ptrace protection enabled by default, such
as Ubuntu, you must disable it. If your Linux distribution does not have ptrace protection enabled by
default, skip this step.

To disable ptrace protection

Do one of the following:

• To temporarily disable ptrace protection for testing purposes, run the following command.

$ sudo sysctl -w kernel.yama.ptrace_scope=0

• To permanently disable ptrace protection, add kernel.yama.ptrace_scope = 0 to /etc/

sysctl.d/10-ptrace.conf and reboot the instance.

902
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and MPI

Step 5: (Optional) Install Intel MPI

Important
If you intend to only use Open MPI, skip this step. Perform this step only if you intend to use
Intel MPI.

Intel MPI requires an additional installation and environment variable configuration.

Prerequisites
Ensure that the user performing the following steps has sudo permissions.

To install Intel MPI

1. To download the Intel MPI installation files, see the Intel Developer Zone website.

You must register before you can download the installation files. After you have registered, do the
following:

a. For Product, choose Intel MPI Library for Linux.

b. For Version, choose 2019 Update 7, and then choose Full Product.
2. The installation files are packaged into a compressed .tar.gz file. Extract the files from the
compressed .tar.gz file and navigate into the extracted directory.

$ tar -xf file_name.tgz

$ cd directory_name

3. Open silent.cfg using your preferred text editor. In line 10, change ACCEPT_EULA=decline to
ACCEPT_EULA=accept. Save the changes and close the file.
4. Run the installation script.

$ sudo ./install.sh -s silent.cfg

Intel MPI is installed in the /opt/intel/impi/ directory by default.

5. Add the Intel MPI environment variables to the corresponding shell startup scripts to ensure that
they are set each time that the instance starts. Do one of the following depending on your shell.

• For bash, add the following environment variable to /home/username/.bashrc and /

home/username/.bash_profile.

source /opt/intel/compilers_and_libraries/linux/mpi/intel64/bin/mpivars.sh

• For csh and tcsh, add the following environment variable to /home/username/.cshrc.

source /opt/intel/compilers_and_libraries/linux/mpi/intel64/bin/mpivars.csh

6. Log out of the instance and then log back in.

7. Run the following command to confirm that Intel MPI was successfully installed.

$ which mpicc

Ensure that the returned path includes the /opt/intel/ subdirectory.

903
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and MPI

Note
If you no longer want to use Intel MPI, remove the environment variables from the shell startup
scripts.

Step 6: Install your HPC application

Install the HPC application on the temporary instance. The installation procedure varies depending on
the specific HPC application. For more information about installing software on your Linux instance, see
Managing Software on Your Linux Instance.
Note
You might need to refer to your HPC application’s documentation for installation instructions.

Step 7: Create an EFA-enabled AMI

After you have installed the required software components, you create an AMI that you can reuse to
launch your EFA-enabled instances.

To create an AMI from your temporary instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the temporary instance that you created and choose Actions, Image, Create image.
4. For Create image, do the following:

a. For Image name, enter a descriptive name for the AMI.

b. (Optional) For Image description, enter a brief description of the purpose of the AMI.
c. Choose Create image.
5. In the navigation pane, choose AMIs.
6. Locate the AMI tht you created in the list. Wait for the status to change from pending to
available before continuing to the next step.

Step 8: Launch EFA-enabled instances into a cluster placement

group
Launch your EFA-enabled instances into a cluster placement group using the EFA-enabled AMI that you
created in Step 7, and the EFA-enabled security group that you created in Step 1.
Note
It is not an absolute requirement to launch your EFA-enabled instances into a cluster placement
group. However, we do recommend running your EFA-enabled instances in a cluster placement
group as it launches the instances into a low-latency group in a single Availability Zone.

To launch your EFA-enabled instances into a cluster placement group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Launch Instance.
3. On the Choose an AMI page, choose My AMIs, find the AMI that you created in Step 7, and then
choose Select.
4. On the Choose an Instance Type page, select one of the supported instance types (p. 897) and
then choose Next: Configure Instance Details.
5. On the Configure Instance Details page, do the following:

a. For Number of instances, enter the number of EFA-enabled instances that you want to launch.

904
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and MPI

b. For Network and Subnet, select the VPC and subnet into which to launch the instances.
c. For Placement group, select Add instance to placement group.
d. For Placement group name, select Add to a new placement group, enter a descriptive name
for the placement group, and then for Placement group strategy, select cluster.
e. For EFA, choose Enable.
f. In the Network Interfaces section, for device eth0, choose New network interface. You can
optionally specify a primary IPv4 address and one or more secondary IPv4 addresses. If you're
launching the instance into a subnet that has an associated IPv6 CIDR block, you can optionally
specify a primary IPv6 address and one or more secondary IPv6 addresses.
g. Choose Next: Add Storage.
6. On the Add Storage page, specify the volumes to attach to the instances in addition to the volumes
specified by the AMI (such as the root device volume), and then choose Next: Add Tags.
7. On the Add Tags page, specify tags for the instances, such as a user-friendly name, and then choose
Next: Configure Security Group.
8. On the Configure Security Group page, for Assign a security group, select Select an existing
security group, and then select the security group that you created in Step 1.
9. Choose Review and Launch.
10. On the Review Instance Launch page, review the settings, and then choose Launch to choose a key
pair and to launch your instances.

Step 9: Terminate the temporary instance

At this point, you no longer need the temporary instance that you launched. You can terminate the
instance to stop incurring charges for it.

To terminate the temporary instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the temporary instance that you created and then choose Actions, Instance state, Terminate
instance.
4. When prompted for confirmation, choose Terminate.

Step 10: Enable passwordless SSH

To enable your applications to run across all of the instances in your cluster, you must enable
passwordless SSH access from the leader node to the member nodes. The leader node is the instance
from which you run your applications. The remaining instances in the cluster are the member nodes.

To enable passwordless SSH between the instances in the cluster

1. Select one instance in the cluster as the leader node, and connect to it.
2. Disable strictHostKeyChecking and enable ForwardAgent on the leader node. Open ~/.ssh/
config using your preferred text editor and add the following.

Host *
ForwardAgent yes
Host *
StrictHostKeyChecking no

3. Generate an RSA key pair.

905
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

$ ssh-keygen -t rsa -N "" -f ~/.ssh/id_rsa

The key pair is created in the $HOME/.ssh/ directory.

4. Change the permissions of the private key on the leader node.

$ chmod 600 ~/.ssh/id_rsa

chmod 600 ~/.ssh/config

5. Open ~/.ssh/id_rsa.pub using your preferred text editor and copy the key.
6. For each member node in the cluster, do the following:

a. Connect to the instance.

b. Open ~/.ssh/authorized_keys using your preferred text editor and add the public key that
you copied earlier.
7. To test that the passwordless SSH is functioning as expected, connect to your leader node and run
the following command.

$ ssh member_node_private_ip

You should connect to the member node without being prompted for a key or password.

Get started with EFA and NCCL

The NVIDIA Collective Communications Library (NCCL) is a library of standard collective communication
routines for multiple GPUs across a single node or multiple nodes. NCCL can be used together with EFA,
Libfabric, and MPI to support various machine learning workloads. For more information, see the NCCL
website.
Note

• NCCL with EFA is supported with p3dn.24xlarge and p4d.24xlarge instances only.
• Only NCCL 2.4.2 and later is supported with EFA.

The following tutorials help you to launch an EFA and NCCL-enabled instance cluster for machine
learning workloads.

• Use a base AMI (p. 906)

• Use an AWS Deep Learning AMI (p. 918)

Use a base AMI

The following steps help you to get started using one of the supported base AMIs (p. 898).
Note
Only the p3dn.24xlarge and p4d.24xlarge instance types are supported.

Contents
• Step 1: Prepare an EFA-enabled security group (p. 907)
• Step 2: Launch a temporary instance (p. 907)
• Step 3: Install Nvidia GPU drivers, Nvidia CUDA toolkit, and cuDNN (p. 908)
• Step 4: Install the EFA software (p. 910)

906
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

• Step 5: Install NCCL (p. 913)

• Step 6: Install the aws-ofi-nccl plugin (p. 913)
• Step 7: Install the NCCL tests (p. 914)
• Step 8: Test your EFA and NCCL configuration (p. 914)
• Step 9: Install your machine learning applications (p. 915)
• Step 10: Create an EFA and NCCL-enabled AMI (p. 916)
• Step 11: Terminate the temporary instance (p. 916)
• Step 12: Launch EFA and NCCL-enabled instances into a cluster placement group (p. 916)
• Step 13: Enable passwordless SSH (p. 917)

Step 1: Prepare an EFA-enabled security group

An EFA requires a security group that allows all inbound and outbound traffic to and from the security
group itself. The following procedure allows all inbound and outbound traffic for testing purposes only.
For other scenarios, see Security group rules for different use cases (p. 1072).

To create an EFA-enabled security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups and then choose Create Security Group.
3. In the Create Security Group window, do the following:

a. For Security group name, enter a descriptive name for the security group, such as EFA-
enabled security group.
b. (Optional) For Description, enter a brief description of the security group.
c. For VPC, select the VPC into which you intend to launch your EFA-enabled instances.
d. Choose Create.
4. Select the security group that you created, and on the Description tab, copy the Group ID.
5. On the Inbound tab, do the following:

a. Choose Edit.
b. For Type, choose All traffic.
c. For Source, choose Custom and paste the security group ID that you copied into the field.
d. Choose Save.
6. On the Outbound tab, do the following:

a. Choose Edit.
b. For Type, choose All traffic.
c. For Destination, choose Custom and paste the security group ID that you copied into the field.
d. Choose Save.

Step 2: Launch a temporary instance

Launch a temporary instance that you can use to install and configure the EFA software components. You
use this instance to create an EFA-enabled AMI from which you can launch your EFA-enabled instances.

To launch a temporary instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Launch Instance.
3. On the Choose an AMI page, choose one of the supported AMIs.

907
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

4. On the Choose an Instance Type page, select p3dn.24xlarge or p4d.24xlarge and then choose
Next: Configure Instance Details.
5. On the Configure Instance Details page, do the following:

a. For Subnet, choose the subnet in which to launch the instance.

b. For Elastic Fabric Adapter, choose Enable.
c. In the Network Interfaces section, for device eth0, choose New network interface.
d. Choose Next: Add Storage.
6. On the Add Storage page, specify the volumes to attach to the instances, in addition to the volumes
specified by the AMI (such as the root device volume). Ensure that you provision enough storage for
the Nvidia CUDA Toolkit. Then choose Next: Add Tags.
7. On the Add Tags page, specify a tag that you can use to identify the temporary instance, and then
choose Next: Configure Security Group.
8. On the Configure Security Group page, for Assign a security group, select Select an existing
security group. Then select the security group that you created in Step 1.
9. On the Review Instance Launch page, review the settings, and then choose Launch to choose a key
pair and to launch your instance.

Step 3: Install Nvidia GPU drivers, Nvidia CUDA toolkit, and cuDNN
To install the Nvidia GPU drivers, Nvidia CUDA toolkit, and cuDNN

1. Install the utilities that are needed to install the Nvidia GPU drivers and the Nvidia CUDA toolkit.

• Amazon Linux, Amazon Linux 2, RHEL, and CentOS

$ sudo yum groupinstall 'Development Tools' -y

• Ubuntu

$ sudo apt-get update

$ sudo apt-get install build-essential -y

2. To use the Nvidia GPU driver, you must first disable the nouveau open source drivers.

a. Install the gcc compiler and the kernel headers package for the version of the kernel that you
are currently running.

• Amazon Linux, Amazon Linux 2, RHEL, and CentOS

$ sudo yum install -y gcc kernel-devel-$(uname -r)

• Ubuntu

$ sudo apt-get install -y gcc make linux-headers-$(uname -r)

b. Add nouveau to the /etc/modprobe.d/blacklist.conf deny list file.

$ cat << EOF | sudo tee --append /etc/modprobe.d/blacklist.conf

blacklist vga16fb
blacklist nouveau
blacklist rivafb
blacklist nvidiafb
blacklist rivatv
EOF

908
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

c. Open /etc/default/grub using your preferred text editor and add the following.
Note
Use sudo to open the file with root privileges.

GRUB_CMDLINE_LINUX="rdblacklist=nouveau"

d. Rebuild the Grub configuration.

• Amazon Linux, Amazon Linux 2, RHEL, and CentOS

$ sudo grub2-mkconfig -o /boot/grub2/grub.cfg

• Ubuntu

$ sudo update-grub

3. Reboot the instance and reconnect to it.

4. Download and add the Nvidia Machine Learning repositories.

• Amazon Linux, Amazon Linux 2, RHEL, and CentOS

$ wget -O /tmp/ml-repo.rpm https://developer.download.nvidia.com/compute/machine-

learning/repos/rhel7/x86_64/nvidia-machine-learning-repo-rhel7-1.0.0-1.x86_64.rpm
$ sudo rpm -Uhv /tmp/ml-repo.rpm
$ sudo yum-config-manager --add-repo https://developer.download.nvidia.com/compute/
cuda/repos/rhel7/x86_64/cuda-rhel7.repo
$ sudo yum clean all
$ sudo yum upgrade -y

• Ubuntu

$ sudo apt-key adv --fetch-keys http://developer.download.nvidia.com/compute/machine-

learning/repos/ubuntu1804/x86_64/7fa2af80.pub
$ wget -O /tmp/deeplearning.deb http://developer.download.nvidia.com/compute/
machine-learning/repos/ubuntu1804/x86_64/nvidia-machine-learning-repo-
ubuntu1804_1.0.0-1_amd64.deb
$ sudo dpkg -i /tmp/deeplearning.deb
$ wget -O /tmp/cuda.pin https://developer.download.nvidia.com/compute/cuda/repos/
ubuntu1804/x86_64/cuda-ubuntu1804.pin
$ sudo mv /tmp/cuda.pin /etc/apt/preferences.d/cuda-repository-pin-600
$ sudo apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/
repos/ubuntu1804/x86_64/7fa2af80.pub
$ sudo add-apt-repository 'deb http://developer.download.nvidia.com/compute/cuda/
repos/ubuntu1804/x86_64/ /'
$ sudo apt update

5. Install the Nvidia GPU drivers, NVIDIA CUDA toolkit, and cuDNN

• Amazon Linux, Amazon Linux 2, RHEL, and CentOS

$ sudo yum -y install cuda-toolkit-11-0 libcudnn8 libcudnn8-devel nvidia-driver-

branch-450

• Ubuntu

$ sudo apt install -o Dpkg::Options::='--force-overwrite' cuda-drivers-450 cuda-

toolkit-11-0 libcudnn8 libcudnn8-dev -y

6. Reboot the instance and reconnect to it.

909
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

7. (p4d.24xlarge instances only) Install the Nvidia Fabric Manager, start the service, and ensure that
it starts automatically when the instance starts. Nvidia Fabric Manager is required for NV Switch
Management.

• Amazon Linux, Amazon Linux 2, RHEL, and CentOS

$ sudo yum -y install nvidia-fabricmanager-450

$ sudo systemctl start nvidia-fabricmanager
$ sudo systemctl enable nvidia-fabricmanager

• Ubuntu

$ sudo apt install -o Dpkg::Options::='--force-overwrite' nvidia-fabricmanager-450

$ sudo systemctl start nvidia-fabricmanager
$ sudo systemctl enable nvidia-fabricmanager

8. Ensure that the CUDA paths are set each time that the instance starts.

• For bash shells, add the following statements to /home/username/.bashrc and /

home/username/.bash_profile.

export PATH=/usr/local/cuda/bin:/usr/local/cuda/NsightCompute-2019.1:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda/lib64:/usr/local/cuda/extras/CUPTI/lib64:
$LD_LIBRARY_PATH

• For tcsh shells, add the following statements to /home/username/.cshrc.

setenv PATH=/usr/local/cuda/bin:/usr/local/cuda/NsightCompute-2019.1:$PATH
setenv LD_LIBRARY_PATH=/usr/local/cuda/lib64:/usr/local/cuda/extras/CUPTI/lib64:
$LD_LIBRARY_PATH

9. To confirm that the Nvidia GPU drivers are functional, run the following command.

$ nvidia-smi -q | head

The command should return information about the Nvidia GPUs, Nvidia GPU drivers, and Nvidia
CUDA toolkit.

Step 4: Install the EFA software

Install the EFA-enabled kernel, EFA drivers, Libfabric, and Open MPI stack that is required to support EFA
on your temporary instance.

To install the EFA software

1. Connect to the instance you launched. For more information, see Connect to your Linux
instance (p. 598).
2. To ensure that all of your software packages are up to date, perform a quick software update on
your instance. This process may take a few minutes.

• Amazon Linux, Amazon Linux 2, RHEL, and CentOS

$ sudo yum update -y --skip-broken

• Ubuntu

910
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

$ sudo apt-get update

$ sudo apt-get upgrade -y

• SUSE Linux Enterprise

$ sudo zypper update -y

3. Download the EFA software installation files. The software installation files are packaged into
a compressed tarball (.tar.gz) file. To download the latest stable version, use the following
command.

$ curl -O https://efa-installer.amazonaws.com/aws-efa-installer-1.11.1.tar.gz

You can also get the latest version by replacing the version number with latest in the preceding
command.
4. (Optional) Verify the authenticity and integrity of the EFA tarball (.tar.gz) file. We recommend
that you do this to verify the identity of the software publisher and to check that the file has not
been altered or corrupted since it was published. If you do not want to verify the tarball file, skip this
step.
Note
Alternatively, if you prefer to verify the tarball file by using an MD5 or SHA256 checksum
instead, see Verify the EFA installer using a checksum (p. 926).

a. Download the public GPG key and import it into your keyring.

$ wget https://efa-installer.amazonaws.com/aws-efa-installer.key

$ gpg --import aws-efa-installer.key

The command should return a key value. Make a note of the key value, because you need it in
the next step.
b. Verify the GPG key's fingerprint. Run the following command and specify the key value from the
previous step.

$ gpg --fingerprint key_value

The command should return a fingerprint that is identical to 4E90 91BC BB97 A96B 26B1
5E59 A054 80B1 DD2D 3CCC. If the fingerprint does not match, don't run the EFA installation
script, and contact AWS Support.
c. Download the signature file and verify the signature of the EFA tarball file.

$ wget https://efa-installer.amazonaws.com/aws-efa-installer-1.11.1.tar.gz.sig

$ gpg --verify ./aws-efa-installer-1.11.1.tar.gz.sig

The following shows example output.

gpg: Signature made Wed 29 Jul 2020 12:50:13 AM UTC using RSA key ID DD2D3CCC
gpg: Good signature from "Amazon EC2 EFA <[email protected]>"

911
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

gpg: WARNING: This key is not certified with a trusted signature!

gpg: There is no indication that the signature belongs to the owner.
Primary key fingerprint: 4E90 91BC BB97 A96B 26B1 5E59 A054 80B1 DD2D 3CCC

If the result includes Good signature, and the fingerprint matches the fingerprint returned
in the previous step, proceed to the next step. If not, don't run the EFA installation script, and
contact AWS Support.
5. Extract the files from the compressed .tar.gz file and navigate into the extracted directory.

$ tar -xf aws-efa-installer-1.11.1.tar.gz

$ cd aws-efa-installer

6. Run the EFA software installation script.

$ sudo ./efa_installer.sh -y -g

Libfabric is installed in the /opt/amazon/efa directory, while Open MPI is installed in the /opt/
amazon/openmpi directory.
7. Reboot the instance and then log back in.
8. Confirm that the EFA software components were successfully installed.

$ fi_info -p efa -t FI_EP_RDM

The command should return information about the Libfabric EFA interfaces. The following example
shows the command output.

• p3dn.24xlarge with single network interface

provider: efa
fabric: EFA-fe80::94:3dff:fe89:1b70
domain: efa_0-rdm
version: 2.0
type: FI_EP_RDM
protocol: FI_PROTO_EFA

• p4d.24xlarge with multiple network interfaces

provider: efa
fabric: EFA-fe80::c6e:8fff:fef6:e7ff
domain: efa_0-rdm
version: 111.0
type: FI_EP_RDM
protocol: FI_PROTO_EFA
provider: efa
fabric: EFA-fe80::c34:3eff:feb2:3c35
domain: efa_1-rdm
version: 111.0
type: FI_EP_RDM
protocol: FI_PROTO_EFA
provider: efa
fabric: EFA-fe80::c0f:7bff:fe68:a775
domain: efa_2-rdm
version: 111.0
type: FI_EP_RDM
protocol: FI_PROTO_EFA
provider: efa

912
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

fabric: EFA-fe80::ca7:b0ff:fea6:5e99
domain: efa_3-rdm
version: 111.0
type: FI_EP_RDM
protocol: FI_PROTO_EFA

Step 5: Install NCCL

Install NCCL. For more information about NCCL, see the NCCL repository.

To install NCCL

1. Navigate to the /opt directory.

$ cd /opt

2. Clone the official NCCL repository to the instance and navigate into the local cloned repository.

$ sudo git clone https://github.com/NVIDIA/nccl.git && cd nccl

3. Build and install NCCL and specify the CUDA installation directory.

$ sudo make -j src.build CUDA_HOME=/usr/local/cuda

Step 6: Install the aws-ofi-nccl plugin

The aws-ofi-nccl plugin maps NCCL's connection-oriented transport APIs to Libfabric's connection-less
reliable interface. This enables you to use Libfabric as a network provider while running NCCL-based
applications. For more information about the aws-ofi-nccl plugin, see the aws-ofi-nccl repository.

To install the aws-ofi-nccl plugin

1. Navigate to your home directory.

$ cd $HOME

2. (Ubuntu only) Install the utilities that are required to install the aws-ofi-nccl plugin. To install the
required utilities, run the following command.

$ sudo apt-get install libtool autoconf -y

3. Clone the aws branch of the official AWS aws-ofi-nccl repository to the instance and navigate into
the local cloned repository.

$ git clone https://github.com/aws/aws-ofi-nccl.git -b aws && cd aws-ofi-nccl

4. To generate the configure script, run the autogen.sh script.

$ ./autogen.sh

5. To generate the make files, run the configure script and specify the MPI, Libfabric, NCCL, and
CUDA installation directories.

$ ./configure --prefix=/opt/aws-ofi-nccl --with-mpi=/opt/amazon/openmpi \

913
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

--with-libfabric=/opt/amazon/efa --with-nccl=/opt/nccl/build \
--with-cuda=/usr/local/cuda

6. Add the Open MPI directory to the PATH variable.

$ export PATH=/opt/amazon/openmpi/bin/:$PATH

7. Install the aws-ofi-nccl plugin.

$ make

$ sudo make install

Step 7: Install the NCCL tests

Install the NCCL tests. The NCCL tests enable you to confirm that NCCL is properly installed and that it is
operating as expected. For more information about the NCCL tests, see the nccl-tests repository.

To install the NCCL tests

1. Navigate to your home directory.

$ cd $HOME

2. Clone the official nccl-tests repository to the instance and navigate into the local cloned repository.

$ git clone https://github.com/NVIDIA/nccl-tests.git

$ cd nccl-tests

3. Add the Libfabric directory to the LD_LIBRARY_PATH variable.

• Amazon Linux, Amazon Linux 2, RHEL , and CentOS

$ export LD_LIBRARY_PATH=/opt/amazon/efa/lib64:$LD_LIBRARY_PATH

• Ubuntu

$ export LD_LIBRARY_PATH=/opt/amazon/efa/lib:$LD_LIBRARY_PATH

4. Install the NCCL tests and specify the MPI, NCCL, and CUDA installation directories.

$ make MPI=1 MPI_HOME=/opt/amazon/openmpi NCCL_HOME=/opt/nccl/build CUDA_HOME=/usr/

local/cuda

Step 8: Test your EFA and NCCL configuration

Run a test to ensure that your temporary instance is properly configured for EFA and NCCL.

To test your EFA and NCCL configuration

1. Create a host file that specifies the hosts on which to run the tests. The following command creates
a host file named my-hosts that includes a reference to the instance itself.

914
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-

aws-ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" –v http://169.254.169.254/latest/
meta-data/local-ipv4 >> my-hosts

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/local-ipv4 >> my-hosts

2. Run the test and specify the host file (--hostfile) and the number of GPUs to use (-n). The
following command runs the all_reduce_perf test on 8 GPUs on the instance itself, and specifies
the following environment variables.

• FI_PROVIDER="efa"—specifies the fabric interface provider. This must be set to "efa".

• FI_EFA_USE_DEVICE_RDMA=1—uses the device's RDMA functionality for one-sided and two-
sided transfer.
• NCCL_DEBUG=INFO—enables detailed debugging output. You can also specify VERSION to print
only the NCCL version at the start of the test, or WARN to receive only error messages.
• NCCL_ALGO=ring—enables ring algorithm for collective operations.

For more information about the NCCL test arguments, see the NCCL Tests README in the official
nccl-tests repository.

$ /opt/amazon/openmpi/bin/mpirun \
-x FI_PROVIDER="efa" \
-x FI_EFA_USE_DEVICE_RDMA=1 \
-x RDMAV_FORK_SAFE=1 \
-x LD_LIBRARY_PATH=/opt/nccl/build/lib:/usr/local/cuda/lib64:/opt/amazon/efa/
lib64:/opt/amazon/openmpi/lib64:/opt/aws-ofi-nccl/lib:$LD_LIBRARY_PATH \
-x NCCL_DEBUG=INFO \
-x NCCL_ALGO=ring \
--hostfile my-hosts -n 8 -N 8 \
--mca pml ^cm --mca btl tcp,self --mca btl_tcp_if_exclude lo,docker0 --bind-to none
\
$HOME/nccl-tests/build/all_reduce_perf -b 8 -e 1G -f 2 -g 1 -c 1 -n 100

3. You can confirm that EFA is active as the underlying provider for NCCL when the NCCL_DEBUG log is
printed.

ip-192-168-2-54:14:14 [0] NCCL INFO NET/OFI Selected Provider is efa*

The following additional information is displayed when using a p4d.24xlarge instance.

ip-192-168-2-54:14:14 [0] NCCL INFO NET/OFI Running on P4d platform, Setting

NCCL_TOPO_FILE environment variable to /home/ec2-user/install/plugin/share/aws-ofi-
nccl/xml/p4d-24xl-topo.xml

Step 9: Install your machine learning applications

Install the machine learning applications on the temporary instance. The installation procedure varies
depending on the specific machine learning application. For more information about installing software
on your Linux instance, see Managing Software on Your Linux Instance.

915
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

Note
You might need to refer to your machine learning application’s documentation for installation
instructions.

Step 10: Create an EFA and NCCL-enabled AMI

After you have installed the required software components, you create an AMI that you can reuse to
launch your EFA-enabled instances.

To create an AMI from your temporary instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the temporary instance that you created and choose Actions, Image, Create image.
4. For Create image, do the following:

a. For Image name, enter a descriptive name for the AMI.

b. (Optional) For Image description, enter a brief description of the purpose of the AMI.
c. Choose Create image.
5. In the navigation pane, choose AMIs.
6. Locate the AMI tht you created in the list. Wait for the status to change from pending to
available before continuing to the next step.

Step 11: Terminate the temporary instance

At this point, you no longer need the temporary instance that you launched. You can terminate the
instance to stop incurring charges for it.

To terminate the temporary instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the temporary instance that you created and then choose Actions, Instance state, Terminate
instance.
4. When prompted for confirmation, choose Terminate.

Step 12: Launch EFA and NCCL-enabled instances into a cluster placement group
Launch your EFA and NCCL-enabled instances into a cluster placement group using the EFA-enabled AMI
and the EFA-enabled security group that you created earlier.

To launch your EFA and NCCL-enabled instances into a cluster placement group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Launch Instance.
3. On the Choose an AMI page, choose My AMIs, find the AMI that you created earlier, and then choose
Select.
4. On the Choose an Instance Type page, select p3dn.24xlarge and then choose Next: Configure
Instance Details.
5. On the Configure Instance Details page, do the following:

a. For Number of instances, enter the number of EFA and NCCL-enabled instances that you want
to launch.

916
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

b. For Network and Subnet, select the VPC and subnet into which to launch the instances.
c. For Placement group, select Add instance to placement group.
d. For Placement group name, select Add to a new placement group, and then enter a
descriptive name for the placement group. Then for Placement group strategy, select cluster.
e. For EFA, choose Enable.
f. In the Network Interfaces section, for device eth0, choose New network interface. You can
optionally specify a primary IPv4 address and one or more secondary IPv4 addresses. If you are
launching the instance into a subnet that has an associated IPv6 CIDR block, you can optionally
specify a primary IPv6 address and one or more secondary IPv6 addresses.
g. Choose Next: Add Storage.
6. On the Add Storage page, specify the volumes to attach to the instances in addition to the volumes
specified by the AMI (such as the root device volume). Then choose Next: Add Tags.
7. On the Add Tags page, specify tags for the instances, such as a user-friendly name, and then choose
Next: Configure Security Group.
8. On the Configure Security Group page, for Assign a security group, select Select an existing
security group, and then select the security group that you created earlier.
9. Choose Review and Launch.
10. On the Review Instance Launch page, review the settings, and then choose Launch to choose a key
pair and to launch your instances.

Step 13: Enable passwordless SSH

To enable your applications to run across all of the instances in your cluster, you must enable
passwordless SSH access from the leader node to the member nodes. The leader node is the instance
from which you run your applications. The remaining instances in the cluster are the member nodes.

To enable passwordless SSH between the instances in the cluster

1. Select one instance in the cluster as the leader node, and connect to it.
2. Disable strictHostKeyChecking and enable ForwardAgent on the leader node. Open ~/.ssh/
config using your preferred text editor and add the following.

Host *
ForwardAgent yes
Host *
StrictHostKeyChecking no

3. Generate an RSA key pair.

$ ssh-keygen -t rsa -N "" -f ~/.ssh/id_rsa

The key pair is created in the $HOME/.ssh/ directory.

4. Change the permissions of the private key on the leader node.

$ chmod 600 ~/.ssh/id_rsa

chmod 600 ~/.ssh/config

5. Open ~/.ssh/id_rsa.pub using your preferred text editor and copy the key.
6. For each member node in the cluster, do the following:

a. Connect to the instance.

b. Open ~/.ssh/authorized_keys using your preferred text editor and add the public key that
you copied earlier.

917
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

7. To test that the passwordless SSH is functioning as expected, connect to your leader node and run
the following command.

$ ssh member_node_private_ip

You should connect to the member node without being prompted for a key or password.

Use an AWS Deep Learning AMI

The following steps help you to get started with one of the following AWS Deep Learning AMIs:

• Deep Learning AMI (Amazon Linux 2) Version 25.0 and later

• Deep Learning AMI (Amazon Linux) Version 25.0 and later
• Deep Learning AMI (Ubuntu 18.04) Version 25.0 and later
• Deep Learning AMI (Ubuntu 16.04) Version 25.0 and later

For more information, see the AWS Deep Learning AMI User Guide.
Note
Only the p3dn.24xlarge and p4d.24xlarge instance types are supported.

Contents
• Step 1: Prepare an EFA-enabled security group (p. 918)
• Step 2: Launch a temporary instance (p. 919)
• Step 3: Test your EFA and NCCL configuration (p. 919)
• Step 4: Install your machine learning applications (p. 920)
• Step 5: Create an EFA and NCCL-enabled AMI (p. 921)
• Step 6: Terminate the temporary instance (p. 921)
• Step 7: Launch EFA and NCCL-enabled instances into a cluster placement group (p. 921)
• Step 8: Enable passwordless SSH (p. 922)

Step 1: Prepare an EFA-enabled security group

An EFA requires a security group that allows all inbound and outbound traffic to and from the security
group itself. The following procedure allows all inbound and outbound traffic for testing purposes only.
For other scenarios, see Security group rules for different use cases (p. 1072).

To create an EFA-enabled security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups and then choose Create Security Group.
3. In the Create Security Group window, do the following:

a. For Security group name, enter a descriptive name for the security group, such as EFA-
enabled security group.
b. (Optional) For Description, enter a brief description of the security group.
c. For VPC, select the VPC into which you intend to launch your EFA-enabled instances.
d. Choose Create.
4. Select the security group that you created, and on the Description tab, copy the Group ID.
5. On the Inbound tab, do the following:

918
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

a. Choose Edit.
b. For Type, choose All traffic.
c. For Source, choose Custom and paste the security group ID that you copied into the field.
d. Choose Save.
6. On the Outbound tab, do the following:

a. Choose Edit.
b. For Type, choose All traffic.
c. For Destination, choose Custom and paste the security group ID that you copied into the field.
d. Choose Save.

Step 2: Launch a temporary instance

Launch a temporary instance that you can use to install and configure the EFA software components. You
use this instance to create an EFA-enabled AMI from which you can launch your EFA-enabled instances.

To launch a temporary instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Launch Instance.
3. On the Choose an AMI page, choose a supported AWS Deep Learning AMI Version 25.0 or later.
4. On the Choose an Instance Type page, select p3dn.24xlarge or p4d.24xlarge and then choose
Next: Configure Instance Details.
5. On the Configure Instance Details page, do the following:

a. For Subnet, choose the subnet in which to launch the instance.

b. For Elastic Fabric Adapter, choose Enable.
c. In the Network Interfaces section, for device eth0, choose New network interface.
d. Choose Next: Add Storage.
6. On the Add Storage page, specify the volumes to attach to the instances, in addition to the volumes
specified by the AMI (such as the root device volume). Then choose Next: Add Tags.
7. On the Add Tags page, specify a tag that you can use to identify the temporary instance, and then
choose Next: Configure Security Group.
8. On the Configure Security Group page, for Assign a security group, select Select an existing
security group. Then select the security group that you created in Step 1.
9. On the Review Instance Launch page, review the settings, and then choose Launch to choose a key
pair and to launch your instance.

Step 3: Test your EFA and NCCL configuration

Run a test to ensure that your temporary instance is properly configured for EFA and NCCL.

To test your EFA and NCCL configuration

1. Create a host file that specifies the hosts on which to run the tests. The following command creates
a host file named my-hosts that includes a reference to the instance itself.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-

aws-ec2-metadata-token-ttl-seconds: 21600"` \

919
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

&& curl -H "X-aws-ec2-metadata-token: $TOKEN" –v http://169.254.169.254/latest/

meta-data/local-ipv4 >> my-hosts

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/local-ipv4 >> my-hosts

2. Run the test and specify the host file (--hostfile) and the number of GPUs to use (-n). The
following command runs the all_reduce_perf test on 8 GPUs on the instance itself, and specifies
the following environment variables.

• FI_PROVIDER="efa"—specifies the fabric interface provider. This must be set to "efa".

• FI_EFA_USE_DEVICE_RDMA=1—uses the device's RDMA functionality for one-sided and two-
sided transfer.
• NCCL_DEBUG=INFO—enables detailed debugging output. You can also specify VERSION to print
only the NCCL version at the start of the test, or WARN to receive only error messages.
• NCCL_ALGO=ring—enables ring algorithm for collective operations.

For more information about the NCCL test arguments, see the NCCL Tests README in the official
nccl-tests repository.

$ /opt/amazon/openmpi/bin/mpirun \
-x FI_PROVIDER="efa" \
-x FI_EFA_USE_DEVICE_RDMA=1 \
-x RDMAV_FORK_SAFE=1 \
-x LD_LIBRARY_PATH=/opt/nccl/build/lib:/usr/local/cuda/lib64:/opt/amazon/efa/
lib64:/opt/amazon/openmpi/lib64:/opt/aws-ofi-nccl/lib:$LD_LIBRARY_PATH \
-x NCCL_DEBUG=INFO \
-x NCCL_ALGO=ring \
--hostfile my-hosts -n 8 -N 8 \
--mca pml ^cm --mca btl tcp,self --mca btl_tcp_if_exclude lo,docker0 --bind-to none
\
$HOME/nccl-tests/build/all_reduce_perf -b 8 -e 1G -f 2 -g 1 -c 1 -n 100

3. You can confirm that EFA is active as the underlying provider for NCCL when the NCCL_DEBUG log is
printed.

ip-192-168-2-54:14:14 [0] NCCL INFO NET/OFI Selected Provider is efa*

The following additional information is displayed when using a p4d.24xlarge instance.

ip-192-168-2-54:14:14 [0] NCCL INFO NET/OFI Running on P4d platform, Setting

NCCL_TOPO_FILE environment variable to /home/ec2-user/install/plugin/share/aws-ofi-
nccl/xml/p4d-24xl-topo.xml

Step 4: Install your machine learning applications

Install the machine learning applications on the temporary instance. The installation procedure varies
depending on the specific machine learning application. For more information about installing software
on your Linux instance, see Managing Software on Your Linux Instance.
Note
You might need to refer to your machine learning application’s documentation for installation
instructions.

920
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

Step 5: Create an EFA and NCCL-enabled AMI

After you have installed the required software components, you create an AMI that you can reuse to
launch your EFA-enabled instances.

To create an AMI from your temporary instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the temporary instance that you created and choose Actions, Image, Create image.
4. For Create image, do the following:

a. For Image name, enter a descriptive name for the AMI.

b. (Optional) For Image description, enter a brief description of the purpose of the AMI.
c. Choose Create image.
5. In the navigation pane, choose AMIs.
6. Locate the AMI tht you created in the list. Wait for the status to change from pending to
available before continuing to the next step.

Step 6: Terminate the temporary instance

At this point, you no longer need the temporary instance that you launched. You can terminate the
instance to stop incurring charges for it.

To terminate the temporary instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the temporary instance that you created and then choose Actions, Instance state, Terminate
instance.
4. When prompted for confirmation, choose Terminate.

Step 7: Launch EFA and NCCL-enabled instances into a cluster placement group
Launch your EFA and NCCL-enabled instances into a cluster placement group using the EFA-enabled AMI
and the EFA-enabled security group that you created earlier.

To launch your EFA and NCCL-enabled instances into a cluster placement group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Launch Instance.
3. On the Choose an AMI page, choose My AMIs, find the AMI that you created earlier, and then choose
Select.
4. On the Choose an Instance Type page, select p3dn.24xlarge and then choose Next: Configure
Instance Details.
5. On the Configure Instance Details page, do the following:

a. For Number of instances, enter the number of EFA and NCCL-enabled instances that you want
to launch.
b. For Network and Subnet, select the VPC and subnet into which to launch the instances.
c. For Placement group, select Add instance to placement group.
d. For Placement group name, select Add to a new placement group, and then enter a
descriptive name for the placement group. Then for Placement group strategy, select cluster.

921
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Get started with EFA and NCCL

e. For EFA, choose Enable.

f. In the Network Interfaces section, for device eth0, choose New network interface. You can
optionally specify a primary IPv4 address and one or more secondary IPv4 addresses. If you are
launching the instance into a subnet that has an associated IPv6 CIDR block, you can optionally
specify a primary IPv6 address and one or more secondary IPv6 addresses.
g. Choose Next: Add Storage.
6. On the Add Storage page, specify the volumes to attach to the instances in addition to the volumes
specified by the AMI (such as the root device volume). Then choose Next: Add Tags.
7. On the Add Tags page, specify tags for the instances, such as a user-friendly name, and then choose
Next: Configure Security Group.
8. On the Configure Security Group page, for Assign a security group, select Select an existing
security group, and then select the security group that you created earlier.
9. Choose Review and Launch.
10. On the Review Instance Launch page, review the settings, and then choose Launch to choose a key
pair and to launch your instances.

Step 8: Enable passwordless SSH

To enable your applications to run across all of the instances in your cluster, you must enable
passwordless SSH access from the leader node to the member nodes. The leader node is the instance
from which you run your applications. The remaining instances in the cluster are the member nodes.

To enable passwordless SSH between the instances in the cluster

1. Select one instance in the cluster as the leader node, and connect to it.
2. Disable strictHostKeyChecking and enable ForwardAgent on the leader node. Open ~/.ssh/
config using your preferred text editor and add the following.

Host *
ForwardAgent yes
Host *
StrictHostKeyChecking no

3. Generate an RSA key pair.

$ ssh-keygen -t rsa -N "" -f ~/.ssh/id_rsa

The key pair is created in the $HOME/.ssh/ directory.

4. Change the permissions of the private key on the leader node.

$ chmod 600 ~/.ssh/id_rsa

chmod 600 ~/.ssh/config

5. Open ~/.ssh/id_rsa.pub using your preferred text editor and copy the key.
6. For each member node in the cluster, do the following:

a. Connect to the instance.

b. Open ~/.ssh/authorized_keys using your preferred text editor and add the public key that
you copied earlier.
7. To test that the passwordless SSH is functioning as expected, connect to your leader node and run
the following command.

$ ssh member_node_private_ip

922
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with EFA

You should connect to the member node without being prompted for a key or password.

Work with EFA

You can create, use, and manage an EFA much like any other elastic network interface in Amazon EC2.
However, unlike elastic network interfaces, EFAs cannot be attached to or detached from an instance in a
running state.

EFA requirements
To use an EFA, you must do the following:

• Choose one of the supported instance types (p. 897).

• Use one of the supported AMIs (p. 898).
• Install the EFA software components. For more information, see Step 3: Install the EFA
software (p. 900) and Step 5: (Optional) Install Intel MPI (p. 903).
• Use a security group that allows all inbound and outbound traffic to and from the security group itself.
For more information, see Step 1: Prepare an EFA-enabled security group (p. 899).

Contents
• Create an EFA (p. 923)
• Attach an EFA to a stopped instance (p. 924)
• Attach an EFA when launching an instance (p. 924)
• Add an EFA to a launch template (p. 924)
• Manage IP addresses for an EFA (p. 924)
• Change the security group for an EFA (p. 925)
• Detach an EFA (p. 925)
• View EFAs (p. 925)
• Delete an EFA (p. 925)

Create an EFA
You can create an EFA in a subnet in a VPC. You can't move the EFA to another subnet after it's created,
and you can only attach it to stopped instances in the same Availability Zone.

To create a new EFA using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Choose Create Network Interface.
4. For Description, enter a descriptive name for the EFA.
5. For Subnet, select the subnet in which to create the EFA.
6. For Private IP, enter the primary private IPv4 address. If you don't specify an IPv4 address, we select
an available private IPv4 address from the selected subnet.
7. (IPv6 only) If you selected a subnet that has an associated IPv6 CIDR block, you can optionally
specify an IPv6 address in the IPv6 IP field.
8. For Security groups, select one or more security groups.
9. For EFA, choose Enabled.

923
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with EFA

10. Choose Yes, Create.

To create a new EFA using the AWS CLI

Use the create-network-interface command and for interface-type, specify efa, as shown in the
following example.

aws ec2 create-network-interface --subnet-id subnet-01234567890 --description example_efa

--interface-type efa

Attach an EFA to a stopped instance

You can attach an EFA to any supported instance that is in the stopped state. You cannot attach an EFA
to an instance that is in the running state. For more information about the supported instance types,
see Supported instance types (p. 897).

You attach an EFA to an instance in the same way that you attach a network interface to an instance. For
more information, see Attach a network interface to an instance (p. 855).

Attach an EFA when launching an instance

To attach an existing EFA when launching an instance (AWS CLI)

Use the run-instances command and for NetworkInterfaceId, specify the ID of the EFA, as shown in the
following example.

aws ec2 run-instances --image-id ami_id --count 1 --instance-

type c5n.18xlarge --key-name my_key_pair --network-interfaces
DeviceIndex=0,NetworkInterfaceId=efa_id,Groups=sg_id,SubnetId=subnet_id

To attach a new EFA when launching an instance (AWS CLI)

Use the run-instances command and for InterfaceType, specify efa, as shown in the following example.

aws ec2 run-instances --image-id ami_id --count 1 --instance-

type c5n.18xlarge --key-name my_key_pair --network-interfaces
DeviceIndex=0,InterfaceType=efa,Groups=sg_id,SubnetId=subnet_id

Add an EFA to a launch template

You can create a launch template that contains the configuration information needed to launch EFA-
enabled instances. To create an EFA-enabled launch template, create a new launch template and
specify a supported instance type, your EFA-enabled AMI, and an EFA-enabled security group. For more
information, see Get started with EFA and MPI (p. 898).

You can leverage launch templates to launch EFA-enabled instances with other AWS services, such as
AWS Batch.

For more information about creating launch templates, see Create a launch template (p. 534).

Manage IP addresses for an EFA

You can change the IP addresses associated with an EFA. If you have an Elastic IP address, you can
associate it with an EFA. If your EFA is provisioned in a subnet that has an associated IPv6 CIDR block, you
can assign one or more IPv6 addresses to the EFA.

924
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Monitor an EFA

You assign an Elastic IP (IPv4) and IPv6 address to an EFA in the same way that you assign an IP address
to an elastic network interface. For more information, see Managing IP addresses (p. 857).

Change the security group for an EFA

You can change the security group that is associated with an EFA. To enable OS-bypass functionality, the
EFA must be a member of a security group that allows all inbound and outbound traffic to and from the
security group itself.

You change the security group that is associated with an EFA in the same way that you change the
security group that is associated with an elastic network interface. For more information, see Changing
the security group (p. 859).

Detach an EFA
To detach an EFA from an instance, you must first stop the instance. You cannot detach an EFA from an
instance that is in the running state.

You detach an EFA from an instance in the same way that you detach an elastic network interface from
an instance. For more information, see Detach a network interface from an instance (p. 856).

View EFAs
You can view all of the EFAs in your account.

You view EFAs in the same way that you view elastic network interfaces. For more information, see View
details about a network interface (p. 855).

Delete an EFA
To delete an EFA, you must first detach it from the instance. You cannot delete an EFA while it is
attached to an instance.

You delete EFAs in the same way that you delete elastic network interfaces. For more information, see
Delete a network interface (p. 860).

Monitor an EFA
You can use the following features to monitor the performance of your Elastic Fabric Adapters.

Amazon VPC flow logs

You can create an Amazon VPC Flow Log to capture information about the traffic going to and from an
EFA. Flow log data can be published to Amazon CloudWatch Logs and Amazon S3. After you create a
flow log, you can retrieve and view its data in the chosen destination. For more information, see VPC
Flow Logs in the Amazon VPC User Guide.

You create a flow log for an EFA in the same way that you create a flow log for an elastic network
interface. For more information, see Creating a Flow Log in the Amazon VPC User Guide.

In the flow log entries, EFA traffic is identified by the srcAddress and destAddress, which are both
formatted as MAC addresses, as shown in the following example.

version accountId eniId srcAddress destAddress sourcePort destPort

protocol packets bytes start end action log-status

925
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Verify the EFA installer using a checksum

2 3794735123 eni-10000001 01:23:45:67:89:ab 05:23:45:67:89:ab - - -

9 5689 1521232534 1524512343 ACCEPT OK

Amazon CloudWatch
Amazon CloudWatch provides metrics that enable you to monitor your EFAs in real time. You can collect
and track metrics, create customized dashboards, and set alarms that notify you or take actions when a
specified metric reaches a threshold that you specify. For more information, see Monitor your instances
using CloudWatch (p. 761).

Verify the EFA installer using a checksum

You can optionally verify the EFA tarball (.tar.gz file) using an MD5 or SHA256 checksum. We recommend
that you do this to verify the identity of the software publisher and to check that the application has not
been altered or corrupted since it was published.

To verify the tarball

Use the md5sum utility for the MD5 checksum, or the sha256sum utility for the SHA256 checksum,
and specify the tarball filename. You must run the command from the directory in which you saved the
tarball file.

• MD5

$ md5sum tarball_filename.tar.gz

• SHA256

$ sha256sum tarball_filename.tar.gz

The commands should return a checksum value in the following format.

checksum_value tarball_filename.tar.gz

Compare the checksum value returned by the command with the checksum value provided in the table
below. If the checksums match, then it is safe to run the installation script. If the checksums do not
match, do not run the installation script, and contact AWS Support.

For example, the following command verifies the EFA 1.9.4 tarball using the SHA256 checksum.

$ sha256sum aws-efa-installer-1.9.4.tar.gz

1009b5182693490d908ef0ed2c1dd4f813cc310a5d2062ce9619c4c12b5a7f14 aws-efa-
installer-1.9.4.tar.gz

The following table lists the checksums for recent versions of EFA.

Version Download URL Checksums

EFA 1.11.1 https://efa-installer.amazonaws.com/ MD5:

aws-efa-installer-1.11.1.tar.gz 026b0d9a0a48780cc7406bd51997b1c0

SHA256:
6cb04baf5ffc58ddf319e956b5461289199c8dd80

926
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Placement groups

Version Download URL Checksums

EFA 1.11.0 https://efa-installer.amazonaws.com/ MD5:

aws-efa-installer-1.11.0.tar.gz 7d9058e010ad65bf2e14259214a36949

SHA256:
7891f6d45ae33e822189511c4ea1d14c9d54d000f

EFA 1.10.1 https://efa-installer.amazonaws.com/ MD5:

aws-efa-installer-1.10.1.tar.gz 78521d3d668be22976f46c6fecc7b730

SHA256:
61564582de7320b21de319f532c3a677d26cc4678

EFA 1.10.0 https://efa-installer.amazonaws.com/ MD5:

aws-efa-installer-1.10.0.tar.gz 46f73f5a7afe41b4bb918c81888fefa9

SHA256:
136612f96f2a085a7d98296da0afb6fa807b38142

EFA 1.9.5 https://efa-installer.amazonaws.com/ MD5:

aws-efa-installer-1.9.5.tar.gz 95edb8a209c18ba8d250409846eb6ef4

SHA256:
a4343308d7ea4dc943ccc21bcebed913e8868e59b

EFA 1.9.4 https://efa-installer.amazonaws.com/ MD5:

aws-efa-installer-1.9.4.tar.gz f26dd5c350422c1a985e35947fa5aa28

SHA256:
1009b5182693490d908ef0ed2c1dd4f813cc310a5

EFA 1.9.3 https://efa-installer.amazonaws.com/ MD5:

aws-efa-installer-1.9.3.tar.gz 95755765a097802d3e6d5018d1a5d3d6

SHA256:
46ce732d6f3fcc9edf6a6e9f9df0ad136054328e2

EFA 1.8.4 https://efa-installer.amazonaws.com/ MD5:

aws-efa-installer-1.8.4.tar.gz 85d594c41e831afc6c9305263140457e

SHA256:
0d974655a09b213d7859e658965e56dc4f23a0eee

Placement groups
When you launch a new EC2 instance, the EC2 service attempts to place the instance in such a way that
all of your instances are spread out across underlying hardware to minimize correlated failures. You can
use placement groups to influence the placement of a group of interdependent instances to meet the
needs of your workload. Depending on the type of workload, you can create a placement group using
one of the following placement strategies:

• Cluster – packs instances close together inside an Availability Zone. This strategy enables workloads
to achieve the low-latency network performance necessary for tightly-coupled node-to-node
communication that is typical of HPC applications.
• Partition – spreads your instances across logical partitions such that groups of instances in one
partition do not share the underlying hardware with groups of instances in different partitions. This

927
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Cluster placement groups

strategy is typically used by large distributed and replicated workloads, such as Hadoop, Cassandra,
and Kafka.
• Spread – strictly places a small group of instances across distinct underlying hardware to reduce
correlated failures.

There is no charge for creating a placement group.

Contents
• Cluster placement groups (p. 928)
• Partition placement groups (p. 929)
• Spread placement groups (p. 930)
• Placement group rules and limitations (p. 930)
• Create a placement group (p. 931)
• Tag a placement group (p. 932)
• Launch instances in a placement group (p. 935)
• Describe instances in a placement group (p. 936)
• Change the placement group for an instance (p. 937)
• Delete a placement group (p. 938)

Cluster placement groups

A cluster placement group is a logical grouping of instances within a single Availability Zone. A cluster
placement group can span peered VPCs in the same Region. Instances in the same cluster placement
group enjoy a higher per-flow throughput limit for TCP/IP traffic and are placed in the same high-
bisection bandwidth segment of the network.

The following image shows instances that are placed into a cluster placement group.

Cluster placement groups are recommended for applications that benefit from low network latency,
high network throughput, or both. They are also recommended when the majority of the network
traffic is between the instances in the group. To provide the lowest latency and the highest packet-per-
second network performance for your placement group, choose an instance type that supports enhanced
networking. For more information, see Enhanced Networking (p. 864).

928
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Partition placement groups

We recommend that you launch your instances in the following way:

• Use a single launch request to launch the number of instances that you need in the placement group.
• Use the same instance type for all instances in the placement group.

If you try to add more instances to the placement group later, or if you try to launch more than one
instance type in the placement group, you increase your chances of getting an insufficient capacity error.

If you stop an instance in a placement group and then start it again, it still runs in the placement group.
However, the start fails if there isn't enough capacity for the instance.

If you receive a capacity error when launching an instance in a placement group that already has running
instances, stop and start all of the instances in the placement group, and try the launch again. Starting
the instances may migrate them to hardware that has capacity for all of the requested instances.

Partition placement groups

Partition placement groups help reduce the likelihood of correlated hardware failures for your
application. When using partition placement groups, Amazon EC2 divides each group into logical
segments called partitions. Amazon EC2 ensures that each partition within a placement group has
its own set of racks. Each rack has its own network and power source. No two partitions within a
placement group share the same racks, allowing you to isolate the impact of hardware failure within your
application.

The following image is a simple visual representation of a partition placement group in a single
Availability Zone. It shows instances that are placed into a partition placement group with three
partitions—Partition 1, Partition 2, and Partition 3. Each partition comprises multiple instances. The
instances in a partition do not share racks with the instances in the other partitions, allowing you to
contain the impact of a single hardware failure to only the associated partition.

Partition placement groups can be used to deploy large distributed and replicated workloads, such as
HDFS, HBase, and Cassandra, across distinct racks. When you launch instances into a partition placement
group, Amazon EC2 tries to distribute the instances evenly across the number of partitions that you
specify. You can also launch instances into a specific partition to have more control over where the
instances are placed.

A partition placement group can have partitions in multiple Availability Zones in the same Region. A
partition placement group can have a maximum of seven partitions per Availability Zone. The number
of instances that can be launched into a partition placement group is limited only by the limits of your
account.

In addition, partition placement groups offer visibility into the partitions — you can see which instances
are in which partitions. You can share this information with topology-aware applications, such as HDFS,
HBase, and Cassandra. These applications use this information to make intelligent data replication
decisions for increasing data availability and durability.

929
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Spread placement groups

If you start or launch an instance in a partition placement group and there is insufficient unique
hardware to fulfill the request, the request fails. Amazon EC2 makes more distinct hardware available
over time, so you can try your request again later.

Spread placement groups

A spread placement group is a group of instances that are each placed on distinct racks, with each rack
having its own network and power source.

The following image shows seven instances in a single Availability Zone that are placed into a spread
placement group. The seven instances are placed on seven different racks.

Spread placement groups are recommended for applications that have a small number of critical
instances that should be kept separate from each other. Launching instances in a spread placement
group reduces the risk of simultaneous failures that might occur when instances share the same racks.
Spread placement groups provide access to distinct racks, and are therefore suitable for mixing instance
types or launching instances over time.

A spread placement group can span multiple Availability Zones in the same Region. You can have a
maximum of seven running instances per Availability Zone per group.

If you start or launch an instance in a spread placement group and there is insufficient unique hardware
to fulfill the request, the request fails. Amazon EC2 makes more distinct hardware available over time, so
you can try your request again later.

Placement group rules and limitations

General rules and limitations
Before you use placement groups, be aware of the following rules:

• The name that you specify for a placement group must be unique within your AWS account for the
Region.
• You can't merge placement groups.
• An instance can be launched in one placement group at a time; it cannot span multiple placement
groups.
• On-Demand Capacity Reservation (p. 503) and zonal Reserved Instances (p. 328) provide a capacity
reservation for EC2 instances in a specific Availability Zone. The capacity reservation can be used
by instances in a placement group. However, it is not possible to explicitly reserve capacity for a
placement group.
• You cannot launch Dedicated Hosts in placement groups.

Cluster placement group rules and limitations

The following rules apply to cluster placement groups:

• Instances in a cluster placement group you must use the following supported instance types:

930
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create a placement group

• Current generation (p. 204) instances, except for burstable performance (p. 224) instances (for
example, T2) and Mac1 instances (p. 258).
• The following previous generation (p. 207) instances: A1, C3, cc2.8xlarge, cr1.8xlarge, G2,
hs1.8xlarge, I2, and R3.
• A cluster placement group can't span multiple Availability Zones.
• The maximum network throughput speed of traffic between two instances in a cluster placement
group is limited by the slower of the two instances. For applications with high-throughput
requirements, choose an instance type with network connectivity that meets your requirements.
• For instances that are enabled for enhanced networking, the following rules apply:
• Instances within a cluster placement group can use up to 10 Gbps for single-flow traffic. Instances
that are not within a cluster placement group can use up to 5 Gbps for single-flow traffic.
• Traffic to and from Amazon S3 buckets within the same Region over the public IP address space or
through a VPC endpoint can use all available instance aggregate bandwidth.
• You can launch multiple instance types into a cluster placement group. However, this reduces the
likelihood that the required capacity will be available for your launch to succeed. We recommend using
the same instance type for all instances in a cluster placement group.
• Network traffic to the internet and over an AWS Direct Connect connection to on-premises resources is
limited to 5 Gbps.

Partition placement group rules and limitations

The following rules apply to partition placement groups:

• A partition placement group supports a maximum of seven partitions per Availability Zone. The
number of instances that you can launch in a partition placement group is limited only by your account
limits.
• When instances are launched into a partition placement group, Amazon EC2 tries to evenly distribute
the instances across all partitions. Amazon EC2 doesn’t guarantee an even distribution of instances
across all partitions.
• A partition placement group with Dedicated Instances can have a maximum of two partitions.

Spread placement group rules and limitations

The following rules apply to spread placement groups:

• A spread placement group supports a maximum of seven running instances per Availability Zone. For
example, in a Region with three Availability Zones, you can run a total of 21 instances in the group
(seven per zone). If you try to start an eighth instance in the same Availability Zone and in the same
spread placement group, the instance will not launch. If you need to have more than seven instances
in an Availability Zone, then the recommendation is to use multiple spread placement groups. Using
multiple spread placement groups does not provide guarantees about the spread of instances between
groups, but it does ensure the spread for each group, thus limiting impact from certain classes of
failures.
• Spread placement groups are not supported for Dedicated Instances.

Create a placement group

You can create a placement group using one of the following methods.
Note
You can tag a placement group on creation using the command line tools only.

931
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tag a placement group

New console

To create a placement group using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Placement Groups, Create placement group.
3. Specify a name for the group.
4. Choose the placement strategy for the group. If you choose Partition, choose the number of
partitions within the group.
5. Choose Create group.

Old console

To create a placement group using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Placement Groups, Create Placement Group.
3. Specify a name for the group.
4. Choose the placement strategy for the group. If you choose Partition, specify the number of
partitions within the group.
5. Choose Create.

AWS CLI

To create a placement group using the AWS CLI

Use the create-placement-group command. The following example creates a placement group
named my-cluster that uses the cluster placement strategy, and it applies a tag with a key of
purpose and a value of production.

aws ec2 create-placement-group --group-name my-cluster --strategy cluster --tag-

specifications 'ResourceType=placement-group,Tags={Key=purpose,Value=production}'

To create a partition placement group using the AWS CLI

Use the create-placement-group command. Specify the --strategy parameter with the value
partition, and specify the --partition-count parameter with the desired number of
partitions. In this example, the partition placement group is named HDFS-Group-A and is created
with five partitions.

aws ec2 create-placement-group --group-name HDFS-Group-A --strategy partition --

partition-count 5

PowerShell

To create a placement group using the AWS Tools for Windows PowerShell

Use the New-EC2PlacementGroup command.

Tag a placement group

To help categorize and manage your existing placement groups, you can tag them with custom
metadata. For more information about how tags work, see Tag your Amazon EC2 resources (p. 1311).

932
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tag a placement group

When you tag a placement group, the instances that are launched into the placement group are not
automatically tagged. You need to explicitly tag the instances that are launched into the placement
group. For more information, see Add a tag when you launch an instance (p. 1319).

You can view, add, and delete tags using the new console and the command line tools.

New console

To view, add, or delete a tag for an existing placement group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Placement Groups.
3. Select a placement group, and then choose Actions, Manage tags.
4. The Manage tags section displays any tags that are assigned to the placement group. Do the
following to add or remove tags:

• To add a tag, choose Add tag, and then enter the tag key and value. You can add up to 50
tags per placement group. For more information, see Tag restrictions (p. 1315).
• To delete a tag, choose Remove next to the tag that you want to delete.
5. Choose Save changes.

AWS CLI

To view placement group tags

Use the describe-tags command to view the tags for the specified resource. In the following
example, you describe the tags for all of your placement groups.

aws ec2 describe-tags \

--filters Name=resource-type,Values=placement-group

{
"Tags": [
{
"Key": "Environment",
"ResourceId": "pg-0123456789EXAMPLE",
"ResourceType": "placement-group",
"Value": "Production"
},
{
"Key": "Environment",
"ResourceId": "pg-9876543210EXAMPLE",
"ResourceType": "placement-group",
"Value": "Production"
}
]
}

You can also use the describe-tags command to view the tags for a placement group by specifying
its ID. In the following example, you describe the tags for pg-0123456789EXAMPLE.

aws ec2 describe-tags \

--filters Name=resource-id,Values=pg-0123456789EXAMPLE

{
"Tags": [

933
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tag a placement group

{
"Key": "Environment",
"ResourceId": "pg-0123456789EXAMPLE",
"ResourceType": "placement-group",
"Value": "Production"
}
]
}

You can also view the tags of a placement group by describing the placement group.

Use the describe-placement-groups command to view the configuration of the specified placement
group, which includes any tags that were specified for the placement group.

aws ec2 describe-placement-groups \

--group-name my-cluster

{
"PlacementGroups": [
{
"GroupName": "my-cluster",
"State": "available",
"Strategy": "cluster",
"GroupId": "pg-0123456789EXAMPLE",
"Tags": [
{
"Key": "Environment",
"Value": "Production"
}
]
}
]
}

To tag an existing placement group using the AWS CLI

You can use the create-tags command to tag existing resources. In the following example, the
existing placement group is tagged with Key=Cost-Center and Value=CC-123.

aws ec2 create-tags \

--resources pg-0123456789EXAMPLE \
--tags Key=Cost-Center,Value=CC-123

To delete a tag from a placement group using the AWS CLI

You can use the delete-tags command to delete tags from existing resources. For examples, see
Examples in the AWS CLI Command Reference.
PowerShell

To view placement group tags

Use the Get-EC2Tag command.

To describe the tags for a specific placement group

Use the Get-EC2PlacementGroup command.

To tag an existing placement group

934
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Launch instances in a placement group

Use the New-EC2Tag command.

To delete a tag from a placement group

Use the Remove-EC2Tag command.

Launch instances in a placement group

You can launch an instance into a placement group if the placement group rules and limitations are
met (p. 930) using one of the following methods.

Console

To launch instances into a placement group using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Choose Launch Instance. Complete the wizard as directed, taking care to do the following:

• On the Choose an Instance Type page, select an instance type that can be launched into a
placement group.
• On the Configure Instance Details page, the following fields are applicable to placement
groups:
• For Number of instances, enter the total number of instances that you need in this
placement group, because you might not be able to add instances to the placement group
later.
• For Placement group, select the Add instance to placement group check box. If you do not
see Placement group on this page, verify that you have selected an instance type that can
be launched into a placement group. Otherwise, this option is not available.
• For Placement group name, you can choose to add the instances to an existing placement
group or to a new placement group that you create.
• For Placement group strategy, choose the appropriate strategy. If you choose partition,
for Target partition, choose Auto distribution to have Amazon EC2 do a best effort to
distribute the instances evenly across all the partitions in the group. Alternatively, specify
the partition in which to launch the instances.

AWS CLI

To launch instances into a placement group using the AWS CLI

Use the run-instances command and specify the placement group name using the --placement
"GroupName = my-cluster" parameter. In this example, the placement group is named my-
cluster.

aws ec2 run-instances --placement "GroupName = my-cluster"

To launch instances into a specific partition of a partition placement group using the AWS CLI

Use the run-instances command and specify the placement group name and partition using the
--placement "GroupName = HDFS-Group-A, PartitionNumber = 3" parameter. In this
example, the placement group is named HDFS-Group-A and the partition number is 3.

aws ec2 run-instances --placement "GroupName = HDFS-Group-A, PartitionNumber = 3"

935
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Describe instances in a placement group

PowerShell

To launch instances into a placement group using AWS Tools for Windows PowerShell

Use the New-EC2Instance command and specify the placement group name using the -
Placement_GroupName parameter.

Describe instances in a placement group

You can view the placement information of your instances using one of the following methods. You can
also filter partition placement groups by the partition number using the AWS CLI.

New console

To view the placement group and partition number of an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance.
4. In the Description tab, under Host and placement group, find Placement group. If the
instance is not in a placement group, the field is empty. Otherwise, it contains the name of
the placement group name. If the placement group is a partition placement group, Partition
number contains the partition number for the instance.

Old console

To view the placement group and partition number of an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance.
4. In the Description tab, find Placement group. If the instance is not in a placement group, the
field is empty. Otherwise, it contains the name of the placement group name. If the placement
group is a partition placement group, Partition number contains the partition number for the
instance.

AWS CLI

To view the partition number for an instance in a partition placement group using the AWS CLI

Use the describe-instances command and specify the --instance-id parameter.

aws ec2 describe-instances --instance-id i-0123a456700123456

The response contains the placement information, which includes the placement group name and
the partition number for the instance.

"Placement": {
"AvailabilityZone": "us-east-1c",
"GroupName": "HDFS-Group-A",
"PartitionNumber": 3,
"Tenancy": "default"
}

936
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Change the placement group for an instance

To filter instances for a specific partition placement group and partition number using the AWS
CLI

Use the describe-instances command and specify the --filters parameter with the placement-
group-name and placement-partition-number filters. In this example, the placement group is
named HDFS-Group-A and the partition number is 7.

aws ec2 describe-instances --filters "Name = placement-group-name, Values = HDFS-Group-

A" "Name = placement-partition-number, Values = 7"

The response lists all the instances that are in the specified partition within the specified placement
group. The following is example output showing only the instance ID, instance type, and placement
information for the returned instances.

"Instances": [
{
"InstanceId": "i-0a1bc23d4567e8f90",
"InstanceType": "r4.large",
},

"Placement": {
"AvailabilityZone": "us-east-1c",
"GroupName": "HDFS-Group-A",
"PartitionNumber": 7,
"Tenancy": "default"
}

{
"InstanceId": "i-0a9b876cd5d4ef321",
"InstanceType": "r4.large",
},

"Placement": {
"AvailabilityZone": "us-east-1c",
"GroupName": "HDFS-Group-A",
"PartitionNumber": 7,
"Tenancy": "default"
}
],

Change the placement group for an instance

You can change the placement group for an instance in any of the following ways:

• Move an existing instance to a placement group

• Move an instance from one placement group to another
• Remove an instance from a placement group

Before you move or remove the instance, the instance must be in the stopped state. You can move or
remove an instance using the AWS CLI or an AWS SDK.

AWS CLI

To move an instance to a placement group using the AWS CLI

1. Stop the instance using the stop-instances command.

2. Use the modify-instance-placement command and specify the name of the placement group to
which to move the instance.

937
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Delete a placement group

aws ec2 modify-instance-placement --instance-id i-0123a456700123456 --group-

name MySpreadGroup

3. Start the instance using the start-instances command.

PowerShell

To move an instance to a placement group using the AWS Tools for Windows PowerShell

1. Stop the instance using the Stop-EC2Instance command.

2. Use the Edit-EC2InstancePlacement command and specify the name of the placement group to
which to move the instance.
3. Start the instance using the Start-EC2Instance command.

AWS CLI

To remove an instance from a placement group using the AWS CLI

1. Stop the instance using the stop-instances command.

2. Use the modify-instance-placement command and specify an empty string for the placement
group name.

aws ec2 modify-instance-placement --instance-id i-0123a456700123456 --group-name ""

3. Start the instance using the start-instances command.

PowerShell

To remove an instance from a placement group using the AWS Tools for Windows
PowerShell

1. Stop the instance using the Stop-EC2Instance command.

2. Use the Edit-EC2InstancePlacement command and specify an empty string for the placement
group name.
3. Start the instance using the Start-EC2Instance command.

Delete a placement group

If you need to replace a placement group or no longer need one, you can delete it. You can delete a
placement group using one of the following methods.

Requirement

Before you can delete a placement group, it must contain no instances. You can terminate (p. 644) all
instances that you launched into the placement group, move (p. 937) them to another placement
group, or remove (p. 938) them from the placement group.

New console

To delete a placement group using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

938
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Network MTU

2. In the navigation pane, choose Placement Groups.

3. Select the placement group and choose Actions, Delete.
4. When prompted for confirmation, enter Delete and then choose Delete.

Old console

To delete a placement group using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Placement Groups.
3. Select the placement group and choose Actions, Delete Placement Group.
4. When prompted for confirmation, choose Delete.

AWS CLI

To delete a placement group using the AWS CLI

Use the delete-placement-group command and specify the placement group name to delete the
placement group. In this example, the placement group name is my-cluster.

aws ec2 delete-placement-group --group-name my-cluster

PowerShell

To delete a placement group using the AWS Tools for Windows PowerShell

Use the Remove-EC2PlacementGroup command to delete the placement group.

Network maximum transmission unit (MTU) for

your EC2 instance
The maximum transmission unit (MTU) of a network connection is the size, in bytes, of the largest
permissible packet that can be passed over the connection. The larger the MTU of a connection, the more
data that can be passed in a single packet. Ethernet packets consist of the frame, or the actual data you
are sending, and the network overhead information that surrounds it.

Ethernet frames can come in different formats, and the most common format is the standard Ethernet
v2 frame format. It supports 1500 MTU, which is the largest Ethernet packet size supported over most of
the internet. The maximum supported MTU for an instance depends on its instance type. All Amazon EC2
instance types support 1500 MTU, and many current instance sizes support 9001 MTU, or jumbo frames.

If your instance runs in a Wavelength Zone, the maximum MTU value is 1300.

To see Network MTU information for Windows instances, switch to this page in the Amazon EC2 User
Guide for Windows Instances guide: Network maximum transmission unit (MTU) for your EC2 instance.

Contents
• Jumbo frames (9001 MTU) (p. 940)
• Path MTU Discovery (p. 940)
• Check the path MTU between two hosts (p. 940)

939
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Jumbo frames (9001 MTU)

• Check and set the MTU on your Linux instance (p. 941)
• Troubleshoot (p. 942)

Jumbo frames (9001 MTU)

Jumbo frames allow more than 1500 bytes of data by increasing the payload size per packet, and thus
increasing the percentage of the packet that is not packet overhead. Fewer packets are needed to send
the same amount of usable data. However, outside of a given AWS Region (EC2-Classic), a single VPC,
or a VPC peering connection, you will experience a maximum path of 1500 MTU. VPN connections and
traffic sent over an internet gateway are limited to 1500 MTU. If packets are over 1500 bytes, they are
fragmented, or they are dropped if the Don't Fragment flag is set in the IP header.

Jumbo frames should be used with caution for internet-bound traffic or any traffic that leaves a VPC.
Packets are fragmented by intermediate systems, which slows down this traffic. To use jumbo frames
inside a VPC and not slow traffic that's bound for outside the VPC, you can configure the MTU size by
route, or use multiple elastic network interfaces with different MTU sizes and different routes.

For instances that are collocated inside a cluster placement group, jumbo frames help to achieve the
maximum network throughput possible, and they are recommended in this case. For more information,
see Placement groups (p. 927).

You can use jumbo frames for traffic between your VPCs and your on-premises networks over AWS Direct
Connect. For more information, and for how to verify Jumbo Frame capability, see Setting Network MTU
in the AWS Direct Connect User Guide.

All current generation instances (p. 210) support jumbo frames. The following previous generation
instances support jumbo frames: A1, C3, G2, I2, M3, and R3.

For more information about supported MTU sizes for transit gateways, see MTU in Amazon VPC Transit
Gateways.

Path MTU Discovery

Path MTU Discovery (PMTUD) is used to determine the maximum transmission unit (MTU) of a network
path. Path MTU is the maximum packet size between the originating host and the receiving host. If a
host sends a packet that's larger than the MTU of the receiving host or that's larger than the MTU of a
device along the path, the receiving host or device returns the following ICMP message: Destination
Unreachable: Fragmentation Needed and Don't Fragment was Set (Type 3, Code 4). This
instructs the original host to adjust the MTU until the packet can be transmitted.

By default, security groups do not allow any inbound ICMP traffic. However, security groups are stateful,
therefore ICMP responses to outbound requests are allowed to flow in, regardless of security group
rules. Therefore, you do not need to explicitly add an inbound ICMP rule to ensure that your instance can
receive the ICMP message response. For more information about configuring ICMP rules in a network
ACL, see Path MTU Discovery in the Amazon VPC User Guide.
Important
Path MTU Discovery does not guarantee that jumbo frames will not be dropped by some
routers. An internet gateway in your VPC will forward packets up to 1500 bytes only. 1500 MTU
packets are recommended for internet traffic.

Check the path MTU between two hosts

You can check the path MTU between two hosts using the tracepath command, which is part of the
iputils package that is available by default on many Linux distributions, including Amazon Linux.

940
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Check and set the MTU on your Linux instance

To check path MTU using tracepath

Use the following command to check the path MTU between your EC2 instance and another host. You
can use a DNS name or an IP address as the destination. If the destination is another EC2 instance, verify
that the security group allows inbound UDP traffic. This example checks the path MTU between an EC2
instance and amazon.com.

[ec2-user ~]$ tracepath amazon.com

1?: [LOCALHOST] pmtu 9001
1: ip-172-31-16-1.us-west-1.compute.internal (172.31.16.1) 0.187ms pmtu 1500
1: no reply
2: no reply
3: no reply
4: 100.64.16.241 (100.64.16.241) 0.574ms
5: 72.21.222.221 (72.21.222.221) 84.447ms asymm 21
6: 205.251.229.97 (205.251.229.97) 79.970ms asymm 19
7: 72.21.222.194 (72.21.222.194) 96.546ms asymm 16
8: 72.21.222.239 (72.21.222.239) 79.244ms asymm 15
9: 205.251.225.73 (205.251.225.73) 91.867ms asymm 16
...
31: no reply
Too many hops: pmtu 1500
Resume: pmtu 1500

In this example, the path MTU is 1500.

Check and set the MTU on your Linux instance

Some instances are configured to use jumbo frames, and others are configured to use standard frame
sizes. You may want to use jumbo frames for network traffic within your VPC or you may want to use
standard frames for internet traffic. Whatever your use case, we recommend verifying that your instance
will behave the way you expect it to. You can use the procedures in this section to check your network
interface's MTU setting and modify it if needed.

To check the MTU setting on a Linux instance

You can check the current MTU value using the following ip command. Note that in the example output,
mtu 9001 indicates that this instance uses jumbo frames.

[ec2-user ~]$ ip link show eth0

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9001 qdisc pfifo_fast state UP mode DEFAULT
group default qlen 1000
link/ether 02:90:c0:b7:9e:d1 brd ff:ff:ff:ff:ff:ff

To set the MTU value on a Linux instance

1. You can set the MTU value using the ip command. The following command sets the desired MTU
value to 1500, but you could use 9001 instead.

[ec2-user ~]$ sudo ip link set dev eth0 mtu 1500

2. (Optional) To persist your network MTU setting after a reboot, modify the following configuration
files, based on your operating system type.

• For Amazon Linux 2, add the following line to the /etc/sysconfig/network-scripts/

ifcfg-eth0 file:

MTU=1500

941
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot

Add the following line to the /etc/dhcp/dhclient.conf file:

request subnet-mask, broadcast-address, time-offset, routers, domain-name, domain-

search, domain-name-servers, host-name, nis-domain, nis-servers, ntp-servers;

• For Amazon Linux, add the following lines to your /etc/dhcp/dhclient-eth0.conf file.

interface "eth0" {
supersede interface-mtu 1500;
}

• For other Linux distributions, consult their specific documentation.

3. (Optional) Reboot your instance and verify that the MTU setting is correct.

Troubleshoot
If you experience connectivity issues between your EC2 instance and an Amazon Redshift cluster when
using jumbo frames, see Queries Appear to Hang in the Amazon Redshift Cluster Management Guide

Virtual private clouds

Amazon Virtual Private Cloud (Amazon VPC) enables you to define a virtual network in your own
logically isolated area within the AWS cloud, known as a virtual private cloud (VPC). You can launch your
Amazon EC2 resources, such as instances, into the subnets of your VPC. Your VPC closely resembles a
traditional network that you might operate in your own data center, with the benefits of using scalable
infrastructure from AWS. You can configure your VPC; you can select its IP address range, create subnets,
and configure route tables, network gateways, and security settings. You can connect instances in your
VPC to the internet or to your own data center.

When you create your AWS account, we create a default VPC for you in each Region. A default VPC is a
VPC that is already configured and ready for you to use. You can launch instances into your default VPC
immediately. Alternatively, you can create your own nondefault VPC and configure it as you need.

If you created your AWS account before 2013-12-04, you might have support for the EC2-Classic
platform in some regions. If you created your AWS account after 2013-12-04, it does not support EC2-
Classic, so you must launch your resources in a VPC. For more information, see EC2-Classic (p. 943).

Amazon VPC documentation

For more information about Amazon VPC, see the following documentation.

Guide Description

Amazon VPC User Guide Describes key concepts and provides instructions
for using the features of Amazon VPC.

Amazon VPC Peering Guide Describes VPC peering connections and provides
instructions for using them.

Amazon VPC Transit Gateways Describes transit gateways and provides

instructions for configuring and using them.

942
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EC2-Classic

Guide Description

AWS Site-to-Site VPN User Guide Describes Site-to-Site VPN connections and
provides instructions for configuring and using
them.

EC2-Classic
With EC2-Classic, your instances run in a single, flat network that you share with other customers. With
Amazon VPC, your instances run in a virtual private cloud (VPC) that's logically isolated to your AWS
account.

The EC2-Classic platform was introduced in the original release of Amazon EC2. If you created your
AWS account after 2013-12-04, it does not support EC2-Classic, so you must launch your Amazon EC2
instances in a VPC.

If your account does not support EC2-Classic, we create a default VPC for you. By default, when you
launch an instance, we launch it into your default VPC. Alternatively, you can create a nondefault VPC
and specify it when you launch an instance.

Detect supported platforms

The Amazon EC2 console indicates which platforms you can launch instances into for the selected region,
and whether you have a default VPC in that Region.

Verify that the Region you'll use is selected in the navigation bar. On the Amazon EC2 console
dashboard, look for Supported Platforms under Account Attributes.

Accounts that support EC2-Classic

The dashboard displays the following under Account Attributes to indicate that the account supports
both the EC2-Classic platform and VPCs in this Region, but the Region does not have a default VPC.

The output of the describe-account-attributes command includes both the EC2 and VPC values for the
supported-platforms attribute.

aws ec2 describe-account-attributes --attribute-names supported-platforms

{
"AccountAttributes": [
{
"AttributeName": "supported-platforms",
"AttributeValues": [
{
"AttributeValue": "EC2"
},
{
"AttributeValue": "VPC"
}
]
}
]

943
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance types available in EC2-Classic

Accounts that require a VPC

The dashboard displays the following under Account Attributes to indicate that the account requires a
VPC to launch instances in this Region, does not support the EC2-Classic platform in this Region, and the
Region has a default VPC with the identifier vpc-1a2b3c4d.

The output of the describe-account-attributes command for the specified Region includes only the VPC
value for the supported-platforms attribute.

aws ec2 describe-account-attributes --attribute-names supported-platforms --region us-

east-2
{
"AccountAttributes": [
{
"AttributeValues": [
{
"AttributeValue": "VPC"
}
]
"AttributeName": "supported-platforms",
}
]
}

Instance types available in EC2-Classic

Most of the newer instance types require a VPC. The following are the only instance types supported in
EC2-Classic:

• General purpose: M1, M3, and T1

• Compute optimized: C1, C3, and CC2
• Memory optimized: CR1, M2, and R3
• Storage optimized: D2, HS1, and I2
• Accelerated computing: G2

If your account supports EC2-Classic but you have not created a nondefault VPC, you can do one of the
following to launch instances that require a VPC:

• Create a nondefault VPC and launch your VPC-only instance into it by specifying a subnet ID or a
network interface ID in the request. Note that you must create a nondefault VPC if you do not have
a default VPC and you are using the AWS CLI, Amazon EC2 API, or AWS SDK to launch a VPC-only
instance.
• Launch your VPC-only instance using the Amazon EC2 console. The Amazon EC2 console creates a
nondefault VPC in your account and launches the instance into the subnet in the first Availability Zone.
The console creates the VPC with the following attributes:

944
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Differences between instances in EC2-Classic and a VPC

• One subnet in each Availability Zone, with the public IPv4 addressing attribute set to true so that
instances receive a public IPv4 address. For more information, see IP Addressing in Your VPC in the
Amazon VPC User Guide.
• An Internet gateway, and a main route table that routes traffic in the VPC to the Internet gateway.
This enables the instances you launch in the VPC to communicate over the Internet. For more
information, see Internet Gateways in the Amazon VPC User Guide.
• A default security group for the VPC and a default network ACL that is associated with each subnet.
For more information, see Security Groups for Your VPC in the Amazon VPC User Guide.

If you have other resources in EC2-Classic, you can take steps to migrate them to a VPC. For more
information, see Migrate from EC2-Classic to a VPC (p. 962).

Differences between instances in EC2-Classic and a

VPC
The following table summarizes the differences between instances launched in EC2-Classic, instances
launched in a default VPC, and instances launched in a nondefault VPC.

Characteristic EC2-Classic Default VPC Nondefault VPC

Public IPv4 Your instance receives a Your instance launched in Your instance doesn't
address (from public IPv4 address from a default subnet receives receive a public IPv4
Amazon's the EC2-Classic public IPv4 a public IPv4 address address by default, unless
public IP address pool. by default, unless you you specify otherwise
address pool) specify otherwise during during launch, or you
launch, or you modify modify the subnet's public
the subnet's public IPv4 IPv4 address attribute.
address attribute.

Private IPv4 Your instance receives a Your instance receives a Your instance receives a
address private IPv4 address from static private IPv4 address static private IPv4 address
the EC2-Classic range each from the address range of from the address range of
time it's started. your default VPC. your VPC.

Multiple We select a single You can assign multiple You can assign multiple
private IPv4 private IP address for private IPv4 addresses to private IPv4 addresses to
addresses your instance; multiple your instance. your instance.
IP addresses are not
supported.

Elastic IP An Elastic IP is An Elastic IP remains An Elastic IP remains

address (IPv4) disassociated from your associated with your associated with your
instance when you stop it. instance when you stop it. instance when you stop it.

Associating You associate an Elastic IP An Elastic IP address is An Elastic IP address is

an Elastic IP address with an instance. a property of a network a property of a network
address interface. You associate interface. You associate
an Elastic IP address with an Elastic IP address with
an instance by updating an instance by updating
the network interface the network interface
attached to the instance. attached to the instance.

Reassociating If the Elastic IP address is If the Elastic IP address is If the Elastic IP address
an Elastic IP already associated with already associated with is already associated
address another instance, the another instance, the with another instance,

945
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Differences between instances in EC2-Classic and a VPC

Characteristic EC2-Classic Default VPC Nondefault VPC

address is automatically address is automatically it succeeds only if you
associated with the new associated with the new allowed reassociation.
instance. instance.

Tagging Elastic You cannot apply tags to You can apply tags to an You can apply tags to an
IP addresses an Elastic IP address. Elastic IP address. Elastic IP address.

DNS DNS hostnames are DNS hostnames are DNS hostnames are
hostnames enabled by default. enabled by default. disabled by default.

Security group A security group can A security group can A security group can
reference security groups reference security groups reference security groups
that belong to other AWS for your VPC, or for a for your VPC only.
accounts. peer VPC in a VPC peering
connection.

Security group You can't change the You can assign up to 5 You can assign up to 5
association security groups of your security groups to an security groups to an
running instance. You can instance. instance.
either modify the rules
of the assigned security You can assign security You can assign security
groups, or replace the groups to your instance groups to your instance
instance with a new one when you launch it and when you launch it and
(create an AMI from the while it's running. while it's running.
instance, launch a new
instance from this AMI
with the security groups
that you need, disassociate
any Elastic IP address
from the original instance
and associate it with the
new instance, and then
terminate the original
instance).

Security group You can add rules for You can add rules for You can add rules for
rules inbound traffic only. inbound and outbound inbound and outbound
traffic. traffic.

Tenancy Your instance runs on You can run your instance You can run your instance
shared hardware. on shared hardware or on shared hardware or
single-tenant hardware. single-tenant hardware.

Accessing the Your instance can access By default, your instance By default, your instance
Internet the Internet. Your instance can access the Internet. cannot access the Internet.
automatically receives a Your instance receives Your instance doesn't
public IP address, and can a public IP address by receive a public IP address
access the Internet directly default. An Internet by default. Your VPC may
through the AWS network gateway is attached to have an Internet gateway,
edge. your default VPC, and your depending on how it was
default subnet has a route created.
to the Internet gateway.

946
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Differences between instances in EC2-Classic and a VPC

Characteristic EC2-Classic Default VPC Nondefault VPC

IPv6 IPv6 addressing is not You can optionally You can optionally
addressing supported. You cannot associate an IPv6 CIDR associate an IPv6 CIDR
assign IPv6 addresses to block with your VPC, and block with your VPC, and
your instances. assign IPv6 addresses to assign IPv6 addresses to
instances in your VPC. instances in your VPC.

Security groups for EC2-Classic

If you're using EC2-Classic, you must use security groups created specifically for EC2-Classic. When you
launch an instance in EC2-Classic, you must specify a security group in the same Region as the instance.
You can't specify a security group that you created for a VPC when you launch an instance in EC2-Classic.

After you launch an instance in EC2-Classic, you can't change its security groups. However, you can
add rules to or remove rules from a security group, and those changes are automatically applied to all
instances that are associated with the security group after a short period.

Your AWS account automatically has a default security group per Region for EC2-Classic. If you try
to delete the default security group, you'll get the following error: Client.InvalidGroup.Reserved: The
security group 'default' is reserved.

You can create custom security groups. The security group name must be unique within your account for
the Region. To create a security group for use in EC2-Classic, choose No VPC for the VPC.

You can add inbound rules to your default and custom security groups. You can't change the outbound
rules for an EC2-Classic security group. When you create a security group rule, you can use a different
security group for EC2-Classic in the same Region as the source or destination. To specify a security
group for another AWS account, add the AWS account ID as a prefix; for example, 111122223333/sg-
edcd9784.

In EC2-Classic, you can have up to 500 security groups in each Region for each account. You can add
up to 100 rules to a security group. You can have up to 800 security group rules per instance. This is
calculated as the multiple of rules per security group and security groups per instance. If you reference
other security groups in your security group rules, we recommend that you use security group names
that are 22 characters or less in length.

IP addressing and DNS

Amazon provides a DNS server that resolves Amazon-provided IPv4 DNS hostnames to IPv4 addresses. In
EC2-Classic, the Amazon DNS server is located at 172.16.0.23.

If you create a custom firewall configuration in EC2-Classic, you must create a rule in your firewall that
allows inbound traffic from port 53 (DNS)—with a destination port from the ephemeral range—from
the address of the Amazon DNS server; otherwise, internal DNS resolution from your instances fails. If
your firewall doesn't automatically allow DNS query responses, then you need to allow traffic from the
IP address of the Amazon DNS server. To get the IP address of the Amazon DNS server, use the following
command from within your instance:

grep nameserver /etc/resolv.conf

Elastic IP addresses
If your account supports EC2-Classic, there's one pool of Elastic IP addresses for use with the EC2-Classic
platform and another for use with your VPCs. You can't associate an Elastic IP address that you allocated

947
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Differences between instances in EC2-Classic and a VPC

for use with a VPC with an instance in EC2-Classic, and vice- versa. However, you can migrate an Elastic
IP address you've allocated for use in the EC2-Classic platform for use with a VPC. You cannot migrate an
Elastic IP address to another Region.

To allocate an Elastic IP address for use in EC2-Classic using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Choose Allocate new address.
4. Select Classic, and then choose Allocate. Close the confirmation screen.

Migrate an Elastic IP Address from EC2-Classic

If your account supports EC2-Classic, you can migrate Elastic IP addresses that you've allocated for use
with EC2-Classic platform to be used with a VPC, within the same Region. This can assist you to migrate
your resources from EC2-Classic to a VPC; for example, you can launch new web servers in your VPC, and
then use the same Elastic IP addresses that you used for your web servers in EC2-Classic for your new
VPC web servers.

After you've migrated an Elastic IP address to a VPC, you cannot use it with EC2-Classic. However, if
required, you can restore it to EC2-Classic. You cannot migrate an Elastic IP address that was originally
allocated for use with a VPC to EC2-Classic.

To migrate an Elastic IP address, it must not be associated with an instance. For more information about
disassociating an Elastic IP address from an instance, see Disassociate an Elastic IP address (p. 836).

You can migrate as many EC2-Classic Elastic IP addresses as you can have in your account. However,
when you migrate an Elastic IP address, it counts against your Elastic IP address limit for VPCs. You
cannot migrate an Elastic IP address if it will result in your exceeding your limit. Similarly, when you
restore an Elastic IP address to EC2-Classic, it counts against your Elastic IP address limit for EC2-Classic.
For more information, see Elastic IP address limit (p. 838).

You cannot migrate an Elastic IP address that has been allocated to your account for less than 24 hours.

You can migrate an Elastic IP address from EC2-Classic using the Amazon EC2 console or the Amazon
VPC console. This option is only available if your account supports EC2-Classic.

To move an Elastic IP address using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select the Elastic IP address, and choose Actions, Move to VPC scope.
4. In the confirmation dialog box, choose Move Elastic IP.

You can restore an Elastic IP address to EC2-Classic using the Amazon EC2 console or the Amazon VPC
console.

To restore an Elastic IP address to EC2-Classic using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic IPs.
3. Select the Elastic IP address, choose Actions, Restore to EC2 scope.
4. In the confirmation dialog box, choose Restore.

948
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Share and access resources between EC2-Classic and a VPC

After you've performed the command to move or restore your Elastic IP address, the process of
migrating the Elastic IP address can take a few minutes. Use the describe-moving-addresses command to
check whether your Elastic IP address is still moving, or has completed moving.

After you've moved your Elastic IP address, you can view its allocation ID on the Elastic IPs page in the
Allocation ID field.

If the Elastic IP address is in a moving state for longer than 5 minutes, contact Premium Support.

To move an Elastic IP address using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• move-address-to-vpc (AWS CLI)

• Move-EC2AddressToVpc (AWS Tools for Windows PowerShell)

To restore an Elastic IP address to EC2-Classic using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• restore-address-to-classic (AWS CLI)

• Restore-EC2AddressToClassic (AWS Tools for Windows PowerShell)

To describe the status of your moving addresses using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• describe-moving-addresses (AWS CLI)

• Get-EC2Address (AWS Tools for Windows PowerShell)

Share and access resources between EC2-Classic and

a VPC
Some resources and features in your AWS account can be shared or accessed between EC2-Classic and a
VPC, for example, through ClassicLink. For more information, see ClassicLink (p. 950).

If your account supports EC2-Classic, you might have set up resources for use in EC2-Classic. If you
want to migrate from EC2-Classic to a VPC, you must recreate those resources in your VPC. For more
information about migrating from EC2-Classic to a VPC, see Migrate from EC2-Classic to a VPC (p. 962).

The following resources can be shared or accessed between EC2-Classic and a VPC.

Resource Notes

AMI  

Bundle task  

EBS volume  

Elastic IP address (IPv4) You can migrate an Elastic IP address from EC2-
Classic to a VPC. You can't migrate an Elastic

949
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

Resource Notes
IP address that was originally allocated for use
in a VPC to EC2-Classic. For more information,
see Migrate an Elastic IP Address from EC2-
Classic (p. 948).

Instance An EC2-Classic instance can communicate with

instances in a VPC using public IPv4 addresses, or
you can use ClassicLink to enable communication
over private IPv4 addresses.

You can't migrate an instance from EC2-Classic to

a VPC. However, you can migrate your application
from an instance in EC2-Classic to an instance in a
VPC. For more information, see Migrate from EC2-
Classic to a VPC (p. 962).

Key pair  

Load balancer If you're using ClassicLink, you can register a

linked EC2-Classic instance with a load balancer in
a VPC, provided that the VPC has a subnet in the
same Availability Zone as the instance.

You can't migrate a load balancer from EC2-

Classic to a VPC. You can't register an instance in a
VPC with a load balancer in EC2-Classic.

Placement group  

Reserved Instance You can change the network platform for your
Reserved Instances from EC2-Classic to a VPC.
For more information, see Modify Reserved
Instances (p. 352).

Security group A linked EC2-Classic instance can use a VPC

security groups through ClassicLink to control
traffic to and from the VPC. VPC instances can't
use EC2-Classic security groups.

You can't migrate a security group from EC2-

Classic to a VPC. You can copy rules from a
security group for EC2-Classic to a security group
for a VPC. For more information, see Create a
security group (p. 1064).

Snapshot  

The following resources can't be shared or moved between EC2-Classic and a VPC:

• Spot Instances

ClassicLink
ClassicLink allows you to link EC2-Classic instances to a VPC in your account, within the same Region. If
you associate the VPC security groups with a EC2-Classic instance, this enables communication between

950
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

your EC2-Classic instance and instances in your VPC using private IPv4 addresses. ClassicLink removes
the need to make use of public IPv4 addresses or Elastic IP addresses to enable communication between
instances in these platforms.

ClassicLink is available to all users with accounts that support the EC2-Classic platform, and can be
used with any EC2-Classic instance. For more information about migrating your resources to a VPC, see
Migrate from EC2-Classic to a VPC (p. 962).

There is no additional charge for using ClassicLink. Standard charges for data transfer and instance usage
apply.

Contents
• ClassicLink basics (p. 951)
• ClassicLink limitations (p. 953)
• Work with ClassicLink (p. 954)
• Example IAM policies for ClassicLink (p. 957)
• Example: ClassicLink security group configuration for a three-tier web application (p. 960)

ClassicLink basics
There are two steps to linking an EC2-Classic instance to a VPC using ClassicLink. First, you must enable
the VPC for ClassicLink. By default, all VPCs in your account are not enabled for ClassicLink, to maintain
their isolation. After you've enabled the VPC for ClassicLink, you can then link any running EC2-Classic
instance in the same Region in your account to that VPC. Linking your instance includes selecting security
groups from the VPC to associate with your EC2-Classic instance. After you've linked the instance, it
can communicate with instances in your VPC using their private IP addresses, provided the VPC security
groups allow it. Your EC2-Classic instance does not lose its private IP address when linked to the VPC.

Linking your instance to a VPC is sometimes referred to as attaching your instance.

A linked EC2-Classic instance can communicate with instances in a VPC, but it does not form part of the
VPC. If you list your instances and filter by VPC, for example, through the DescribeInstances API
request, or by using the Instances screen in the Amazon EC2 console, the results do not return any EC2-
Classic instances that are linked to the VPC. For more information about viewing your linked EC2-Classic
instances, see View your ClassicLink-enabled VPCs and linked instances (p. 956).

By default, if you use a public DNS hostname to address an instance in a VPC from a linked EC2-Classic
instance, the hostname resolves to the instance's public IP address. The same occurs if you use a public
DNS hostname to address a linked EC2-Classic instance from an instance in the VPC. If you want the
public DNS hostname to resolve to the private IP address, you can enable ClassicLink DNS support for
the VPC. For more information, see Enable ClassicLink DNS support (p. 956).

If you no longer require a ClassicLink connection between your instance and the VPC, you can unlink
the EC2-Classic instance from the VPC. This disassociates the VPC security groups from the EC2-Classic
instance. A linked EC2-Classic instance is automatically unlinked from a VPC when it's stopped. After
you've unlinked all linked EC2-Classic instances from the VPC, you can disable ClassicLink for the VPC.

Use other AWS services in your VPC with ClassicLink

Linked EC2-Classic instances can access the following AWS services in the VPC: Amazon Redshift,
Amazon ElastiCache, Elastic Load Balancing, and Amazon RDS. However, instances in the VPC cannot
access the AWS services provisioned by the EC2-Classic platform using ClassicLink.

If you use Elastic Load Balancing, you can register your linked EC2-Classic instances with the load
balancer. You must create your load balancer in the ClassicLink-enabled VPC and enable the Availability

951
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

Zone in which the instance runs. If you terminate the linked EC2-Classic instance, the load balancer
deregisters the instance.

If you use Amazon EC2 Auto Scaling, you can create an Amazon EC2 Auto Scaling group with instances
that are automatically linked to a specified ClassicLink-enabled VPC at launch. For more information, see
Linking EC2-Classic Instances to a VPC in the Amazon EC2 Auto Scaling User Guide.

If you use Amazon RDS instances or Amazon Redshift clusters in your VPC, and they are publicly
accessible (accessible from the Internet), the endpoint you use to address those resources from a linked
EC2-Classic instance by default resolves to a public IP address. If those resources are not publicly
accessible, the endpoint resolves to a private IP address. To address a publicly accessible RDS instance
or Redshift cluster over private IP using ClassicLink, you must use their private IP address or private DNS
hostname, or you must enable ClassicLink DNS support for the VPC.

If you use a private DNS hostname or a private IP address to address an RDS instance, the linked EC2-
Classic instance cannot use the failover support available for Multi-AZ deployments.

You can use the Amazon EC2 console to find the private IP addresses of your Amazon Redshift, Amazon
ElastiCache, or Amazon RDS resources.

To locate the private IP addresses of AWS resources in your VPC

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Network Interfaces.
3. Check the descriptions of the network interfaces in the Description column. A network interface
that's used by Amazon Redshift, Amazon ElastiCache, or Amazon RDS will have the name of the
service in the description. For example, a network interface that's attached to an Amazon RDS
instance will have the following description: RDSNetworkInterface.
4. Select the required network interface.
5. In the details pane, get the private IP address from the Primary private IPv4 IP field.

Control the use of ClassicLink

By default, IAM users do not have permission to work with ClassicLink. You can create an IAM policy
that grants users permissions to enable or disable a VPC for ClassicLink, link or unlink an instance to a
ClassicLink-enabled VPC, and to view ClassicLink-enabled VPCs and linked EC2-Classic instances. For
more information about IAM policies for Amazon EC2, see IAM policies for Amazon EC2 (p. 979).

For more information about policies for working with ClassicLink, see the following example: Example
IAM policies for ClassicLink (p. 957).

Security groups in ClassicLink

Linking your EC2-Classic instance to a VPC does not affect your EC2-Classic security groups. They
continue to control all traffic to and from the instance. This excludes traffic to and from instances in the
VPC, which is controlled by the VPC security groups that you associated with the EC2-Classic instance.
EC2-Classic instances that are linked to the same VPC cannot communicate with each other through
the VPC; regardless of whether they are associated with the same VPC security group. Communication
between EC2-Classic instances is controlled by the EC2-Classic security groups associated with those
instances. For an example of a security group configuration, see Example: ClassicLink security group
configuration for a three-tier web application (p. 960).

After you've linked your instance to a VPC, you cannot change which VPC security groups are associated
with the instance. To associate different security groups with your instance, you must first unlink the
instance, and then link it to the VPC again, choosing the required security groups.

952
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

Routing for ClassicLink

When you enable a VPC for ClassicLink, a static route is added to all of the VPC route tables with a
destination of 10.0.0.0/8 and a target of local. This allows communication between instances in the
VPC and any EC2-Classic instances that are then linked to the VPC. If you add a custom route table to a
ClassicLink-enabled VPC, a static route is automatically added with a destination of 10.0.0.0/8 and a
target of local. When you disable ClassicLink for a VPC, this route is automatically deleted in all of the
VPC route tables.

VPCs that are in the 10.0.0.0/16 and 10.1.0.0/16 IP address ranges can be enabled for ClassicLink
only if they do not have any existing static routes in route tables in the 10.0.0.0/8 IP address range,
excluding the local routes that were automatically added when the VPC was created. Similarly, if you've
enabled a VPC for ClassicLink, you may not be able to add any more specific routes to your route tables
within the 10.0.0.0/8 IP address range.
Important
If your VPC CIDR block is a publicly routable IP address range, consider the security implications
before you link an EC2-Classic instance to your VPC. For example, if your linked EC2-Classic
instance receives an incoming Denial of Service (DoS) request flood attack from a source IP
address that falls within the VPC’s IP address range, the response traffic is sent into your VPC.
We strongly recommend that you create your VPC using a private IP address range as specified
in RFC 1918.

For more information about route tables and routing in your VPC, see Route Tables in the Amazon VPC
User Guide.

Enable a VPC peering connection for ClassicLink

If you have a VPC peering connection between two VPCs, and there are one or more EC2-Classic
instances that are linked to one or both of the VPCs via ClassicLink, you can extend the VPC peering
connection to enable communication between the EC2-Classic instances and the instances in the
VPC on the other side of the VPC peering connection. This enables the EC2-Classic instances and the
instances in the VPC to communicate using private IP addresses. To do this, you can enable a local VPC
to communicate with a linked EC2-Classic instance in a peer VPC, or you can enable a local linked EC2-
Classic instance to communicate with instances in a peer VPC.

If you enable a local VPC to communicate with a linked EC2-Classic instance in a peer VPC, a static route
is automatically added to your route tables with a destination of 10.0.0.0/8 and a target of local.

For more information and examples, see Configurations With ClassicLink in the Amazon VPC Peering
Guide.

ClassicLink limitations
To use the ClassicLink feature, you need to be aware of the following limitations:

• You can link an EC2-Classic instance to only one VPC at a time.

• If you stop your linked EC2-Classic instance, it's automatically unlinked from the VPC and the VPC
security groups are no longer associated with the instance. You can link your instance to the VPC again
after you've restarted it.
• You cannot link an EC2-Classic instance to a VPC that's in a different Region or a different AWS
account.
• You cannot use ClassicLink to link a VPC instance to a different VPC, or to a EC2-Classic resource.
To establish a private connection between VPCs, you can use a VPC peering connection. For more
information, see the Amazon VPC Peering Guide.
• You cannot associate a VPC Elastic IP address with a linked EC2-Classic instance.

953
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

• You cannot enable EC2-Classic instances for IPv6 communication. You can associate an IPv6 CIDR block
with your VPC and assign IPv6 address to resources in your VPC, however, communication between a
ClassicLinked instance and resources in the VPC is over IPv4 only.
• VPCs with routes that conflict with the EC2-Classic private IP address range of 10/8 cannot be
enabled for ClassicLink. This does not include VPCs with 10.0.0.0/16 and 10.1.0.0/16 IP address
ranges that already have local routes in their route tables. For more information, see Routing for
ClassicLink (p. 953).
• VPCs configured for dedicated hardware tenancy cannot be enabled for ClassicLink. Contact AWS
support to request that your dedicated tenancy VPC be allowed to be enabled for ClassicLink.
Important
EC2-Classic instances are run on shared hardware. If you've set the tenancy of your VPC to
dedicated because of regulatory or security requirements, then linking an EC2-Classic
instance to your VPC might not conform to those requirements, as this allows a shared
tenancy resource to address your isolated resources directly using private IP addresses. If you
need to enable your dedicated VPC for ClassicLink, provide a detailed reason in your request
to AWS support.
• If you link your EC2-Classic instance to a VPC in the 172.16.0.0/16 range, and you have a DNS
server running on the 172.16.0.23/32 IP address within the VPC, then your linked EC2-Classic
instance can't access the VPC DNS server. To work around this issue, run your DNS server on a different
IP address within the VPC.
• ClassicLink doesn't support transitive relationships out of the VPC. Your linked EC2-Classic instance
doesn't have access to any VPN connection, VPC gateway endpoint, NAT gateway, or Internet gateway
associated with the VPC. Similarly, resources on the other side of a VPN connection or an Internet
gateway don't have access to a linked EC2-Classic instance.

Work with ClassicLink

You can use the Amazon EC2 and Amazon VPC consoles to work with the ClassicLink feature. You can
enable or disable a VPC for ClassicLink, and link and unlink EC2-Classic instances to a VPC.
Note
The ClassicLink features are only visible in the consoles for accounts and Regions that support
EC2-Classic.

Tasks
• Enable a VPC for ClassicLink (p. 954)
• Create a VPC with ClassicLink enabled (p. 955)
• Link an instance to a VPC (p. 955)
• Link an instance to a VPC at launch (p. 955)
• View your ClassicLink-enabled VPCs and linked instances (p. 956)
• Enable ClassicLink DNS support (p. 956)
• Disable ClassicLink DNS support (p. 956)
• Unlink an instance from a VPC (p. 957)
• Disable ClassicLink for a VPC (p. 957)

Enable a VPC for ClassicLink

To link an EC2-Classic instance to a VPC, you must first enable the VPC for ClassicLink. You cannot enable
a VPC for ClassicLink if the VPC has routing that conflicts with the EC2-Classic private IP address range.
For more information, see Routing for ClassicLink (p. 953).

954
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

To enable a VPC for ClassicLink

1. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.

2. In the navigation pane, choose Your VPCs.
3. Select the VPC.
4. Choose Actions, Enable ClassicLink.
5. When prompted for confirmation, choose Enable ClassicLink.
6. (Optional) If you want the public DNS hostname to resolve to the private IP address, enable
ClassicLink DNS support for the VPC before you link any instances. For more information, see Enable
ClassicLink DNS support (p. 956).

Create a VPC with ClassicLink enabled

You can create a new VPC and immediately enable it for ClassicLink by using the VPC wizard in the
Amazon VPC console.

To create a VPC with ClassicLink enabled

1. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.

2. From the Amazon VPC dashboard, choose Launch VPC Wizard.
3. Select one of the VPC configuration options and choose Select.
4. On the next page of the wizard, choose Yes for Enable ClassicLink. Complete the rest of the steps in
the wizard to create your VPC. For more information about using the VPC wizard, see Scenarios for
Amazon VPC in the Amazon VPC User Guide.
5. (Optional) If you want the public DNS hostname to resolve to the private IP address, enable
ClassicLink DNS support for the VPC before you link any instances. For more information, see Enable
ClassicLink DNS support (p. 956).

Link an instance to a VPC

After you've enabled a VPC for ClassicLink, you can link an EC2-Classic instance to it. The instance must
be in the running state.

If you want the public DNS hostname to resolve to the private IP address, enable ClassicLink DNS
support for the VPC before you link the instance. For more information, see Enable ClassicLink DNS
support (p. 956).

To link an instance to a VPC

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select one or more running EC2-Classic instances.
4. Choose Actions, ClassicLink, Link to VPC.
5. Choose the VPC. The console displays only VPCs that are enabled for ClassicLink.
6. Select one or more security groups to associate with your instances. The console displays security
groups only for VPCs enabled for ClassicLink.
7. Choose Link.

Link an instance to a VPC at launch

You can use the launch wizard in the Amazon EC2 console to launch an EC2-Classic instance and
immediately link it to a ClassicLink-enabled VPC.

955
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

To link an instance to a VPC at launch

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the Amazon EC2 dashboard, choose Launch Instance.
3. Select an AMI, and then choose an instance type that is supported on EC2-Classic. For more
information, see Instance types available in EC2-Classic (p. 944).
4. On the Configure Instance Details page, do the following:

a. For Network, choose Launch into EC2-Classic. If this option is disabled, then the instance type
is not supported on EC2-Classic.
b. Expand Link to VPC (ClassicLink) and choose a VPC from Link to VPC. The console displays
only VPCs with ClassicLink enabled.
5. Complete the rest of the steps in the wizard to launch your instance. For more information, see
Launch an instance using the Launch Instance Wizard (p. 526).

View your ClassicLink-enabled VPCs and linked instances

You can view all of your ClassicLink-enabled VPCs in the Amazon VPC console, and your linked EC2-
Classic instances in the Amazon EC2 console.

To view your ClassicLink-enabled VPCs

1. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.

2. In the navigation pane, choose Your VPCs.
3. Select the VPC.
4. If the value of ClassicLink is Enabled, then the VPC is enabled for ClassicLink.

Enable ClassicLink DNS support

You can enable ClassicLink DNS support for your VPC so that DNS hostnames that are addressed
between linked EC2-Classic instances and instances in the VPC resolve to private IP addresses and not
public IP addresses. For this feature to work, your VPC must be enabled for DNS hostnames and DNS
resolution.
Note
If you enable ClassicLink DNS support for your VPC, your linked EC2-Classic instance can access
any private hosted zone associated with the VPC. For more information, see Working with
Private Hosted Zones in the Amazon Route 53 Developer Guide.

To enable ClassicLink DNS support

1. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.

2. In the navigation pane, choose Your VPCs.
3. Select the VPC.
4. Choose Actions, Edit ClassicLink DNS Support.
5. For ClassicLink DNS support, select Enable.
6. Choose Save changes.

Disable ClassicLink DNS support

You can disable ClassicLink DNS support for your VPC so that DNS hostnames that are addressed
between linked EC2-Classic instances and instances in the VPC resolve to public IP addresses and not
private IP addresses.

956
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

To disable ClassicLink DNS support

1. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.

2. In the navigation pane, choose Your VPCs.
3. Select the VPC.
4. Choose Actions, Edit ClassicLink DNS Support.
5. For ClassicLink DNS Support, clear Enable.
6. Choose Save changes.

Unlink an instance from a VPC

If you no longer require a ClassicLink connection between your EC2-Classic instance and your VPC, you
can unlink the instance from the VPC. Unlinking the instance disassociates the VPC security groups from
the instance.

A stopped instance is automatically unlinked from a VPC.

To unlink an instance from a VPC

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select one or more of your instances.
4. Choose Actions, ClassicLink, Unlink from VPC.
5. When prompted for confirmation, choose Unlink.

Disable ClassicLink for a VPC

If you no longer require a connection between EC2-Classic instances and your VPC, you can disable
ClassicLink on the VPC. You must first unlink all linked EC2-Classic instances that are linked to the VPC.

To disable ClassicLink for a VPC

1. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.

2. In the navigation pane, choose Your VPCs.
3. Select your VPC.
4. Choose Actions, Disable ClassicLink.
5. When prompted for confirmation, choose Disable ClassicLink.

Example IAM policies for ClassicLink

You can enable a VPC for ClassicLink and then link an EC2-Classic instance to the VPC. You can
also view your ClassicLink-enabled VPCs, and all of your EC2-Classic instances that are linked to a
VPC. You can create policies with resource-level permission for the ec2:EnableVpcClassicLink,
ec2:DisableVpcClassicLink, ec2:AttachClassicLinkVpc, and ec2:DetachClassicLinkVpc
actions to control how users are able to use those actions. Resource-level permissions are not supported
for ec2:Describe* actions.

Examples
• Full permissions to work with ClassicLink (p. 958)
• Enable and disable a VPC for ClassicLink (p. 958)

957
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

• Link instances (p. 958)

• Unlink instances (p. 959)

Full permissions to work with ClassicLink

The following policy grants users permissions to view ClassicLink-enabled VPCs and linked EC2-
Classic instances, to enable and disable a VPC for ClassicLink, and to link and unlink instances from a
ClassicLink-enabled VPC.

{
"Version": "2012-10-17",
"Statement": [{
"Effect": "Allow",
"Action": [
"ec2:DescribeClassicLinkInstances", "ec2:DescribeVpcClassicLink",
"ec2:EnableVpcClassicLink", "ec2:DisableVpcClassicLink",
"ec2:AttachClassicLinkVpc", "ec2:DetachClassicLinkVpc"
],
"Resource": "*"
}
]
}

Enable and disable a VPC for ClassicLink

The following policy allows user to enable and disable VPCs for ClassicLink that have the specific tag
'purpose=classiclink'. Users cannot enable or disable any other VPCs for ClassicLink.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:*VpcClassicLink",
"Resource": "arn:aws:ec2:region:account:vpc/*",
"Condition": {
"StringEquals": {
"ec2:ResourceTag/purpose":"classiclink"
}
}
}
]
}

Link instances
The following policy grants users permissions to link instances to a VPC only if the instance is an
m3.large instance type. The second statement allows users to use the VPC and security group
resources, which are required to link an instance to a VPC.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:AttachClassicLinkVpc",
"Resource": "arn:aws:ec2:region:account:instance/*",
"Condition": {

958
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

"StringEquals": {
"ec2:InstanceType":"m3.large"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:AttachClassicLinkVpc",
"Resource": [
"arn:aws:ec2:region:account:vpc/*",
"arn:aws:ec2:region:account:security-group/*"
]
}
]
}

The following policy grants users permissions to link instances to a specific VPC (vpc-1a2b3c4d) only,
and to associate only specific security groups from the VPC to the instance (sg-1122aabb and sg-
aabb2233). Users cannot link an instance to any other VPC, and they cannot specify any other of the
VPC security groups to associate with the instance in the request.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:AttachClassicLinkVpc",
"Resource": [
"arn:aws:ec2:region:account:vpc/vpc-1a2b3c4d",
"arn:aws:ec2:region:account:instance/*",
"arn:aws:ec2:region:account:security-group/sg-1122aabb",
"arn:aws:ec2:region:account:security-group/sg-aabb2233"
]
}
]
}

Unlink instances
The following grants users permission to unlink any linked EC2-Classic instance from a VPC, but only if
the instance has the tag "unlink=true". The second statement grants users permissions to use the VPC
resource, which is required to unlink an instance from a VPC.

{
"Version": "2012-10-17",
"Statement": [{
"Effect": "Allow",
"Action": "ec2:DetachClassicLinkVpc",
"Resource": [
"arn:aws:ec2:region:account:instance/*"
],
"Condition": {
"StringEquals": {
"ec2:ResourceTag/unlink":"true"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:DetachClassicLinkVpc",
"Resource": [
"arn:aws:ec2:region:account:vpc/*"

959
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

]
}
]
}

Example: ClassicLink security group configuration for a three-

tier web application
In this example, you have an application with three instances: a public-facing web server, an application
server, and a database server. Your web server accepts HTTPS traffic from the Internet, and then
communicates with your application server over TCP port 6001. Your application server then
communicates with your database server over TCP port 6004. You're in the process of migrating your
entire application to a VPC in your account. You've already migrated your application server and your
database server to your VPC. Your web server is still in EC2-Classic and linked to your VPC via ClassicLink.

You want a security group configuration that allows traffic to flow only between these instances. You
have four security groups: two for your web server (sg-1a1a1a1a and sg-2b2b2b2b), one for your
application server (sg-3c3c3c3c), and one for your database server (sg-4d4d4d4d).

The following diagram displays the architecture of your instances, and their security group configuration.

960
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ClassicLink

Security groups for your web server (sg-1a1a1a1a and sg-2b2b2b2b)

You have one security group in EC2-Classic, and the other in your VPC. You associated the VPC security
group with your web server instance when you linked the instance to your VPC via ClassicLink. The VPC
security group enables you to control the outbound traffic from your web server to your application
server.

The following are the security group rules for the EC2-Classic security group (sg-1a1a1a1a).

Inbound

Source Type Port Range Comments

0.0.0.0/0 HTTPS 443 Allows Internet traffic

to reach your web
server.

The following are the security group rules for the VPC security group (sg-2b2b2b2b).

Outbound

Destination Type Port Range Comments

sg-3c3c3c3c TCP 6001 Allows outbound traffic

from your web server
to your application
server in your VPC
(or to any other
instance associated with
sg-3c3c3c3c).

Security group for your application server (sg-3c3c3c3c)

The following are the security group rules for the VPC security group that's associated with your
application server.

Inbound

Source Type Port Range Comments

sg-2b2b2b2b TCP 6001 Allows the specified

type of traffic from
your web server (or
any other instance
associated with
sg-2b2b2b2b) to reach
your application server.

Outbound

Destination Type Port Range Comments

sg-4d4d4d4d TCP 6004 Allows outbound traffic

from the application
server to the database
server (or to any other

961
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Migrate from EC2-Classic to a VPC

instance associated with

sg-4d4d4d4d).

Security group for your database server (sg-4d4d4d4d)

The following are the security group rules for the VPC security group that's associated with your
database server.

Inbound

Source Type Port Range Comments

sg-3c3c3c3c TCP 6004 Allows the specified

type of traffic from
your application
server (or any other
instance associated with
sg-3c3c3c3c) to reach
your database server.

Migrate from EC2-Classic to a VPC

If you created your AWS account before December 4, 2013, you might have support for EC2-Classic in
some AWS Regions. Some Amazon EC2 resources and features, such as enhanced networking and newer
instance types, require a virtual private cloud (VPC). Some resources can be shared between EC2-Classic
and a VPC, while some can't. For more information, see Share and access resources between EC2-Classic
and a VPC (p. 949). We recommend that you migrate to a VPC to take advantage of VPC-only features.

To migrate from EC2-Classic to a VPC, you must migrate or recreate your EC2-Classic resources in a VPC.
You can migrate and recreate your resources in full, or you can perform an incremental migration over
time using ClassicLink.

Contents
• Options for getting a default VPC (p. 962)
• Migrate your resources to a VPC (p. 963)
• Use ClassicLink for an incremental migration (p. 967)
• Example: Migrate a simple web application (p. 968)

Options for getting a default VPC

A default VPC is a VPC that is configured and ready for you to use, and is only available in Regions that
are VPC-only. For Regions that support EC2-Classic, you can create a nondefault VPC to set up your
resources. However, you might want to use a default VPC if you prefer not to set up a VPC yourself, or if
you do not have specific requirements for your VPC configuration. For more information about default
VPCs, see Default VPC and Default Subnets in the Amazon VPC User Guide.

The following are options for using a default VPC when you have an AWS account that supports EC2-
Classic.

Options
• Switch to a VPC-only Region (p. 963)
• Create a new AWS account (p. 963)

962
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Migrate from EC2-Classic to a VPC

• Convert your existing AWS account to VPC-only (p. 963)

Switch to a VPC-only Region

Use this option if you want to use your existing account to set up your resources in a default VPC and
you do not need to use a specific Region. To find a Region that has a default VPC, see Detect supported
platforms (p. 943).

Create a new AWS account

New AWS accounts support VPC only. Use this option if you want an account that has a default VPC in
every Region.

Convert your existing AWS account to VPC-only

Use this option if you want a default VPC in every Region in your existing account. Before you can
convert your account, you must delete all of your EC2-Classic resources. You can also migrate some
resources to a VPC. For more information, see Migrate your resources to a VPC (p. 963).

To convert your EC2-Classic account

1. Delete or migrate (if applicable) the resources that you have created for use in EC2-Classic. These
include the following:

• Amazon EC2 instances

• EC2-Classic security groups (excluding the default security group, which you cannot delete
yourself)
• EC2-Classic Elastic IP addresses
• Classic Load Balancers
• Amazon RDS resources
• Amazon ElastiCache resources
• Amazon Redshift resources
• AWS Elastic Beanstalk resources
• AWS Data Pipeline resources
• Amazon EMR resources
• AWS OpsWorks resources
2. Go to the AWS Support Center at console.aws.amazon.com/support.
3. Choose Create case.
4. Choose Account and billing support.
5. For Type, choose Account. For Category, choose Convert EC2 Classic to VPC.
6. Fill in the other details as required, and choose Submit. We will review your request and contact you
to guide you through the next steps.

Migrate your resources to a VPC

You can migrate or move some of your resources to a VPC. Some resources can only be migrated from
EC2-Classic to a VPC that's in the same Region and in the same AWS account. If the resource cannot be
migrated, you must create a new resource for use in your VPC.

Prerequisites

Before you begin, you must have a VPC. If you don't have a default VPC, you can create a nondefault VPC
using one of these methods:

963
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Migrate from EC2-Classic to a VPC

• In the Amazon VPC console, use the VPC wizard to create a new VPC. For more information, see
Amazon VPC Console Wizard Configurations. Use this option if you want to set up a VPC quickly, using
one of the available configuration options.
• In the Amazon VPC console, set up the components of a VPC according to your requirements. For more
information, see VPCs and Subnets. Use this option if you have specific requirements for your VPC,
such as a particular number of subnets.

Topics
• Security groups (p. 964)
• Elastic IP addresses (p. 964)
• AMIs and instances (p. 965)
• Amazon RDS DB instances (p. 967)

Security groups
If you want your instances in your VPC to have the same security group rules as your EC2-Classic
instances, you can use the Amazon EC2 console to copy your existing EC2-Classic security group rules to
a new VPC security group.

You can only copy security group rules to a new security group in the same AWS account in the same
Region. If you are using a different Region or a different AWS account, you must create a new security
group and manually add the rules yourself. For more information, see Amazon EC2 security groups for
Linux instances (p. 1058).

To copy your security group rules to a new security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Select the security group that's associated with your EC2-Classic instance, then choose Actions, and
select Copy to new.
Note
To identify an EC2-Classic security group, check the VPC ID column. For each EC2-Classic
security group, the value in the column is blank or a - symbol.
4. In the Create Security Group dialog box, specify a name and description for your new security
group. Select your VPC from the VPC list.
5. The Inbound tab is populated with the rules from your EC2-Classic security group. You can modify
the rules as required. In the Outbound tab, a rule that allows all outbound traffic has automatically
been created for you. For more information about modifying security group rules, see Amazon EC2
security groups for Linux instances (p. 1058).
Note
If you've defined a rule in your EC2-Classic security group that references another security
group, you cannot use the same rule in your VPC security group. Modify the rule to
reference a security group in the same VPC.
6. Choose Create.

Elastic IP addresses
You can migrate an Elastic IP address that is allocated for use in EC2-Classic for use with a VPC. You
cannot migrate an Elastic IP address to another Region or AWS account. For more information, see
Migrate an Elastic IP Address from EC2-Classic (p. 948).

To identify an Elastic IP address that is allocated for use in EC2-Classic

964
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Migrate from EC2-Classic to a VPC

In the Amazon EC2 console, choose Elastic IPs in the navigation pane. In the Scope column, the value is
standard.

Alternatively, use the following describe-addresses command.

aws ec2 describe-addresses --filters Name=domain,Values=standard

AMIs and instances

An AMI is a template for launching your Amazon EC2 instance. You can create your own AMI based on an
existing EC2-Classic instance, then use that AMI to launch instances into your VPC.

Contents
• Identify EC2-Classic instances (p. 965)
• Create an AMI (p. 965)
• (Optional) Share or copy your AMI (p. 966)
• (Optional) Store your data on Amazon EBS volumes (p. 966)
• Launch an instance into your VPC (p. 966)

Identify EC2-Classic instances

If you have instances running in both EC2-Classic and a VPC, you can identify your EC2-Classic instances.

Amazon EC2 console

Choose Instances in the navigation pane. In the VPC ID column, the value for each EC2-Classic instance
is blank or a - symbol. If the VPC ID column is not present, choose the gear icon and make the column
visible.

AWS CLI

Use the following describe-instances AWS CLI command. The --query parameter displays only instances
where the value for VpcId is null.

aws ec2 describe-instances --query 'Reservations[*].Instances[?VpcId==`null`]'

Create an AMI
After you've identified your EC2-Classic instance, you can create an AMI from it.

To create a Windows AMI

For more information, see Creating a custom Windows AMI.

To create a Linux AMI

The method that you use to create your Linux AMI depends on the root device type of your instance,
and the operating system platform on which your instance runs. To find out the root device type of your
instance, go to the Instances page, select your instance, and look at the information in the Root device
type field in the Description tab. If the value is ebs, then your instance is EBS-backed. If the value is
instance-store, then your instance is instance store-backed. You can also use the describe-instances
AWS CLI command to find out the root device type.

The following table provides options for you to create your Linux AMI based on the root device type of
your instance, and the software platform.
Important
Some instance types support both PV and HVM virtualization, while others support only
one or the other. If you plan to use your AMI to launch a different instance type than your

965
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Migrate from EC2-Classic to a VPC

current instance type, verify that the instance type supports the type of virtualization that your
AMI offers. If your AMI supports PV virtualization, and you want to use an instance type that
supports HVM virtualization, you might have to reinstall your software on a base HVM AMI. For
more information about PV and HVM virtualization, see Linux AMI virtualization types.

Instance root device Action

type

EBS Create an EBS-backed AMI from your instance. For more information, see
Creating an Amazon EBS-backed Linux AMI.

Instance store Create an instance store-backed AMI from your instance using the AMI tools.
For more information, see Creating an instance store-backed Linux AMI.

Instance store Convert your instance store-backed instance to an EBS-backed instance.

For more information, see Converting your instance store-backed AMI to an
Amazon EBS-backed AMI.

(Optional) Share or copy your AMI

To use your AMI to launch an instance in a new AWS account, you must first share the AMI with your new
account. For more information, see Share an AMI with specific AWS accounts (p. 115).

To use your AMI to launch an instance in a VPC in a different Region, you must first copy the AMI to that
Region. For more information, see Copy an AMI (p. 166).

(Optional) Store your data on Amazon EBS volumes

You can create an Amazon EBS volume and use it to back up and store the data on your instance—
like you would use a physical hard drive. Amazon EBS volumes can be attached and detached from any
instance in the same Availability Zone. You can detach a volume from your instance in EC2-Classic, and
attach it to a new instance that you launch into your VPC in the same Availability Zone.

For more information about Amazon EBS volumes, see the following topics:

• Amazon EBS volumes (p. 1082)

• Create an Amazon EBS volume (p. 1104)
• Attach an Amazon EBS volume to an instance (p. 1106)

To back up the data on your Amazon EBS volume, you can take periodic snapshots of your volume.
For more information, see Create Amazon EBS snapshots (p. 1129). If you need to, you can create
an Amazon EBS volume from your snapshot. For more information, see Create a volume from a
snapshot (p. 1105).

Launch an instance into your VPC

After you've created an AMI, you can use the Amazon EC2 launch wizard to launch an instance into your
VPC. The instance will have the same data and configurations as your existing EC2-Classic instance.
Note
You can use this opportunity to upgrade to a current generation instance type. However, verify
that the instance type supports the type of virtualization that your AMI offers (PV or HVM). For
more information about PV and HVM virtualization, see Linux AMI virtualization types (p. 104).

To launch an instance into your VPC

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. On the dashboard, choose Launch instance.

966
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Migrate from EC2-Classic to a VPC

3. On the Choose an Amazon Machine Image page, select the My AMIs category, and select the AMI
you created. Alternatively, if you shared an AMI from another account, in the Ownership filter list,
choose Shared with me. Select the AMI that you shared from your EC2-Classic account.
4. On the Choose an Instance Type page, select the type of instance, and choose Next: Configure
Instance Details.
5. On the Configure Instance Details page, select your VPC from the Network list. Select the required
subnet from the Subnet list. Configure any other details that you require, then go through the next
pages of the wizard until you reach the Configure Security Group page.
6. Select Select an existing group, and select the security group that you created for your VPC. Choose
Review and Launch.
7. Review your instance details, then choose Launch to specify a key pair and launch your instance.

For more information about the parameters that you can configure in each step of the wizard, see
Launch an instance using the Launch Instance Wizard (p. 526).

Amazon RDS DB instances

You can move your EC2-Classic DB instance to a VPC in the same Region, in the same account. For more
information, see Updating the VPC for a DB Instance in the Amazon RDS User Guide.

Use ClassicLink for an incremental migration

The ClassicLink feature makes it easier to manage an incremental migration to a VPC. ClassicLink enables
you to link an EC2-Classic instance to a VPC in your account in the same Region, allowing your new VPC
resources to communicate with the EC2-Classic instance using private IPv4 addresses. You can then
migrate functionality one component at a time until your application is running fully in your VPC.

Use this option if you cannot afford downtime during the migration, for example, if you have a multi-tier
application with processes that cannot be interrupted.

For more information about ClassicLink, see ClassicLink (p. 950).

Tasks
• Step 1: Prepare your migration sequence (p. 967)
• Step 2: Enable your VPC for ClassicLink (p. 967)
• Step 3: Link your EC2-Classic instances to your VPC (p. 968)
• Step 4: Complete the VPC migration (p. 968)

Step 1: Prepare your migration sequence

To use ClassicLink effectively, you must first identify the components of your application that must be
migrated to the VPC, and then confirm the order in which to migrate that functionality.

For example, you have an application that relies on a presentation web server, a backend database
server, and authentication logic for transactions. You may decide to start the migration process with the
authentication logic, then the database server, and finally, the web server.

Then, you can start migrating or recreating your resources. For more information, see Migrate your
resources to a VPC (p. 963).

Step 2: Enable your VPC for ClassicLink

After you've configured your new VPC instances and made the functionality of your application available
in the VPC, you can use ClassicLink to enable private IP communication between your new VPC instances
and your EC2-Classic instances. First, you must enable your VPC for ClassicLink.

967
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Migrate from EC2-Classic to a VPC

To enable a VPC for ClassicLink

1. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.

2. In the navigation pane, choose Your VPCs.
3. Select a VPC.
4. Choose Actions, Enable ClassicLink.
5. When prompted for confirmation, choose Enable ClassicLink.

Step 3: Link your EC2-Classic instances to your VPC

After you've enabled ClassicLink in your VPC, you can link your EC2-Classic instances to the VPC. The
instance must be in the running state.

To link an instance to a VPC

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select one or more running EC2-Classic instances.
4. Choose Actions, ClassicLink, Link to VPC.
5. Choose a VPC. The console displays only VPCs that are enabled for ClassicLink.
6. Select one or more security groups to associate with your instances. The console displays security
groups only for VPCs enabled for ClassicLink.
7. Choose Link.

Step 4: Complete the VPC migration

Depending on the size of your application and the functionality that must be migrated, repeat the
preceding steps until you've moved all of the components of your application from EC2-Classic into your
VPC.

After you've enabled internal communication between the EC2-Classic and VPC instances, you must
update your application to point to your migrated service in your VPC, instead of your service in the EC2-
Classic platform. The exact steps for this depend on your application’s design. Generally, this includes
updating your destination IP addresses to point to the IP addresses of your VPC instances instead of your
EC2-Classic instances.

After you've completed this step and you've tested that the application is functioning from your VPC, you
can terminate your EC2-Classic instances, and disable ClassicLink for your VPC. You can also clean up any
EC2-Classic resources that you no longer need to avoid incurring charges for them. For example, you can
release Elastic IP addresses and delete the volumes that were associated with your EC2-Classic instances.

Example: Migrate a simple web application

In this example, you use AWS to host your gardening website. To manage your website, you have three
running instances in EC2-Classic. Instances A and B host your public-facing web application, and you use
Elastic Load Balancing to load balance the traffic between these instances. You've assigned Elastic IP
addresses to instances A and B so that you have static IP addresses for configuration and administration
tasks on those instances. Instance C holds your MySQL database for your website. You've registered the
domain name www.garden.example.com, and you've used Route 53 to create a hosted zone with an
alias record set that's associated with the DNS name of your load balancer.

968
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Migrate from EC2-Classic to a VPC

The first part of migrating to a VPC is deciding what kind of VPC architecture suits your needs. In this
case, you've decided on the following: one public subnet for your web servers, and one private subnet for
your database server. As your website grows, you can add more web servers and database servers to your
subnets. By default, instances in the private subnet cannot access the internet; however, you can enable
internet access through a Network Address Translation (NAT) device in the public subnet. You might
want to set up a NAT device to support periodic updates and patches from the internet for your database
server. You'll migrate your Elastic IP addresses to a VPC, and create a load balancer in your public subnet
to load balance the traffic between your web servers.

969
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Migrate from EC2-Classic to a VPC

To migrate your web application to a VPC, you can follow these steps:

• Create a VPC: In this case, you can use the VPC wizard in the Amazon VPC console to create your VPC
and subnets. The second wizard configuration creates a VPC with one private and one public subnet,
and launches and configures a NAT device in your public subnet for you. For more information, see VPC
with public and private subnets (NAT) in the Amazon VPC User Guide.
• Configure your security groups: In your EC2-Classic environment, you have one security group for
your web servers, and another security group for your database server. You can use the Amazon EC2
console to copy the rules from each security group into new security groups for your VPC. For more
information, see Security groups (p. 964).
Tip
Create the security groups that are referenced by other security groups first.
• Create AMIs and launch new instances: Create an AMI from one of your web servers, and a second
AMI from your database server. Then, launch replacement web servers into your public subnet, and
launch your replacement database server into your private subnet. For more information, see Create
an AMI (p. 965).
• Configure your NAT device: If you are using a NAT instance, you must create a security group for
it that allows HTTP and HTTPS traffic from your private subnet. For more information, see NAT
instances. If you are using a NAT gateway, traffic from your private subnet is automatically allowed.
• Configure your database: When you created an AMI from your database server in EC2-Classic, all
of the configuration information that was stored in that instance was copied to the AMI. You might
have to connect to your new database server and update the configuration details. For example, if you

970
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Migrate from EC2-Classic to a VPC

configured your database to grant full read, write, and modification permissions to your web servers in
EC2-Classic, you need to update the configuration files to grant the same permissions to your new VPC
web servers instead.
• Configure your web servers: Your web servers will have the same configuration settings as your
instances in EC2-Classic. For example, if you configured your web servers to use the database in EC2-
Classic, update your web servers' configuration settings to point to your new database instance.
Note
By default, instances launched into a nondefault subnet are not assigned a public IP address,
unless you specify otherwise at launch. Your new database server might not have a public
IP address. In this case, you can update your web servers' configuration file to use your new
database server's private DNS name. Instances in the same VPC can communicate with each
other via private IP address.
• Migrate your Elastic IP addresses: Disassociate your Elastic IP addresses from your web servers in EC2-
Classic, and then migrate them to a VPC. After you've migrated them, you can associate them with
your new web servers in your VPC. For more information, see Migrate an Elastic IP Address from EC2-
Classic (p. 948).
• Create a new load balancer: To continue using Elastic Load Balancing to load balance the traffic to
your instances, make sure you understand the various ways to configure your load balancer in VPC. For
more information, see the Elastic Load Balancing User Guide.
• Update your DNS records: After you've set up your load balancer in your public subnet, verify that
your www.garden.example.com domain points to your new load balancer. To do this, update your
DNS records and your alias record set in Route 53. For more information about using Route 53, see
Getting Started with Route 53.
• Shut down your EC2-Classic resources: After you've verified that your web application is working from
within the VPC architecture, you can shut down your EC2-Classic resources to stop incurring charges
for them.

971
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Infrastructure security

Security in Amazon EC2

Cloud security at AWS is the highest priority. As an AWS customer, you benefit from a data center
and network architecture that are built to meet the requirements of the most security-sensitive
organizations.

Security is a shared responsibility between AWS and you. The shared responsibility model describes this
as security of the cloud and security in the cloud:

• Security of the cloud – AWS is responsible for protecting the infrastructure that runs AWS services in
the AWS Cloud. AWS also provides you with services that you can use securely. Third-party auditors
regularly test and verify the effectiveness of our security as part of the AWS Compliance Programs.
To learn about the compliance programs that apply to Amazon EC2, see AWS Services in Scope by
Compliance Program.
• Security in the cloud – Your responsibility includes the following areas:
• Controlling network access to your instances, for example, through configuring your VPC and
security groups. For more information, see Controlling network traffic (p. 973).
• Managing the credentials used to connect to your instances.
• Managing the guest operating system and software deployed to the guest operating system,
including updates and security patches. For more information, see Update management in Amazon
EC2 (p. 1077).
• Configuring the IAM roles that are attached to the instance and the permissions associated with
those roles. For more information, see IAM roles for Amazon EC2 (p. 1032).

This documentation helps you understand how to apply the shared responsibility model when using
Amazon EC2. It shows you how to configure Amazon EC2 to meet your security and compliance
objectives. You also learn how to use other AWS services that help you to monitor and secure your
Amazon EC2 resources.

Contents
• Infrastructure security in Amazon EC2 (p. 972)
• Amazon EC2 and interface VPC endpoints (p. 974)
• Resilience in Amazon EC2 (p. 975)
• Data protection in Amazon EC2 (p. 976)
• Identity and access management for Amazon EC2 (p. 977)
• Amazon EC2 key pairs and Linux instances (p. 1044)
• Amazon EC2 security groups for Linux instances (p. 1058)
• Update management in Amazon EC2 (p. 1077)
• Compliance validation for Amazon EC2 (p. 1078)

Infrastructure security in Amazon EC2

As a managed service, Amazon EC2 is protected by the AWS global network security procedures that are
described in the Amazon Web Services: Overview of Security Processes whitepaper.

You use AWS published API calls to access Amazon EC2 through the network. Clients must support
Transport Layer Security (TLS) 1.0 or later. We recommend TLS 1.2 or later. Clients must also support
cipher suites with perfect forward secrecy (PFS) such as Ephemeral Diffie-Hellman (DHE) or Elliptic Curve
Ephemeral Diffie-Hellman (ECDHE). Most modern systems such as Java 7 and later support these modes.

972
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Network isolation

Additionally, requests must be signed using an access key ID and a secret access key that is associated
with an IAM principal. Or you can use the AWS Security Token Service (AWS STS) to generate temporary
security credentials to sign requests.

Network isolation
A virtual private cloud (VPC) is a virtual network in your own logically isolated area in the AWS Cloud.
Use separate VPCs to isolate infrastructure by workload or organizational entity.

A subnet is a range of IP addresses in a VPC. When you launch an instance, you launch it into a subnet
in your VPC. Use subnets to isolate the tiers of your application (for example, web, application, and
database) within a single VPC. Use private subnets for your instances if they should not be accessed
directly from the internet.

To call the Amazon EC2 API from your VPC without sending traffic over the public internet, use AWS
PrivateLink.

Isolation on physical hosts

Different EC2 instances on the same physical host are isolated from each other as though they are
on separate physical hosts. The hypervisor isolates CPU and memory, and the instances are provided
virtualized disks instead of access to the raw disk devices.

When you stop or terminate an instance, the memory allocated to it is scrubbed (set to zero) by the
hypervisor before it is allocated to a new instance, and every block of storage is reset. This ensures that
your data is not unintentionally exposed to another instance.

Network MAC addresses are dynamically assigned to instances by the AWS network infrastructure. IP
addresses are either dynamically assigned to instances by the AWS network infrastructure, or assigned
by an EC2 administrator through authenticated API requests. The AWS network allows instances to send
traffic only from the MAC and IP addresses assigned to them. Otherwise, the traffic is dropped.

By default, an instance cannot receive traffic that is not specifically addressed to it. If you need to run
network address translation (NAT), routing, or firewall services on your instance, you can disable source/
destination checking for the network interface.

Controlling network traffic

Consider the following options for controlling network traffic to your EC2 instances:

• Restrict access to your instances using security groups (p. 1058). For example, you can allow traffic
only from the address ranges for your corporate network.
• Use private subnets for your instances if they should not be accessed directly from the internet. Use a
bastion host or NAT gateway for internet access from an instance in a private subnet.
• Use AWS Virtual Private Network or AWS Direct Connect to establish private connections from your
remote networks to your VPCs. For more information, see Network-to-Amazon VPC Connectivity
Options.
• Use VPC Flow Logs to monitor the traffic that reaches your instances.
• Use AWS Security Hub to check for unintended network accessibility from your instances.
• Use EC2 Instance Connect (p. 604) to connect to your instances using Secure Shell (SSH) without the
need to share and manage SSH keys.
• Use AWS Systems Manager Session Manager to access your instances remotely instead of opening
inbound SSH ports and managing SSH keys.
• Use AWS Systems Manager Run Command to automate common administrative tasks instead of
opening inbound SSH ports and managing SSH keys.

973
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Interface VPC endpoints

In addition to restricting network access to each Amazon EC2 instance, Amazon VPC supports
implementing additional network security controls like in-line gateways, proxy servers, and various
network monitoring options.

For more information, see the AWS Security Best Practices whitepaper.

Amazon EC2 and interface VPC endpoints

You can improve the security posture of your VPC by configuring Amazon EC2 to use an interface
VPC endpoint. Interface endpoints are powered by AWS PrivateLink, a technology that enables you to
privately access Amazon EC2 APIs by restricting all network traffic between your VPC and Amazon EC2 to
the Amazon network. With interface endpoints, you also don't need an internet gateway, a NAT device,
or a virtual private gateway.

You are not required to configure AWS PrivateLink, but it's recommended. For more information about
AWS PrivateLink and VPC endpoints, see Interface VPC Endpoints (AWS PrivateLink).

Topics
• Create an interface VPC endpoint (p. 974)
• Create an interface VPC endpoint policy (p. 974)

Create an interface VPC endpoint

Create an endpoint for Amazon EC2 using the following service name:

• com.amazonaws.region.ec2 — Creates an endpoint for the Amazon EC2 API actions.

For more information, see Creating an Interface Endpoint in the Amazon VPC User Guide.

Create an interface VPC endpoint policy

You can attach a policy to your VPC endpoint to control access to the Amazon EC2 API. The policy
specifies:

• The principal that can perform actions.

• The actions that can be performed.
• The resource on which the actions can be performed.

Important
When a non-default policy is applied to an interface VPC endpoint for Amazon EC2, certain
failed API requests, such as those failing from RequestLimitExceeded, might not be logged
to AWS CloudTrail or Amazon CloudWatch.

For more information, see Controlling Access to Services with VPC Endpoints in the Amazon VPC User
Guide.

The following example shows a VPC endpoint policy that denies permission to create unencrypted
volumes or to launch instances with unencrypted volumes. The example policy also grants permission to
perform all other Amazon EC2 actions.

{
"Version": "2012-10-17",
"Statement": [

974
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Resilience

{
"Action": "ec2:*",
"Effect": "Allow",
"Resource": "*",
"Principal": "*"
},
{
"Action": [
"ec2:CreateVolume"
],
"Effect": "Deny",
"Resource": "*",
"Principal": "*",
"Condition": {
"Bool": {
"ec2:Encrypted": "false"
}
}
},
{
"Action": [
"ec2:RunInstances"
],
"Effect": "Deny",
"Resource": "*",
"Principal": "*",
"Condition": {
"Bool": {
"ec2:Encrypted": "false"
}
}
}]
}

Resilience in Amazon EC2

The AWS global infrastructure is built around AWS Regions and Availability Zones. Regions provide
multiple physically separated and isolated Availability Zones, which are connected through low-latency,
high-throughput, and highly redundant networking. With Availability Zones, you can design and operate
applications and databases that automatically fail over between zones without interruption. Availability
Zones are more highly available, fault tolerant, and scalable than traditional single or multiple data
center infrastructures.

If you need to replicate your data or applications over greater geographic distances, use AWS Local
Zones. An AWS Local Zone is an extension of an AWS Region in geographic proximity to your users. Local
Zones have their own connections to the internet and support AWS Direct Connect. Like all AWS Regions,
AWS Local Zones are completely isolated from other AWS Zones.

If you need to replicate your data or applications in an AWS Local Zone, AWS recommends that you use
one of the following zones as the failover zone:

• Another Local Zone

• An Availability Zone in the Region that is not the parent zone. You can use the describe-availability-
zones command to view the parent zone.

For more information about AWS Regions and Availability Zones, see AWS Global Infrastructure.

In addition to the AWS global infrastructure, Amazon EC2 offers the following features to support your
data resiliency:

975
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Data protection

• Copying AMIs across Regions

• Copying EBS snapshots across Regions
• Automating EBS-backed AMIs using Amazon Data Lifecycle Manager
• Automating EBS snapshots using Amazon Data Lifecycle Manager
• Maintaining the health and availability of your fleet using Amazon EC2 Auto Scaling
• Distributing incoming traffic across multiple instances in a single Availability Zone or multiple
Availability Zones using Elastic Load Balancing

Data protection in Amazon EC2

The AWS shared responsibility model applies to data protection in Amazon Elastic Compute Cloud.
As described in this model, AWS is responsible for protecting the global infrastructure that runs all
of the AWS Cloud. You are responsible for maintaining control over your content that is hosted on
this infrastructure. This content includes the security configuration and management tasks for the
AWS services that you use. For more information about data privacy, see the Data Privacy FAQ. For
information about data protection in Europe, see the AWS Shared Responsibility Model and GDPR blog
post on the AWS Security Blog.

For data protection purposes, we recommend that you protect AWS account credentials and set up
individual user accounts with AWS Identity and Access Management (IAM). That way each user is given
only the permissions necessary to fulfill their job duties. We also recommend that you secure your data
in the following ways:

• Use multi-factor authentication (MFA) with each account.

• Use SSL/TLS to communicate with AWS resources. We recommend TLS 1.2 or later.
• Set up API and user activity logging with AWS CloudTrail.
• Use AWS encryption solutions, along with all default security controls within AWS services.
• Use advanced managed security services such as Amazon Macie, which assists in discovering and
securing personal data that is stored in Amazon S3.
• If you require FIPS 140-2 validated cryptographic modules when accessing AWS through a command
line interface or an API, use a FIPS endpoint. For more information about the available FIPS endpoints,
see Federal Information Processing Standard (FIPS) 140-2.

We strongly recommend that you never put sensitive identifying information, such as your customers'
account numbers, into free-form fields such as a Name field. This includes when you work with Amazon
EC2 or other AWS services using the console, API, AWS CLI, or AWS SDKs. Any data that you enter into
Amazon EC2 or other services might get picked up for inclusion in diagnostic logs. When you provide a
URL to an external server, don't include credentials information in the URL to validate your request to
that server.

Encryption at rest
Amazon EBS encryption is an encryption solution for your EBS volumes and snapshots. It uses AWS Key
Management Service (AWS KMS) customer master keys (CMK). For more information, see Amazon EBS
encryption (p. 1176).

The data on NVMe instance store volumes is encrypted using an XTS-AES-256 cipher implemented on
a hardware module on the instance. The encryption keys are generated using the hardware module and
are unique to each NVMe instance storage device. All encryption keys are destroyed when the instance
is stopped or terminated and cannot be recovered. You cannot disable this encryption and you cannot
provide your own encryption key.

976
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Encryption in transit

Encryption in transit
SSH provides a secure communications channel for remote access to your Linux instances. Remote access
to your instances using AWS Systems Manager Session Manager and Run Command is encrypted using
TLS 1.2, and requests to create a connection are signed using SigV4.

Use an encryption protocol such as Transport Layer Security (TLS) to encrypt sensitive data in transit
between clients and your instances.

AWS provides secure and private connectivity between EC2 instances of all types. In addition, some
instance types use the offload capabilities of the underlying hardware to automatically encrypt in-
transit traffic between instances, using AEAD algorithms with 256-bit encryption. There is no impact on
network performance. The following requirements must be met to ensure the additional in-transit traffic
encryption:

• The instances use the following instance types: C5a, C5ad, C5n, C6gn, D3, D3en, G4ad, G4dn, I3en,
M5dn, M5n, M5zn, P3dn, P4d, R5dn, and R5n.
• The instances are in the same Region.
• The instances are in the same VPC or peered VPCs, and the traffic does not pass through a virtual
network device, such as a load balancer or a transit gateway.

Identity and access management for Amazon EC2

Your security credentials identify you to services in AWS and grant you unlimited use of your AWS
resources, such as your Amazon EC2 resources. You can use features of Amazon EC2 and AWS Identity
and Access Management (IAM) to allow other users, services, and applications to use your Amazon EC2
resources without sharing your security credentials. You can use IAM to control how other users use
resources in your AWS account, and you can use security groups to control access to your Amazon EC2
instances. You can choose to allow full use or limited use of your Amazon EC2 resources.

Contents
• Network access to your instance (p. 977)
• Amazon EC2 permission attributes (p. 977)
• IAM and Amazon EC2 (p. 978)
• IAM policies for Amazon EC2 (p. 979)
• IAM roles for Amazon EC2 (p. 1032)
• Authorize inbound traffic for your Linux instances (p. 1041)

Network access to your instance

A security group acts as a firewall that controls the traffic allowed to reach one or more instances. When
you launch an instance, you assign it one or more security groups. You add rules to each security group
that control traffic for the instance. You can modify the rules for a security group at any time; the new
rules are automatically applied to all instances to which the security group is assigned.

For more information, see Authorize inbound traffic for your Linux instances (p. 1041).

Amazon EC2 permission attributes

Your organization might have multiple AWS accounts. Amazon EC2 enables you to specify additional
AWS accounts that can use your Amazon Machine Images (AMIs) and Amazon EBS snapshots. These
permissions work at the AWS account level only; you can't restrict permissions for specific users within

977
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM and Amazon EC2

the specified AWS account. All users in the AWS account that you've specified can use the AMI or
snapshot.

Each AMI has a LaunchPermission attribute that controls which AWS accounts can access the AMI. For
more information, see Make an AMI public (p. 114).

Each Amazon EBS snapshot has a createVolumePermission attribute that controls which AWS
accounts can use the snapshot. For more information, see Share an Amazon EBS snapshot (p. 1139).

IAM and Amazon EC2

IAM enables you to do the following:

• Create users and groups under your AWS account

• Assign unique security credentials to each user under your AWS account
• Control each user's permissions to perform tasks using AWS resources
• Allow the users in another AWS account to share your AWS resources
• Create roles for your AWS account and define the users or services that can assume them
• Use existing identities for your enterprise to grant permissions to perform tasks using AWS resources

By using IAM with Amazon EC2, you can control whether users in your organization can perform a task
using specific Amazon EC2 API actions and whether they can use specific AWS resources.

This topic helps you answer the following questions:

• How do I create groups and users in IAM?

• How do I create a policy?
• What IAM policies do I need to carry out tasks in Amazon EC2?
• How do I grant permissions to perform actions in Amazon EC2?
• How do I grant permissions to perform actions on specific resources in Amazon EC2?

Create an IAM group and users

To create an IAM group

1. Open the IAM console at https://console.aws.amazon.com/iam/.

2. In the navigation pane, choose Groups and then choose Create New Group.
3. For Group Name, enter a name for your group, and then choose Next Step.
4. On the Attach Policy page, select an AWS managed policy and then choose Next Step. For example,
for Amazon EC2, one of the following AWS managed policies might meet your needs:

• PowerUserAccess
• ReadOnlyAccess
• AmazonEC2FullAccess
• AmazonEC2ReadOnlyAccess
5. Choose Create Group.

Your new group is listed under Group Name.

To create an IAM user, add the user to your group, and create a password for the user

1. In the navigation pane, choose Users, Add user.

978
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

2. For User name, enter a user name.

3. For Access type, select both Programmatic access and AWS Management Console access.
4. For Console password, choose one of the following:

• Autogenerated password. Each user gets a randomly generated password that meets the current
password policy in effect (if any). You can view or download the passwords when you get to the
Final page.
• Custom password. Each user is assigned the password that you enter in the box.
5. Choose Next: Permissions.
6. On the Set permissions page, choose Add user to group. Select the check box next to the group
that you created earlier and choose Next: Review.
7. Choose Create user.
8. To view the users' access keys (access key IDs and secret access keys), choose Show next to each
password and secret access key to see. To save the access keys, choose Download .csv and then save
the file to a safe location.
Important
You cannot retrieve the secret access key after you complete this step; if you misplace it you
must create a new one.
9. Choose Close.
10. Give each user his or her credentials (access keys and password); this enables them to use services
based on the permissions you specified for the IAM group.

Related topics
For more information about IAM, see the following:

• IAM policies for Amazon EC2 (p. 979)

• IAM roles for Amazon EC2 (p. 1032)
• AWS Identity and Access Management (IAM)
• IAM User Guide

IAM policies for Amazon EC2

By default, IAM users don't have permission to create or modify Amazon EC2 resources, or perform tasks
using the Amazon EC2 API. (This means that they also can't do so using the Amazon EC2 console or CLI.)
To allow IAM users to create or modify resources and perform tasks, you must create IAM policies that
grant IAM users permission to use the specific resources and API actions they'll need, and then attach
those policies to the IAM users or groups that require those permissions.

When you attach a policy to a user or group of users, it allows or denies the users permission to perform
the specified tasks on the specified resources. For more general information about IAM policies, see
Permissions and Policies in the IAM User Guide. For more information about managing and creating
custom IAM policies, see Managing IAM Policies.

Getting Started

An IAM policy must grant or deny permissions to use one or more Amazon EC2 actions. It must also
specify the resources that can be used with the action, which can be all resources, or in some cases,
specific resources. The policy can also include conditions that you apply to the resource.

Amazon EC2 partially supports resource-level permissions. This means that for some EC2 API actions,
you cannot specify which resource a user is allowed to work with for that action. Instead, you have to
allow users to work with all resources for that action.

979
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

Task Topic

Understand the basic structure of a policy Policy syntax (p. 980)

Define actions in your policy Actions for Amazon EC2 (p. 981)

Define specific resources in your policy Amazon Resource Names (ARNs) for Amazon
EC2 (p. 982)

Apply conditions to the use of the resources Condition keys for Amazon EC2 (p. 983)

Work with the available resource-level Actions, resources, and condition keys for Amazon
permissions for Amazon EC2 EC2

Test your policy Check that users have the required

permissions (p. 984)

Example policies for a CLI or SDK Example policies for working with the AWS CLI or
an AWS SDK (p. 987)

Example policies for the Amazon EC2 console Example policies for working in the Amazon EC2
console (p. 1023)

Policy structure
The following topics explain the structure of an IAM policy.

Contents
• Policy syntax (p. 980)
• Actions for Amazon EC2 (p. 981)
• Supported resource-level permissions for Amazon EC2 API actions (p. 981)
• Amazon Resource Names (ARNs) for Amazon EC2 (p. 982)
• Condition keys for Amazon EC2 (p. 983)
• Check that users have the required permissions (p. 984)

Policy syntax
An IAM policy is a JSON document that consists of one or more statements. Each statement is structured
as follows.

{
"Statement":[{
"Effect":"effect",
"Action":"action",
"Resource":"arn",
"Condition":{
"condition":{
"key":"value"
}
}
}
]
}

There are various elements that make up a statement:

980
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

• Effect: The effect can be Allow or Deny. By default, IAM users don't have permission to use resources
and API actions, so all requests are denied. An explicit allow overrides the default. An explicit deny
overrides any allows.
• Action: The action is the specific API action for which you are granting or denying permission. To learn
about specifying action, see Actions for Amazon EC2 (p. 981).
• Resource: The resource that's affected by the action. Some Amazon EC2 API actions allow you to
include specific resources in your policy that can be created or modified by the action. You specify
a resource using an Amazon Resource Name (ARN) or using the wildcard (*) to indicate that the
statement applies to all resources. For more information, see Supported resource-level permissions for
Amazon EC2 API actions (p. 981).
• Condition: Conditions are optional. They can be used to control when your policy is in effect. For
more information about specifying conditions for Amazon EC2, see Condition keys for Amazon
EC2 (p. 983).

For more information about example IAM policy statements for Amazon EC2, see Example policies for
working with the AWS CLI or an AWS SDK (p. 987).

Actions for Amazon EC2

In an IAM policy statement, you can specify any API action from any service that supports IAM.
For Amazon EC2, use the following prefix with the name of the API action: ec2:. For example:
ec2:RunInstances and ec2:CreateImage.

To specify multiple actions in a single statement, separate them with commas as follows:

"Action": ["ec2:action1", "ec2:action2"]

You can also specify multiple actions using wildcards. For example, you can specify all actions whose
name begins with the word "Describe" as follows:

"Action": "ec2:Describe*"

To specify all Amazon EC2 API actions, use the * wildcard as follows:

"Action": "ec2:*"

For a list of Amazon EC2 actions, see Actions in the Amazon EC2 API Reference.

Supported resource-level permissions for Amazon EC2 API actions

Resource-level permissions refers to the ability to specify which resources users are allowed to perform
actions on. Amazon EC2 has partial support for resource-level permissions. This means that for certain
Amazon EC2 actions, you can control when users are allowed to use those actions based on conditions
that have to be fulfilled, or specific resources that users are allowed to use. For example, you can grant
users permissions to launch instances, but only of a specific type, and only using a specific AMI.

To specify a resource in an IAM policy statement, use its Amazon Resource Name (ARN). For more
information about specifying the ARN value, see Amazon Resource Names (ARNs) for Amazon
EC2 (p. 982). If an API action does not support individual ARNs, you must use a wildcard (*) to specify
that all resources can be affected by the action.

To see tables that identify which Amazon EC2 API actions support resource-level permissions, and the
ARNs and condition keys that you can use in a policy, see Actions, resources, and condition keys for
Amazon EC2.

981
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

Keep in mind that you can apply tag-based resource-level permissions in the IAM policies you use for
Amazon EC2 API actions. This gives you better control over which resources a user can create, modify, or
use. For more information, see Grant permission to tag resources during creation (p. 984).

Amazon Resource Names (ARNs) for Amazon EC2

Each IAM policy statement applies to the resources that you specify using their ARNs.

An ARN has the following general syntax:

arn:aws:[service]:[region]:[account]:resourceType/resourcePath

service

The service (for example, ec2).

region

The Region for the resource (for example, us-east-1).

account

The AWS account ID, with no hyphens (for example, 123456789012).

resourceType

The type of resource (for example, instance).

resourcePath

A path that identifies the resource. You can use the * wildcard in your paths.

For example, you can indicate a specific instance (i-1234567890abcdef0) in your statement using its
ARN as follows.

"Resource": "arn:aws:ec2:us-east-1:123456789012:instance/i-1234567890abcdef0"

You can specify all instances that belong to a specific account by using the * wildcard as follows.

"Resource": "arn:aws:ec2:us-east-1:123456789012:instance/*"

You can also specify all Amazon EC2 resources that belong to a specific account by using the * wildcard
as follows.

"Resource": "arn:aws:ec2:us-east-1:123456789012:*"

To specify all resources, or if a specific API action does not support ARNs, use the * wildcard in the
Resource element as follows.

"Resource": "*"

Many Amazon EC2 API actions involve multiple resources. For example, AttachVolume attaches an
Amazon EBS volume to an instance, so an IAM user must have permissions to use the volume and the
instance. To specify multiple resources in a single statement, separate their ARNs with commas, as
follows.

"Resource": ["arn1", "arn2"]

982
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

For a list of ARNs for Amazon EC2 resources, see Resource types defined by Amazon EC2.

Condition keys for Amazon EC2

In a policy statement, you can optionally specify conditions that control when it is in effect. Each
condition contains one or more key-value pairs. Condition keys are not case-sensitive. We've defined
AWS-wide condition keys, plus additional service-specific condition keys.

For a list of service-specific condition keys for Amazon EC2, see Condition keys for Amazon EC2. Amazon
EC2 also implements the AWS-wide condition keys. For more information, see Information Available in
All Requests in the IAM User Guide.

To use a condition key in your IAM policy, use the Condition statement. For example, the following
policy grants users permission to add and remove inbound and outbound rules for any security group. It
uses the ec2:Vpc condition key to specify that these actions can only be performed on security groups
in a specific VPC.

{
"Version": "2012-10-17",
"Statement":[{
"Effect":"Allow",
"Action": [
"ec2:AuthorizeSecurityGroupIngress",
"ec2:AuthorizeSecurityGroupEgress",
"ec2:RevokeSecurityGroupIngress",
"ec2:RevokeSecurityGroupEgress"],
"Resource": "arn:aws:ec2:region:account:security-group/*",
"Condition": {
"StringEquals": {
"ec2:Vpc": "arn:aws:ec2:region:account:vpc/vpc-11223344556677889"
}
}
}
]
}

If you specify multiple conditions, or multiple keys in a single condition, we evaluate them using a
logical AND operation. If you specify a single condition with multiple values for one key, we evaluate the
condition using a logical OR operation. For permissions to be granted, all conditions must be met.

You can also use placeholders when you specify conditions. For example, you can grant an IAM user
permission to use resources with a tag that specifies his or her IAM user name. For more information, see
Policy Variables in the IAM User Guide.
Important
Many condition keys are specific to a resource, and some API actions use multiple resources.
If you write a policy with a condition key, use the Resource element of the statement to
specify the resource to which the condition key applies. If not, the policy may prevent users
from performing the action at all, because the condition check fails for the resources to which
the condition key does not apply. If you do not want to specify a resource, or if you've written
the Action element of your policy to include multiple API actions, then you must use the
...IfExists condition type to ensure that the condition key is ignored for resources that do
not use it. For more information, see ...IfExists Conditions in the IAM User Guide.

All Amazon EC2 actions support the aws:RequestedRegion and ec2:Region condition keys. For more
information, see Example: Restrict access to a specific Region (p. 988).

The ec2:SourceInstanceARN key can be used for conditions that specify the ARN of the instance
from which a request is made. This condition key is available AWS-wide and is not service-specific. For
policy examples, see Allows an EC2 Instance to Attach or Detach Volumes and Example: Allow a specific

983
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

instance to view resources in other AWS services (p. 1019). The ec2:SourceInstanceARN key cannot
be used as a variable to populate the ARN for the Resource element in a statement.

For example policy statements for Amazon EC2, see Example policies for working with the AWS CLI or an
AWS SDK (p. 987).

Check that users have the required permissions

After you've created an IAM policy, we recommend that you check whether it grants users the
permissions to use the particular API actions and resources they need before you put the policy into
production.

First, create an IAM user for testing purposes, and then attach the IAM policy that you created to the test
user. Then, make a request as the test user.

If the Amazon EC2 action that you are testing creates or modifies a resource, you should make the
request using the DryRun parameter (or run the AWS CLI command with the --dry-run option). In
this case, the call completes the authorization check, but does not complete the operation. For example,
you can check whether the user can terminate a particular instance without actually terminating it. If
the test user has the required permissions, the request returns DryRunOperation; otherwise, it returns
UnauthorizedOperation.

If the policy doesn't grant the user the permissions that you expected, or is overly permissive, you can
adjust the policy as needed and retest until you get the desired results.
Important
It can take several minutes for policy changes to propagate before they take effect. Therefore,
we recommend that you allow five minutes to pass before you test your policy updates.

If an authorization check fails, the request returns an encoded message with diagnostic information. You
can decode the message using the DecodeAuthorizationMessage action. For more information, see
DecodeAuthorizationMessage in the AWS Security Token Service API Reference, and decode-authorization-
message in the AWS CLI Command Reference.

Grant permission to tag resources during creation

Some resource-creating Amazon EC2 API actions enable you to specify tags when you create the
resource. For more information, see Tag your resources (p. 1313).

To enable users to tag resources on creation, they must have permissions to use the action that creates
the resource, such as ec2:RunInstances or ec2:CreateVolume. If tags are specified in the resource-
creating action, Amazon performs additional authorization on the ec2:CreateTags action to verify
if users have permissions to create tags. Therefore, users must also have explicit permissions to use the
ec2:CreateTags action.

In the IAM policy definition for the ec2:CreateTags action, use the Condition element with the
ec2:CreateAction condition key to give tagging permissions to the action that creates the resource.

The following example demonstrates a policy that allows users to launch instances and apply any tags to
instances and volumes during launch. Users are not permitted to tag any existing resources (they cannot
call the ec2:CreateTags action directly).

{
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": "*"

984
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

},
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],
"Resource": "arn:aws:ec2:region:account:*/*",
"Condition": {
"StringEquals": {
"ec2:CreateAction" : "RunInstances"
}
}
}
]
}

Similarly, the following policy allows users to create volumes and apply any tags to the volumes
during volume creation. Users are not permitted to tag any existing resources (they cannot call the
ec2:CreateTags action directly).

{
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:CreateVolume"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],
"Resource": "arn:aws:ec2:region:account:*/*",
"Condition": {
"StringEquals": {
"ec2:CreateAction" : "CreateVolume"
}
}
}
]
}

The ec2:CreateTags action is only evaluated if tags are applied during the resource-creating action.
Therefore, a user that has permissions to create a resource (assuming there are no tagging conditions)
does not require permissions to use the ec2:CreateTags action if no tags are specified in the request.
However, if the user attempts to create a resource with tags, the request fails if the user does not have
permissions to use the ec2:CreateTags action.

The ec2:CreateTags action is also evaluated if tags are provided in a launch template. For an example
policy, see Tags in a launch template (p. 1007).

Control access to specific tags

You can use additional conditions in the Condition element of your IAM policies to control the tag keys
and values that can be applied to resources.

The following condition keys can be used with the examples in the preceding section:

• aws:RequestTag: To indicate that a particular tag key or tag key and value must be present in a
request. Other tags can also be specified in the request.

985
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

• Use with the StringEquals condition operator to enforce a specific tag key and value combination,
for example, to enforce the tag cost-center=cc123:

"StringEquals": { "aws:RequestTag/cost-center": "cc123" }

• Use with the StringLike condition operator to enforce a specific tag key in the request; for
example, to enforce the tag key purpose:

"StringLike": { "aws:RequestTag/purpose": "*" }

• aws:TagKeys: To enforce the tag keys that are used in the request.
• Use with the ForAllValues modifier to enforce specific tag keys if they are provided in the request
(if tags are specified in the request, only specific tag keys are allowed; no other tags are allowed). For
example, the tag keys environment or cost-center are allowed:

"ForAllValues:StringEquals": { "aws:TagKeys": ["environment","cost-center"] }

• Use with the ForAnyValue modifier to enforce the presence of at least one of the specified tag
keys in the request. For example, at least one of the tag keys environment or webserver must be
present in the request:

"ForAnyValue:StringEquals": { "aws:TagKeys": ["environment","webserver"] }

These condition keys can be applied to resource-creating actions that support tagging, as well as the
ec2:CreateTags and ec2:DeleteTags actions. To learn whether an Amazon EC2 API action supports
tagging, see Actions, resources, and condition keys for Amazon EC2.

To force users to specify tags when they create a resource, you must use the aws:RequestTag condition
key or the aws:TagKeys condition key with the ForAnyValue modifier on the resource-creating action.
The ec2:CreateTags action is not evaluated if a user does not specify tags for the resource-creating
action.

For conditions, the condition key is not case-sensitive and the condition value is case-sensitive. Therefore,
to enforce the case-sensitivity of a tag key, use the aws:TagKeys condition key, where the tag key is
specified as a value in the condition.

For example IAM policies, see Example policies for working with the AWS CLI or an AWS SDK (p. 987).
For more information about multi-value conditions, see Creating a Condition That Tests Multiple Key
Values in the IAM User Guide.

Control access to EC2 resources using resource tags

When you create an IAM policy that grants IAM users permission to use EC2 resources, you can include
tag information in the Condition element of the policy to control access based on tags. This gives you
better control over which EC2 resources a user can modify, use, or delete.

For example, you can create a policy that allows users to terminate an instance but denies the action
if the instance has the tag environment=production. To do this, you use the ec2:ResourceTag
condition key to allow or deny access to the resource based on the tags that are attached to the resource.

"StringEquals": { "ec2:ResourceTag/environment": "production" }

To learn whether an Amazon EC2 API action supports controlling access using the ec2:ResourceTag
condition key, see Actions, resources, and condition keys for Amazon EC2. Note that the Describe

986
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

actions do not support resource-level permissions, so you must specify them in a separate statement
without conditions.

For example IAM policies, see Example policies for working with the AWS CLI or an AWS SDK (p. 987).

If you allow or deny users access to resources based on tags, you must consider explicitly denying users
the ability to add those tags to or remove them from the same resources. Otherwise, it's possible for a
user to circumvent your restrictions and gain access to a resource by modifying its tags.

Example policies for working with the AWS CLI or an AWS SDK
The following examples show policy statements that you could use to control the permissions that IAM
users have to Amazon EC2. These policies are designed for requests that are made with the AWS CLI
or an AWS SDK. For example policies for working in the Amazon EC2 console, see Example policies for
working in the Amazon EC2 console (p. 1023). For examples of IAM policies specific to Amazon VPC, see
Identity and Access Management for Amazon VPC.

Examples
• Example: Read-only access (p. 987)
• Example: Restrict access to a specific Region (p. 988)
• Work with instances (p. 988)
• Work with volumes (p. 990)
• Work with snapshots (p. 992)
• Launch instances (RunInstances) (p. 999)
• Work with Spot Instances (p. 1011)
• Example: Work with Reserved Instances (p. 1015)
• Example: Tag resources (p. 1016)
• Example: Work with IAM roles (p. 1018)
• Example: Work with route tables (p. 1019)
• Example: Allow a specific instance to view resources in other AWS services (p. 1019)
• Example: Work with launch templates (p. 1020)
• Work with instance metadata (p. 1021)

Example: Read-only access

The following policy grants users permissions to use all Amazon EC2 API actions whose names begin
with Describe. The Resource element uses a wildcard to indicate that users can specify all resources
with these API actions. The * wildcard is also necessary in cases where the API action does not support
resource-level permissions. For more information about which ARNs you can use with which Amazon EC2
API actions, see Actions, resources, and condition keys for Amazon EC2.

Users don't have permission to perform any actions on the resources (unless another statement grants
them permission to do so) because they're denied permission to use API actions by default.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:Describe*",
"Resource": "*"
}
]
}

987
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

Example: Restrict access to a specific Region

The following policy denies users permission to use all Amazon EC2 API actions unless the Region is
Europe (Frankfurt). It uses the global condition key aws:RequestedRegion, which is supported by all
Amazon EC2 API actions.

{
"Version":"2012-10-17",
"Statement": [
{
"Effect": "Deny",
"Action": "ec2:*",
"Resource": "*",
"Condition": {
"StringNotEquals": {
"aws:RequestedRegion": "eu-central-1"
}
}
}
]
}

Alternatively, you can use the condition key ec2:Region, which is specific to Amazon EC2 and is
supported by all Amazon EC2 API actions.

{
"Version":"2012-10-17",
"Statement": [
{
"Effect": "Deny",
"Action": "ec2:*",
"Resource": "*",
"Condition": {
"StringNotEquals": {
"ec2:Region": "eu-central-1"
}
}
}
]
}

Work with instances

Examples
• Example: Describe, launch, stop, start, and terminate all instances (p. 988)
• Example: Describe all instances, and stop, start, and terminate only particular instances (p. 989)

Example: Describe, launch, stop, start, and terminate all instances

The following policy grants users permissions to use the API actions specified in the Action element.
The Resource element uses a * wildcard to indicate that users can specify all resources with these API
actions. The * wildcard is also necessary in cases where the API action does not support resource-level
permissions. For more information about which ARNs you can use with which Amazon EC2 API actions,
see Actions, resources, and condition keys for Amazon EC2.

The users don't have permission to use any other API actions (unless another statement grants them
permission to do so) because users are denied permission to use API actions by default.

988
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:DescribeInstances",
"ec2:DescribeImages",
"ec2:DescribeKeyPairs",
"ec2:DescribeSecurityGroups",
"ec2:DescribeAvailabilityZones",
"ec2:RunInstances",
"ec2:TerminateInstances",
"ec2:StopInstances",
"ec2:StartInstances"
],
"Resource": "*"
}
]
}

Example: Describe all instances, and stop, start, and terminate only particular instances

The following policy allows users to describe all instances, to start and stop only instances
i-1234567890abcdef0 and i-0598c7d356eba48d7, and to terminate only instances in the US East (N.
Virginia) Region (us-east-1) with the resource tag "purpose=test".

The first statement uses a * wildcard for the Resource element to indicate that users can
specify all resources with the action; in this case, they can list all instances. The * wildcard is also
necessary in cases where the API action does not support resource-level permissions (in this case,
ec2:DescribeInstances). For more information about which ARNs you can use with which Amazon
EC2 API actions, see Actions, resources, and condition keys for Amazon EC2.

The second statement uses resource-level permissions for the StopInstances and StartInstances
actions. The specific instances are indicated by their ARNs in the Resource element.

The third statement allows users to terminate all instances in the US East (N. Virginia) Region
(us-east-1) that belong to the specified AWS account, but only where the instance has the tag
"purpose=test". The Condition element qualifies when the policy statement is in effect.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:DescribeInstances",
"Resource": "*"
},
{
"Effect": "Allow",
"Action": [
"ec2:StopInstances",
"ec2:StartInstances"
],
"Resource": [
"arn:aws:ec2:us-east-1:123456789012:instance/i-1234567890abcdef0",
"arn:aws:ec2:us-east-1:123456789012:instance/i-0598c7d356eba48d7"
]
},
{
"Effect": "Allow",
"Action": "ec2:TerminateInstances",
"Resource": "arn:aws:ec2:us-east-1:123456789012:instance/*",

989
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Condition": {
"StringEquals": {
"ec2:ResourceTag/purpose": "test"
}
}
}

]
}

Work with volumes

Examples
• Example: Attach and detach volumes (p. 990)
• Example: Create a volume (p. 991)
• Example: Create a volume with tags (p. 991)

Example: Attach and detach volumes

When an API action requires a caller to specify multiple resources, you must create a policy statement
that allows users to access all required resources. If you need to use a Condition element with one or
more of these resources, you must create multiple statements as shown in this example.

The following policy allows users to attach volumes with the tag "volume_user=iam-user-name" to
instances with the tag "department=dev", and to detach those volumes from those instances. If you
attach this policy to an IAM group, the aws:username policy variable gives each IAM user in the group
permission to attach or detach volumes from the instances with a tag named volume_user that has his
or her IAM user name as a value.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:AttachVolume",
"ec2:DetachVolume"
],
"Resource": "arn:aws:ec2:us-east-1:123456789012:instance/*",
"Condition": {
"StringEquals": {
"ec2:ResourceTag/department": "dev"
}
}
},
{
"Effect": "Allow",
"Action": [
"ec2:AttachVolume",
"ec2:DetachVolume"
],
"Resource": "arn:aws:ec2:us-east-1:123456789012:volume/*",
"Condition": {
"StringEquals": {
"ec2:ResourceTag/volume_user": "${aws:username}"
}
}
}
]
}

990
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

Example: Create a volume

The following policy allows users to use the CreateVolume API action. The user is allowed to create a
volume only if the volume is encrypted and only if the volume size is less than 20 GiB.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:CreateVolume"
],
"Resource": "arn:aws:ec2:us-east-1:123456789012:volume/*",
"Condition":{
"NumericLessThan": {
"ec2:VolumeSize" : "20"
},
"Bool":{
"ec2:Encrypted" : "true"
}
}
}
]
}

Example: Create a volume with tags

The following policy includes the aws:RequestTag condition key that requires users to tag any volumes
they create with the tags costcenter=115 and stack=prod. The aws:TagKeys condition key uses
the ForAllValues modifier to indicate that only the keys costcenter and stack are allowed in the
request (no other tags can be specified). If users don't pass these specific tags, or if they don't specify
tags at all, the request fails.

For resource-creating actions that apply tags, users must also have permissions to use the CreateTags
action. The second statement uses the ec2:CreateAction condition key to allow users to create tags
only in the context of CreateVolume. Users cannot tag existing volumes or any other resources. For
more information, see Grant permission to tag resources during creation (p. 984).

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowCreateTaggedVolumes",
"Effect": "Allow",
"Action": "ec2:CreateVolume",
"Resource": "arn:aws:ec2:us-east-1:123456789012:volume/*",
"Condition": {
"StringEquals": {
"aws:RequestTag/costcenter": "115",
"aws:RequestTag/stack": "prod"
},
"ForAllValues:StringEquals": {
"aws:TagKeys": ["costcenter","stack"]
}
}
},
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],

991
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Resource": "arn:aws:ec2:us-east-1:123456789012:volume/*",
"Condition": {
"StringEquals": {
"ec2:CreateAction" : "CreateVolume"
}
}
}
]
}

The following policy allows users to create a volume without having to specify tags. The CreateTags
action is only evaluated if tags are specified in the CreateVolume request. If users do specify tags, the
tag must be purpose=test. No other tags are allowed in the request.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:CreateVolume",
"Resource": "*"
},
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],
"Resource": "arn:aws:ec2:us-east-1:1234567890:volume/*",
"Condition": {
"StringEquals": {
"aws:RequestTag/purpose": "test",
"ec2:CreateAction" : "CreateVolume"
},
"ForAllValues:StringEquals": {
"aws:TagKeys": "purpose"
}
}
}
]
}

Work with snapshots

The following are example policies for both CreateSnapshot (point-in-time snapshot of an EBS
volume) and CreateSnapshots (multi-volume snapshots).

Examples
• Example: Create a snapshot (p. 992)
• Example: Create snapshots (p. 993)
• Example: Create a snapshot with tags (p. 993)
• Example: Create snapshots with tags (p. 994)
• Example: Modify permission settings for snapshots (p. 999)

Example: Create a snapshot

The following policy allows customers to use the CreateSnapshot API action. The customer can create
snapshots only if the volume is encrypted and only if the volume size is less than 20 GiB.

992
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshot",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*"
},
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshot",
"Resource":"arn:aws:ec2:us-east-1:123456789012:volume/*",
"Condition":{
"NumericLessThan":{
"ec2:VolumeSize":"20"
},
"Bool":{
"ec2:Encrypted":"true"
}
}
}
]
}

Example: Create snapshots

The following policy allows customers to use the CreateSnapshots API action. The customer can create
snapshots only if all of the volumes on the instance are type GP2.

{
"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshots",
"Resource":[
"arn:aws:ec2:us-east-1::snapshot/*",
"arn:aws:ec2:*:*:instance/*"
]
},
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshots",
"Resource":"arn:aws:ec2:us-east-1:*:volume/*",
"Condition":{
"StringLikeIfExists":{
"ec2:VolumeType":"gp2"
}
}

}
]
}

Example: Create a snapshot with tags

The following policy includes the aws:RequestTag condition key that requires the customer to apply
the tags costcenter=115 and stack=prod to any new snapshot. The aws:TagKeys condition
key uses the ForAllValues modifier to indicate that only the keys costcenter and stack can be
specified in the request. The request fails if either of these conditions is not met.

For resource-creating actions that apply tags, customers must also have permissions to use the
CreateTags action. The third statement uses the ec2:CreateAction condition key to allow
customers to create tags only in the context of CreateSnapshot. Customers cannot tag existing

993
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

volumes or any other resources. For more information, see Grant permission to tag resources during
creation (p. 984).

{
"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshot",
"Resource":"arn:aws:ec2:us-east-1:123456789012:volume/*"
},
{
"Sid":"AllowCreateTaggedSnapshots",
"Effect":"Allow",
"Action":"ec2:CreateSnapshot",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"StringEquals":{
"aws:RequestTag/costcenter":"115",
"aws:RequestTag/stack":"prod"
},
"ForAllValues:StringEquals":{
"aws:TagKeys":[
"costcenter",
"stack"
]
}
}
},
{
"Effect":"Allow",
"Action":"ec2:CreateTags",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"StringEquals":{
"ec2:CreateAction":"CreateSnapshot"
}
}
}
]
}

Example: Create snapshots with tags

The following policy includes the aws:RequestTag condition key that requires the customer to apply
the tags costcenter=115 and stack=prod to any new snapshot. The aws:TagKeys condition
key uses the ForAllValues modifier to indicate that only the keys costcenter and stack can be
specified in the request. The request fails if either of these conditions is not met.

{
"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshots",
"Resource":[
"arn:aws:ec2:us-east-1::snapshot/*",
"arn:aws:ec2:*:*:instance/*",
"arn:aws:ec2:*:*:volume/*"

]
},
{

994
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Sid":"AllowCreateTaggedSnapshots",
"Effect":"Allow",
"Action":"ec2:CreateSnapshots",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"StringEquals":{
"aws:RequestTag/costcenter":"115",
"aws:RequestTag/stack":"prod"
},
"ForAllValues:StringEquals":{
"aws:TagKeys":[
"costcenter",
"stack"
]
}
}
},
{
"Effect":"Allow",
"Action":"ec2:CreateTags",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"StringEquals":{
"ec2:CreateAction":"CreateSnapshots"
}
}
}
]
}

The following policy allows customers to create a snapshot without having to specify tags.
The CreateTags action is evaluated only if tags are specified in the CreateSnapshot or
CreateSnapshots request. If a tag is specified, the tag must be purpose=test. No other tags are
allowed in the request.

{
"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshot",
"Resource":"*"
},
{
"Effect":"Allow",
"Action":"ec2:CreateTags",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"StringEquals":{
"aws:RequestTag/purpose":"test",
"ec2:CreateAction":"CreateSnapshot"
},
"ForAllValues:StringEquals":{
"aws:TagKeys":"purpose"
}
}
}
]
}

{
"Version":"2012-10-17",
"Statement": [

995
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

{
"Effect":"Allow",
"Action":"ec2:CreateSnapshots",
"Resource":"*"
},
{
"Effect":"Allow",
"Action":"ec2:CreateTags",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"StringEquals":{
"aws:RequestTag/purpose":"test",
"ec2:CreateAction":"CreateSnapshots"
},
"ForAllValues:StringEquals":{
"aws:TagKeys":"purpose"
}
}
}
]
}

The following policy allows snapshots to be created only if the source volume is tagged with
User:username for the customer, and the snapshot itself is tagged with Environment:Dev and
User:username. The customer can add additional tags to the snapshot.

{
"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshot",
"Resource":"arn:aws:ec2:us-east-1:123456789012:volume/*",
"Condition":{
"StringEquals":{
"ec2:ResourceTag/User":"${aws:username}"
}
}
},
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshot",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"StringEquals":{
"aws:RequestTag/Environment":"Dev",
"aws:RequestTag/User":"${aws:username}"
}
}
},
{
"Effect":"Allow",
"Action":"ec2:CreateTags",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*"
}
]
}

The following policy for CreateSnapshots allows snapshots to be created only if the source
volume is tagged with User:username for the customer, and the snapshot itself is tagged with
Environment:Dev and User:username.

996
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshots",
"Resource":"arn:aws:ec2:us-east-1:*:instance/*",
},
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshots",
"Resource":"arn:aws:ec2:us-east-1:123456789012:volume/*",
"Condition":{
"StringEquals":{
"ec2:ResourceTag/User":"${aws:username}"
}
}
},
{
"Effect":"Allow",
"Action":"ec2:CreateSnapshots",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"StringEquals":{
"aws:RequestTag/Environment":"Dev",
"aws:RequestTag/User":"${aws:username}"
}
}
},
{
"Effect":"Allow",
"Action":"ec2:CreateTags",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*"
}
]
}

The following policy allows deletion of a snapshot only if the snapshot is tagged with User:username for
the customer.

{
"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",
"Action":"ec2:DeleteSnapshot",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"StringEquals":{
"ec2:ResourceTag/User":"${aws:username}"
}
}
}
]
}

The following policy allows a customer to create a snapshot but denies the action if the snapshot being
created has a tag key value=stack.

{
"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",

997
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Action":[
"ec2:CreateSnapshot",
"ec2:CreateTags"
],
"Resource":"*"
},
{
"Effect":"Deny",
"Action":"ec2:CreateSnapshot",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"ForAnyValue:StringEquals":{
"aws:TagKeys":"stack"
}
}
}
]
}

The following policy allows a customer to create snapshots but denies the action if the snapshots being
created have a tag key value=stack.

{
"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",
"Action":[
"ec2:CreateSnapshots",
"ec2:CreateTags"
],
"Resource":"*"
},
{
"Effect":"Deny",
"Action":"ec2:CreateSnapshots",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"ForAnyValue:StringEquals":{
"aws:TagKeys":"stack"
}
}
}
]
}

The following policy allows you to combine multiple actions into a single policy. You can only create a
snapshot (in the context of CreateSnapshots) when the snapshot is created in Region us-east-1. You
can only create snapshots (in the context of CreateSnapshots) when the snapshots are being created
in the Region us-east-1 and when the instance type is t2*.

{
"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",
"Action":[
"ec2:CreateSnapshots",
"ec2:CreateSnapshot",
"ec2:CreateTags"
],
"Resource": [
"arn:aws:ec2:*:*:instance/*",

998
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"arn:aws:ec2:*:*:snapshot/*",
"arn:aws:ec2:*:*:volume/*"
],
"Condition":{
"StringEqualsIgnoreCase": {
"ec2:Region": "us-east-1"
},
"StringLikeIfExists": {
"ec2:InstanceType": ["t2.*"]
}
}
}
]
}

Example: Modify permission settings for snapshots

The following policy allows modification of a snapshot only if the snapshot is tagged with
User:username, where username is the customer's AWS account user name. The request fails if this
condition is not met.

{
"Version":"2012-10-17",
"Statement": [
{
"Effect":"Allow",
"Action":"ec2: ModifySnapshotAttribute",
"Resource":"arn:aws:ec2:us-east-1::snapshot/*",
"Condition":{
"StringEquals":{
"ec2:ResourceTag/user-name":"${aws:username}"
}
}
}
]
}

Launch instances (RunInstances)

The RunInstances API action launches one or more On-Demand Instances or one or more Spot Instances.
RunInstances requires an AMI and creates an instance. Users can specify a key pair and security group
in the request. Launching into a VPC requires a subnet, and creates a network interface. Launching from
an Amazon EBS-backed AMI creates a volume. Therefore, the user must have permissions to use these
Amazon EC2 resources. You can create a policy statement that requires users to specify an optional
parameter on RunInstances, or restricts users to particular values for a parameter.

For more information about the resource-level permissions that are required to launch an instance, see
Actions, resources, and condition keys for Amazon EC2.

By default, users don't have permissions to describe, start, stop, or terminate the resulting instances. One
way to grant the users permission to manage the resulting instances is to create a specific tag for each
instance, and then create a statement that enables them to manage instances with that tag. For more
information, see Work with instances (p. 988).

Resources
• AMIs (p. 1000)
• Instance types (p. 1001)
• Subnets (p. 1002)
• EBS volumes (p. 1003)

999
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

• Tags (p. 1003)

• Tags in a launch template (p. 1007)
• Elastic GPUs (p. 1008)
• Launch templates (p. 1009)

AMIs

The following policy allows users to launch instances using only the specified AMIs, ami-9e1670f7 and
ami-45cf5c3c. The users can't launch an instance using other AMIs (unless another statement grants
the users permission to do so).

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:region::image/ami-9e1670f7",
"arn:aws:ec2:region::image/ami-45cf5c3c",
"arn:aws:ec2:region:account:instance/*",
"arn:aws:ec2:region:account:volume/*",
"arn:aws:ec2:region:account:key-pair/*",
"arn:aws:ec2:region:account:security-group/*",
"arn:aws:ec2:region:account:subnet/*",
"arn:aws:ec2:region:account:network-interface/*"
]
}
]
}

Alternatively, the following policy allows users to launch instances from all AMIs owned by Amazon. The
Condition element of the first statement tests whether ec2:Owner is amazon. The users can't launch
an instance using other AMIs (unless another statement grants the users permission to do so).

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:region::image/ami-*"
],
"Condition": {
"StringEquals": {
"ec2:Owner": "amazon"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:region:account:instance/*",
"arn:aws:ec2:region:account:subnet/*",
"arn:aws:ec2:region:account:volume/*",
"arn:aws:ec2:region:account:network-interface/*",
"arn:aws:ec2:region:account:key-pair/*",
"arn:aws:ec2:region:account:security-group/*"
]

1000
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

}
]
}

Instance types
The following policy allows users to launch instances using only the t2.micro or t2.small instance
type, which you might do to control costs. The users can't launch larger instances because the
Condition element of the first statement tests whether ec2:InstanceType is either t2.micro or
t2.small.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:region:account:instance/*"
],
"Condition": {
"StringEquals": {
"ec2:InstanceType": ["t2.micro", "t2.small"]
}
}
},
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:region::image/ami-*",
"arn:aws:ec2:region:account:subnet/*",
"arn:aws:ec2:region:account:network-interface/*",
"arn:aws:ec2:region:account:volume/*",
"arn:aws:ec2:region:account:key-pair/*",
"arn:aws:ec2:region:account:security-group/*"
]
}
]
}

Alternatively, you can create a policy that denies users permissions to launch any instances except
t2.micro and t2.small instance types.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Deny",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:region:account:instance/*"
],
"Condition": {
"StringNotEquals": {
"ec2:InstanceType": ["t2.micro", "t2.small"]
}
}
},
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [

1001
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"arn:aws:ec2:region::image/ami-*",
"arn:aws:ec2:region:account:network-interface/*",
"arn:aws:ec2:region:account:instance/*",
"arn:aws:ec2:region:account:subnet/*",
"arn:aws:ec2:region:account:volume/*",
"arn:aws:ec2:region:account:key-pair/*",
"arn:aws:ec2:region:account:security-group/*"
]
}
]
}

Subnets

The following policy allows users to launch instances using only the specified subnet,
subnet-12345678. The group can't launch instances into any another subnet (unless another statement
grants the users permission to do so).

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:region:account:subnet/subnet-12345678",
"arn:aws:ec2:region:account:network-interface/*",
"arn:aws:ec2:region:account:instance/*",
"arn:aws:ec2:region:account:volume/*",
"arn:aws:ec2:region::image/ami-*",
"arn:aws:ec2:region:account:key-pair/*",
"arn:aws:ec2:region:account:security-group/*"
]
}
]
}

Alternatively, you could create a policy that denies users permissions to launch an instance into any other
subnet. The statement does this by denying permission to create a network interface, except where
subnet subnet-12345678 is specified. This denial overrides any other policies that are created to allow
launching instances into other subnets.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Deny",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:region:account:network-interface/*"
],
"Condition": {
"ArnNotEquals": {
"ec2:Subnet": "arn:aws:ec2:region:account:subnet/subnet-12345678"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:region::image/ami-*",

1002
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"arn:aws:ec2:region:account:network-interface/*",
"arn:aws:ec2:region:account:instance/*",
"arn:aws:ec2:region:account:subnet/*",
"arn:aws:ec2:region:account:volume/*",
"arn:aws:ec2:region:account:key-pair/*",
"arn:aws:ec2:region:account:security-group/*"
]
}
]
}

EBS volumes

The following policy allows users to launch instances only if the EBS volumes for the instance are
encrypted. The user must launch an instance from an AMI that was created with encrypted snapshots, to
ensure that the root volume is encrypted. Any additional volume that the user attaches to the instance
during launch must also be encrypted.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:*:*:volume/*"
],
"Condition": {
"Bool": {
"ec2:Encrypted": "true"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:*::image/ami-*",
"arn:aws:ec2:*:*:network-interface/*",
"arn:aws:ec2:*:*:instance/*",
"arn:aws:ec2:*:*:subnet/*",
"arn:aws:ec2:*:*:key-pair/*",
"arn:aws:ec2:*:*:security-group/*"
]
}
]
}

Tags

Tag instances on creation

The following policy allows users to launch instances and tag the instances during creation. For resource-
creating actions that apply tags, users must have permissions to use the CreateTags action. The second
statement uses the ec2:CreateAction condition key to allow users to create tags only in the context
of RunInstances, and only for instances. Users cannot tag existing resources, and users cannot tag
volumes using the RunInstances request.

For more information, see Grant permission to tag resources during creation (p. 984).

1003
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],
"Resource": "arn:aws:ec2:us-east-1:123456789012:instance/*",
"Condition": {
"StringEquals": {
"ec2:CreateAction" : "RunInstances"
}
}
}
]
}

Tag instances and volumes on creation with specific tags

The following policy includes the aws:RequestTag condition key that requires users to tag any
instances and volumes that are created by RunInstances with the tags environment=production
and purpose=webserver. The aws:TagKeys condition key uses the ForAllValues modifier to
indicate that only the keys environment and purpose are allowed in the request (no other tags can be
specified). If no tags are specified in the request, the request fails.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:region::image/*",
"arn:aws:ec2:region:account:subnet/*",
"arn:aws:ec2:region:account:network-interface/*",
"arn:aws:ec2:region:account:security-group/*",
"arn:aws:ec2:region:account:key-pair/*"
]
},
{
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:region:account:volume/*",
"arn:aws:ec2:region:account:instance/*"
],
"Condition": {
"StringEquals": {
"aws:RequestTag/environment": "production" ,
"aws:RequestTag/purpose": "webserver"
},
"ForAllValues:StringEquals": {
"aws:TagKeys": ["environment","purpose"]

1004
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

}
}
},
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],
"Resource": "arn:aws:ec2:region:account:*/*",
"Condition": {
"StringEquals": {
"ec2:CreateAction" : "RunInstances"
}
}
}
]
}

Tag instances and volumes on creation with at least one specific tag

The following policy uses the ForAnyValue modifier on the aws:TagKeys condition to indicate that at
least one tag must be specified in the request, and it must contain the key environment or webserver.
The tag must be applied to both instances and volumes. Any tag values can be specified in the request.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:region::image/*",
"arn:aws:ec2:region:account:subnet/*",
"arn:aws:ec2:region:account:network-interface/*",
"arn:aws:ec2:region:account:security-group/*",
"arn:aws:ec2:region:account:key-pair/*"
]
},
{
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:region:account:volume/*",
"arn:aws:ec2:region:account:instance/*"
],
"Condition": {
"ForAnyValue:StringEquals": {
"aws:TagKeys": ["environment","webserver"]
}
}
},
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],
"Resource": "arn:aws:ec2:region:account:*/*",
"Condition": {
"StringEquals": {
"ec2:CreateAction" : "RunInstances"

1005
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

}
}
}
]
}

If instances are tagged on creation, they must be tagged with a specific tag

In the following policy, users do not have to specify tags in the request, but if they do, the tag must be
purpose=test. No other tags are allowed. Users can apply the tags to any taggable resource in the
RunInstances request.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],
"Resource": "arn:aws:ec2:region:account:*/*",
"Condition": {
"StringEquals": {
"aws:RequestTag/purpose": "test",
"ec2:CreateAction" : "RunInstances"
},
"ForAllValues:StringEquals": {
"aws:TagKeys": "purpose"
}
}
}
]
}

To disallow anyone called tag on create for RunInstances

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowRun",
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:us-east-1::image/*",
"arn:aws:ec2:us-east-1:*:subnet/*",
"arn:aws:ec2:us-east-1:*:network-interface/*",
"arn:aws:ec2:us-east-1:*:security-group/*",
"arn:aws:ec2:us-east-1:*:key-pair/*",
"arn:aws:ec2:us-east-1:*:volume/*",
"arn:aws:ec2:us-east-1:*:instance/*",
"arn:aws:ec2:us-east-1:*:spot-instances-request/*"
]

1006
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

},
{
"Sid": "VisualEditor0",
"Effect": "Deny",
"Action": "ec2:CreateTags",
"Resource": "*"
}
]
}

Only allow specific tags for spot-instances-request. Surprise inconsistency number 2 comes into play
here. Under normal circumstances, specifying no tags will result in Unauthenticated. In the case of spot-
instances-request, this policy will not be evaluated if there are no spot-instances-request tags, so a non-
tag Spot on Run request will succeed.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowRun",
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:us-east-1::image/*",
"arn:aws:ec2:us-east-1:*:subnet/*",
"arn:aws:ec2:us-east-1:*:network-interface/*",
"arn:aws:ec2:us-east-1:*:security-group/*",
"arn:aws:ec2:us-east-1:*:key-pair/*",
"arn:aws:ec2:us-east-1:*:volume/*",
"arn:aws:ec2:us-east-1:*:instance/*",
]
},
{
"Sid": "VisualEditor0",
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": "arn:aws:ec2:us-east-1:*:spot-instances-request/*",
"Condition": {
"StringEquals": {
"aws:RequestTag/environment": "production"
}
}
}
]
}

Tags in a launch template

In the following example, users can launch instances, but only if they use a specific launch template
(lt-09477bcd97b0d310e). The ec2:IsLaunchTemplateResource condition key prevents users from
overriding any of the resources specified in the launch template. The second part of the statement allows
users to tag instances on creation—this part of the statement is necessary if tags are specified for the
instance in the launch template.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",

1007
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Action": "ec2:RunInstances",
"Resource": "*",
"Condition": {
"ArnLike": {
"ec2:LaunchTemplate": "arn:aws:ec2:region:account:launch-template/
lt-09477bcd97b0d310e"
},
"Bool": {
"ec2:IsLaunchTemplateResource": "true"
}
}
},
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],
"Resource": "arn:aws:ec2:region:account:instance/*",
"Condition": {
"StringEquals": {
"ec2:CreateAction" : "RunInstances"
}
}
}
]
}

Elastic GPUs

In the following policy, users can launch an instance and specify an elastic GPU to attach to the instance.
Users can launch instances in any Region, but they can only attach an elastic GPU during a launch in the
us-east-2 Region.

The ec2:ElasticGpuType condition key uses the ForAnyValue modifier to indicate that only the
elastic GPU types eg1.medium and eg1.large are allowed in the request.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:*:account:elastic-gpu/*"
],
"Condition": {
"StringEquals": {
"ec2:Region": "us-east-2"
},
"ForAnyValue:StringLike": {
"ec2:ElasticGpuType": [
"eg1.medium",
"eg1.large"
]
}
}
},
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:*::image/ami-*",

1008
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"arn:aws:ec2:*:account:network-interface/*",
"arn:aws:ec2:*:account:instance/*",
"arn:aws:ec2:*:account:subnet/*",
"arn:aws:ec2:*:account:volume/*",
"arn:aws:ec2:*:account:key-pair/*",
"arn:aws:ec2:*:account:security-group/*"
]
}
]
}

Launch templates

In the following example, users can launch instances, but only if they use a specific launch template
(lt-09477bcd97b0d310e). Users can override any parameters in the launch template by specifying the
parameters in the RunInstances action.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": "*",
"Condition": {
"ArnLike": {
"ec2:LaunchTemplate": "arn:aws:ec2:region:account:launch-template/
lt-09477bcd97b0d310e"
}
}
}
]
}

In this example, users can launch instances only if they use a launch template. The policy uses the
ec2:IsLaunchTemplateResource condition key to prevent users from overriding any pre-existing
ARNs in the launch template.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": "*",
"Condition": {
"ArnLike": {
"ec2:LaunchTemplate": "arn:aws:ec2:region:account:launch-template/*"
},
"Bool": {
"ec2:IsLaunchTemplateResource": "true"
}
}
}
]
}

The following example policy allows user to launch instances, but only if they use a launch template.
Users cannot override the subnet and network interface parameters in the request; these parameters
can only be specified in the launch template. The first part of the statement uses the NotResource
element to allow all other resources except subnets and network interfaces. The second part of the

1009
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

statement allows the subnet and network interface resources, but only if they are sourced from the
launch template.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"NotResource": ["arn:aws:ec2:region:account:subnet/*",
"arn:aws:ec2:region:account:network-interface/*" ],
"Condition": {
"ArnLike": {
"ec2:LaunchTemplate": "arn:aws:ec2:region:account:launch-template/*"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": ["arn:aws:ec2:region:account:subnet/*",
"arn:aws:ec2:region:account:network-interface/*" ],
"Condition": {
"ArnLike": {
"ec2:LaunchTemplate": "arn:aws:ec2:region:account:launch-template/*"
},
"Bool": {
"ec2:IsLaunchTemplateResource": "true"
}
}
}
]
}

The following example allows users to launch instances only if they use a launch template, and only
if the launch template has the tag Purpose=Webservers. Users cannot override any of the launch
template parameters in the RunInstances action.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"NotResource": "arn:aws:ec2:region:account:launch-template/*",
"Condition": {
"ArnLike": {
"ec2:LaunchTemplate": "arn:aws:ec2:region:account:launch-template/*"
},
"Bool": {
"ec2:IsLaunchTemplateResource": "true"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": "arn:aws:ec2:region:account:launch-template/*",
"Condition": {
"StringEquals": {
"ec2:ResourceTag/Purpose": "Webservers"
}
}
}

1010
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

]
}

Work with Spot Instances

You can use the RunInstances action to create Spot Instance requests, and tag the Spot Instance requests
on create. The resource to specify for RunInstances is spot-instances-request.

The spot-instances-request resource is evaluated in the IAM policy as follows:

• If you don't tag a Spot Instance request on create, Amazon EC2 does not evaluate the spot-
instances-request resource in the RunInstances statement.
• If you tag a Spot Instance request on create, Amazon EC2 evaluates the spot-instances-request
resource in the RunInstances statement.

Therefore, for the spot-instances-request resource, the following rules apply to the IAM policy:

• If you use RunInstances to create a Spot Instance request and you don't intend to tag the Spot Instance
request on create, you don’t need to explicitly allow the spot-instances-request resource; the call
will succeed.
• If you use RunInstances to create a Spot Instance request and intend to tag the Spot Instance request
on create, you must include the spot-instances-request resource in the RunInstances allow
statement, otherwise the call will fail.
• If you use RunInstances to create a Spot Instance request and intend to tag the Spot Instance request
on create, you must specify the spot-instances-request resource or * wildcard in the CreateTags
allow statement, otherwise the call will fail.

You can request Spot Instances using RunInstances or RequestSpotInstances. The following example IAM
policies apply only when requesting Spot Instances using RunInstances.

Example: Request Spot Instances using RunInstances

The following policy allows users to request Spot Instances by using the RunInstances action. The spot-
instances-request resource, which is created by RunInstances, requests Spot Instances.
Note
To use RunInstances to create Spot Instance requests, you can omit spot-instances-
request from the Resource list if you do not intend to tag the Spot Instance requests on
create. This is because Amazon EC2 does not evaluate the spot-instances-request resource
in the RunInstances statement if the Spot Instance request is not tagged on create.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowRun",
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:us-east-1::image/*",
"arn:aws:ec2:us-east-1:*:subnet/*",
"arn:aws:ec2:us-east-1:*:network-interface/*",
"arn:aws:ec2:us-east-1:*:security-group/*",
"arn:aws:ec2:us-east-1:*:key-pair/*",
"arn:aws:ec2:us-east-1:*:volume/*",
"arn:aws:ec2:us-east-1:*:instance/*",
"arn:aws:ec2:us-east-1:*:spot-instances-request/*"

1011
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

]
}
]
}

Warning
NOT SUPPORTED – Example: Deny users permission to request Spot Instances using
RunInstances
The following policy is not supported for the spot-instances-request resource.
The following policy is meant to give users the permission to launch On-Demand Instances, but
deny users the permission to request Spot Instances. The spot-instances-request resource,
which is created by RunInstances, is the resource that requests Spot Instances. The second
statement is meant to deny the RunInstances action for the spot-instances-request
resource. However, this condition is not supported because Amazon EC2 does not evaluate
the spot-instances-request resource in the RunInstances statement if the Spot Instance
request is not tagged on create.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowRun",
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:us-east-1::image/*",
"arn:aws:ec2:us-east-1:*:subnet/*",
"arn:aws:ec2:us-east-1:*:network-interface/*",
"arn:aws:ec2:us-east-1:*:security-group/*",
"arn:aws:ec2:us-east-1:*:key-pair/*",
"arn:aws:ec2:us-east-1:*:volume/*",
"arn:aws:ec2:us-east-1:*:instance/*"
]
},
{
"Sid": "DenySpotInstancesRequests - NOT SUPPORTED - DO NOT USE!",
"Effect": "Deny",
"Action": "ec2:RunInstances",
"Resource": "arn:aws:ec2:us-east-1:*:spot-instances-request/*"
}
]
}

Example: Tag Spot Instance requests on create

The following policy allows users to tag all resources that are created during instance launch. The first
statement allows RunInstances to create the listed resources. The spot-instances-request resource,
which is created by RunInstances, is the resource that requests Spot Instances. The second statement
provides a * wildcard to allow all resources to be tagged when they are created at instance launch.
Note
If you tag a Spot Instance request on create, Amazon EC2 evaluates the spot-instances-
request resource in the RunInstances statement. Therefore, you must explicitly allow the
spot-instances-request resource for the RunInstances action, otherwise the call will fail.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowRun",

1012
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:us-east-1::image/*",
"arn:aws:ec2:us-east-1:*:subnet/*",
"arn:aws:ec2:us-east-1:*:network-interface/*",
"arn:aws:ec2:us-east-1:*:security-group/*",
"arn:aws:ec2:us-east-1:*:key-pair/*",
"arn:aws:ec2:us-east-1:*:volume/*",
"arn:aws:ec2:us-east-1:*:instance/*",
"arn:aws:ec2:us-east-1:*:spot-instances-request/*"
]
},
{
"Sid": "TagResources",
"Effect": "Allow",
"Action": "ec2:CreateTags",
"Resource": "*"
}
]
}

Example: Deny tag on create for Spot Instance requests

The following policy denies users the permission to tag the resources that are created during instance
launch.

The first statement allows RunInstances to create the listed resources. The spot-instances-request
resource, which is created by RunInstances, is the resource that requests Spot Instances. The second
statement provides a * wildcard to deny all resources being tagged when they are created at instance
launch. If spot-instances-request or any other resource is tagged on create, the RunInstances call
will fail.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowRun",
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:us-east-1::image/*",
"arn:aws:ec2:us-east-1:*:subnet/*",
"arn:aws:ec2:us-east-1:*:network-interface/*",
"arn:aws:ec2:us-east-1:*:security-group/*",
"arn:aws:ec2:us-east-1:*:key-pair/*",
"arn:aws:ec2:us-east-1:*:volume/*",
"arn:aws:ec2:us-east-1:*:instance/*",
"arn:aws:ec2:us-east-1:*:spot-instances-request/*"
]
},
{
"Sid": "DenyTagResources",
"Effect": "Deny",
"Action": "ec2:CreateTags",
"Resource": "*"
}
]
}

1013
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

Warning
NOT SUPPORTED – Example: Allow creating a Spot Instance request only if it is assigned a
specific tag
The following policy is not supported for the spot-instances-request resource.
The following policy is meant to grant RunInstances the permission to create a Spot Instance
request only if the request is tagged with a specific tag.
The first statement allows RunInstances to create the listed resources.
The second statement is meant to grant users the permission to create a Spot Instance request
only if the request has the tag environment=production. If this condition is applied to other
resources created by RunInstances, specifying no tags results in an Unauthenticated error.
However, if no tags are specified for the Spot Instance request, Amazon EC2 does not evaluate
the spot-instances-request resource in the RunInstances statement, which results in non-
tagged Spot Instance requests being created by RunInstances.
Note that specifying another tag other than environment=production results in an
Unauthenticated error, because if a user tags a Spot Instance request, Amazon EC2 evaluates
the spot-instances-request resource in the RunInstances statement.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowRun",
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:us-east-1::image/*",
"arn:aws:ec2:us-east-1:*:subnet/*",
"arn:aws:ec2:us-east-1:*:network-interface/*",
"arn:aws:ec2:us-east-1:*:security-group/*",
"arn:aws:ec2:us-east-1:*:key-pair/*",
"arn:aws:ec2:us-east-1:*:volume/*",
"arn:aws:ec2:us-east-1:*:instance/*"
]
},
{
"Sid": "RequestSpotInstancesOnlyIfTagIs_environment=production - NOT
SUPPORTED - DO NOT USE!",
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": "arn:aws:ec2:us-east-1:*:spot-instances-request/*",
"Condition": {
"StringEquals": {
"aws:RequestTag/environment": "production"
}
}
},
{
"Sid": "TagResources",
"Effect": "Allow",
"Action": "ec2:CreateTags",
"Resource": "*"
}

]
}

Example: Deny creating a Spot Instance request if it is assigned a specific tag

The following policy denies RunInstances the permission to create a Spot Instance request if the request
is tagged with environment=production.

1014
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

The first statement allows RunInstances to create the listed resources.

The second statement denies users the permission to create a Spot Instance request if the request has
the tag environment=production. Specifying environment=production as a tag results in an
Unauthenticated error. Specifying other tags or specifying no tags will result in the creation of a Spot
Instance request.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowRun",
"Effect": "Allow",
"Action": [
"ec2:RunInstances"
],
"Resource": [
"arn:aws:ec2:us-east-1::image/*",
"arn:aws:ec2:us-east-1:*:subnet/*",
"arn:aws:ec2:us-east-1:*:network-interface/*",
"arn:aws:ec2:us-east-1:*:security-group/*",
"arn:aws:ec2:us-east-1:*:key-pair/*",
"arn:aws:ec2:us-east-1:*:volume/*",
"arn:aws:ec2:us-east-1:*:instance/*",
"arn:aws:ec2:us-east-1:*:spot-instances-request/*"
]
},
{
"Sid": "DenySpotInstancesRequests",
"Effect": "Deny",
"Action": "ec2:RunInstances",
"Resource": "arn:aws:ec2:us-east-1:*:spot-instances-request/*",
"Condition": {
"StringEquals": {
"aws:RequestTag/environment": "production"
}
}
},
{
"Sid": "TagResources",
"Effect": "Allow",
"Action": "ec2:CreateTags",
"Resource": "*"
}
]
}

Example: Work with Reserved Instances

The following policy gives users permission to view, modify, and purchase Reserved Instances in your
account.

It is not possible to set resource-level permissions for individual Reserved Instances. This policy means
that users have access to all the Reserved Instances in the account.

The Resource element uses a * wildcard to indicate that users can specify all resources with the action;
in this case, they can list and modify all Reserved Instances in the account. They can also purchase
Reserved Instances using the account credentials. The * wildcard is also necessary in cases where the API
action does not support resource-level permissions.

{
"Version": "2012-10-17",

1015
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:DescribeReservedInstances",
"ec2:ModifyReservedInstances",
"ec2:PurchaseReservedInstancesOffering",
"ec2:DescribeAvailabilityZones",
"ec2:DescribeReservedInstancesOfferings"
],
"Resource": "*"
}
]
}

To allow users to view and modify the Reserved Instances in your account, but not purchase new
Reserved Instances.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:DescribeReservedInstances",
"ec2:ModifyReservedInstances",
"ec2:DescribeAvailabilityZones"
],
"Resource": "*"
}
]
}

Example: Tag resources

The following policy allows users to use the CreateTags action to apply tags to an instance only if
the tag contains the key environment and the value production. The ForAllValues modifier is
used with the aws:TagKeys condition key to indicate that only the key environment is allowed in the
request (no other tags are allowed). The user cannot tag any other resource types.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],
"Resource": "arn:aws:ec2:region:account:instance/*",
"Condition": {
"StringEquals": {
"aws:RequestTag/environment": "production"
},
"ForAllValues:StringEquals": {
"aws:TagKeys": [
"environment"
]
}
}
}
]

1016
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

The following policy allows users to tag any taggable resource that already has a tag with a key
of owner and a value of the IAM username. In addition, users must specify a tag with a key of
anycompany:environment-type and a value of either test or prod in the request. Users can specify
additional tags in the request.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],
"Resource": "arn:aws:ec2:region:account:*/*",
"Condition": {
"StringEquals": {
"aws:RequestTag/anycompany:environment-type": ["test","prod"],
"ec2:ResourceTag/owner": "${aws:username}"
}
}
}
]
}

You can create an IAM policy that allows users to delete specific tags for a resource. For example, the
following policy allows users to delete tags for a volume if the tag keys specified in the request are
environment or cost-center. Any value can be specified for the tag but the tag key must match
either of the specified keys.
Note
If you delete a resource, all tags associated with the resource are also deleted. Users do not need
permissions to use the ec2:DeleteTags action to delete a resource that has tags; they only
need permissions to perform the deleting action.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:DeleteTags",
"Resource": "arn:aws:ec2:us-east-1:123456789012:volume/*",
"Condition": {
"ForAllValues:StringEquals": {
"aws:TagKeys": ["environment","cost-center"]
}
}
}
]
}

This policy allows users to delete only the environment=prod tag on any resource, and only if the
resource is already tagged with a key of owner and a value of the IAM username. Users cannot delete
any other tags for a resource.

{
"Version": "2012-10-17",
"Statement": [
{

1017
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Effect": "Allow",
"Action": [
"ec2:DeleteTags"
],
"Resource": "arn:aws:ec2:region:account:*/*",
"Condition": {
"StringEquals": {
"aws:RequestTag/environment": "prod",
"ec2:ResourceTag/owner": "${aws:username}"
},
"ForAllValues:StringEquals": {
"aws:TagKeys": ["environment"]
}
}
}
]
}

Example: Work with IAM roles

The following policy allows users to attach, replace, and detach an IAM role to instances that have the
tag department=test. Replacing or detaching an IAM role requires an association ID, therefore the
policy also grants users permission to use the ec2:DescribeIamInstanceProfileAssociations
action.

IAM users must have permission to use the iam:PassRole action in order to pass the role to the
instance.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:AssociateIamInstanceProfile",
"ec2:ReplaceIamInstanceProfileAssociation",
"ec2:DisassociateIamInstanceProfile"
],
"Resource": "arn:aws:ec2:region:account:instance/*",
"Condition": {
"StringEquals": {
"ec2:ResourceTag/department":"test"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:DescribeIamInstanceProfileAssociations",
"Resource": "*"
},
{
"Effect": "Allow",
"Action": "iam:PassRole",
"Resource": "*"
}
]
}

The following policy allows users to attach or replace an IAM role for any instance. Users can only attach
or replace IAM roles with names that begin with TestRole-. For the iam:PassRole action, ensure that
you specify the name of the IAM role and not the instance profile (if the names are different). For more
information, see Instance profiles (p. 1032).

1018
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:AssociateIamInstanceProfile",
"ec2:ReplaceIamInstanceProfileAssociation"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": "ec2:DescribeIamInstanceProfileAssociations",
"Resource": "*"
},
{
"Effect": "Allow",
"Action": "iam:PassRole",
"Resource": "arn:aws:iam::account:role/TestRole-*"
}
]
}

Example: Work with route tables

The following policy allows users to add, remove, and replace routes for route tables that are associated
with VPC vpc-ec43eb89 only. To specify a VPC for the ec2:Vpc condition key, you must specify the full
ARN of the VPC.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:DeleteRoute",
"ec2:CreateRoute",
"ec2:ReplaceRoute"
],
"Resource": [
"arn:aws:ec2:region:account:route-table/*"
],
"Condition": {
"StringEquals": {
"ec2:Vpc": "arn:aws:ec2:region:account:vpc/vpc-ec43eb89"
}
}
}
]
}

Example: Allow a specific instance to view resources in other AWS services

The following is an example of a policy that you might attach to an IAM role. The policy allows an
instance to view resources in various AWS services. It uses the ec2:SourceInstanceARN condition key
to specify that the instance from which the request is made must be instance i-093452212644b0dd6.
If the same IAM role is associated with another instance, the other instance cannot perform any of these
actions.

The ec2:SourceInstanceARN key is an AWS-wide condition key, therefore it can be used for other
service actions, not just Amazon EC2.

1019
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:DescribeVolumes",
"s3:ListAllMyBuckets",
"dynamodb:ListTables",
"rds:DescribeDBInstances"
],
"Resource": [
"*"
],
"Condition": {
"ArnEquals": {
"ec2:SourceInstanceARN": "arn:aws:ec2:region:account:instance/
i-093452212644b0dd6"
}
}
}
]
}

Example: Work with launch templates

The following policy allows users to create a launch template version and modify a launch template, but
only for a specific launch template (lt-09477bcd97b0d3abc). Users cannot work with other launch
templates.

{
"Version": "2012-10-17",
"Statement": [
{
"Action": [
"ec2:CreateLaunchTemplateVersion",
"ec2:ModifyLaunchTemplate"
],
"Effect": "Allow",
"Resource": "arn:aws:ec2:region:account:launch-template/lt-09477bcd97b0d3abc"
}
]
}

The following policy allows users to delete any launch template and launch template version, provided
that the launch template has the tag Purpose=Testing.

{
"Version": "2012-10-17",
"Statement": [
{
"Action": [
"ec2:DeleteLaunchTemplate",
"ec2:DeleteLaunchTemplateVersions"
],
"Effect": "Allow",
"Resource": "arn:aws:ec2:region:account:launch-template/*",
"Condition": {
"StringEquals": {
"ec2:ResourceTag/Purpose": "Testing"
}
}

1020
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

}
]
}

Work with instance metadata

The following policies ensure that users can only retrieve instance metadata (p. 704) using Instance
Metadata Service Version 2 (IMDSv2). You can combine the following four policies into one policy with
four statements. When combined as one policy, you can use the policy as a service control policy (SCP). It
can work equally well as a deny policy that you apply to an existing IAM policy (taking away and limiting
existing permission), or as an SCP that is applied globally across an account, an organizational unit (OU),
or an entire organization.
Note
The following RunInstances metadata options policies must be used in conjunction with a policy
that gives the principal permissions to launch an instance with RunInstances. If the principal
does not also have RunInstances permissions, it will not be able to launch an instance. For
more information, see the policies in Work with instances (p. 988) and Launch instances
(RunInstances) (p. 999).
Important
If you use Auto Scaling groups and you need to require the use of IMDSv2 on all new instances,
your Auto Scaling groups must use launch templates.
When an Auto Scaling group uses a launch template, the ec2:RunInstances permissions of
the IAM principal are checked when a new Auto Scaling group is created. They are also checked
when an existing Auto Scaling group is updated to use a new launch template or a new version
of a launch template.
Restrictions on the use of IMDSv1 on IAM principals for RunInstances are only checked
when an Auto Scaling group that is using a launch template, is created or updated. For an
Auto Scaling group that is configured to use the Latest or Default launch template, the
permissions are not checked when a new version of the launch template is created. For
permissions to be checked, you must configure the Auto Scaling group to use a specific version
of the launch template.

To enforce the use of IMDSv2 on instances launched by Auto Scaling groups, the
following additional steps are required:

1. Disable the use of launch configurations for all accounts in your organization by using
either service control policies (SCPs) or IAM permissions boundaries for new principals
that are created. For existing IAM principals with Auto Scaling group permissions,
update their associated policies with this condition key. To disable the use of launch
configurations, create or modify the relevant SCP, permissions boundary, or IAM policy with
the "autoscaling:LaunchConfigurationName" condition key with the value specified as
null.
2. For new launch templates, configure the instance metadata options in the launch template.
For existing launch templates, create a new version of the launch template and configure the
instance metadata options in the new version.
3. In the policy that gives any principal the permission to use a launch
template, restrict association of $latest and $default by specifying
"autoscaling:LaunchTemplateVersionSpecified": "true". By restricting the
use to a specific version of a launch template, you can ensure that new instances will be
launched using the version in which the instance metadata options are configured. For more
information, see LaunchTemplateSpecification in the Amazon EC2 Auto Scaling API Reference,
specifically the Version parameter.
4. For an Auto Scaling group that uses a launch configuration, replace the launch configuration
with a launch template. For more information, see Replacing a Launch Configuration with a
Launch Template in the Amazon EC2 Auto Scaling User Guide.

1021
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

5. For an Auto Scaling group that uses a launch template, make sure that it uses a new launch
template with the instance metadata options configured, or uses a new version of the current
launch template with the instance metadata options configured. For more information, see
update-auto-scaling-group in the AWS CLI Command Reference.

Examples
• Require the use of IMDSv2 (p. 1022)
• Specify maximum hop limit (p. 1022)
• Limit who can modify the instance metadata options (p. 1023)
• Require role credentials to be retrieved from IMDSv2 (p. 1023)

Require the use of IMDSv2

The following policy specifies that you can’t call the RunInstances API unless the instance is also opted in
to require the use of IMDSv2 (indicated by "ec2:MetadataHttpTokens": "required"). If you do not
specify that the instance requires IMDSv2, you get an UnauthorizedOperation error when you call the
RunInstances API.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "RequireImdsV2",
"Effect": "Deny",
"Action": "ec2:RunInstances",
"Resource": "arn:aws:ec2:*:*:instance/*",
"Condition": {
"StringNotEquals": {
"ec2:MetadataHttpTokens": "required"
}
}
}
]
}

Specify maximum hop limit

The following policy specifies that you can’t call the RunInstances API unless you also specify a hop limit,
and the hop limit can’t be more than 3. If you fail to do that, you get an UnauthorizedOperation
error when you call the RunInstances API.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "MaxImdsHopLimit",
"Effect": "Deny",
"Action": "ec2:RunInstances",
"Resource": "arn:aws:ec2:*:*:instance/*",
"Condition": {
"NumericGreaterThan": {
"ec2:MetadataHttpPutResponseHopLimit": "3"
}
}
}
]
}

1022
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

Limit who can modify the instance metadata options

The following policy removes the ability for the general population of administrators to modify instance
metadata options, and permits only users with the role ec2-imds-admins to make changes. If any
principal other than the ec2-imds-admins role tries to call the ModifyInstanceMetadataOptions API,
it will get an UnauthorizedOperation error. This statement could be used to control the use of the
ModifyInstanceMetadataOptions API; there are currently no fine-grained access controls (conditions) for
the ModifyInstanceMetadataOptions API.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "AllowOnlyImdsAdminsToModifySettings",
"Effect": "Deny",
"Action": "ec2:ModifyInstanceMetadataOptions",
"Resource": "*",
"Condition": {
"StringNotLike": {
"aws:PrincipalARN": "arn:aws:iam::*:role/ec2-imds-admins"
}
}
}
]
}

Require role credentials to be retrieved from IMDSv2

The following policy specifies that if this policy is applied to a role, and the role is assumed by
the EC2 service and the resulting credentials are used to sign a request, then the request must
be signed by EC2 role credentials retrieved from IMDSv2. Otherwise, all of its API calls will get an
UnauthorizedOperation error. This statement/policy can be applied generally because, if the request
is not signed by EC2 role credentials, it has no effect.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "RequireAllEc2RolesToUseV2",
"Effect": "Deny",
"Action": "*",
"Resource": "*",
"Condition": {
"NumericLessThan": {
"ec2:RoleDelivery": "2.0"
}
}
}
]
}

Example policies for working in the Amazon EC2 console

You can use IAM policies to grant users permissions to view and work with specific resources in the
Amazon EC2 console. You can use the example policies in the previous section; however, they are
designed for requests that are made with the AWS CLI or an AWS SDK. The console uses additional
API actions for its features, so these policies may not work as expected. For example, a user that has
permission to use only the DescribeVolumes API action will encounter errors when trying to view
volumes in the console. This section demonstrates policies that enable users to work with specific parts
of the console.

1023
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

Tip
To help you work out which API actions are required to perform tasks in the console, you can
use a service such as AWS CloudTrail. For more information, see the AWS CloudTrail User Guide.
If your policy does not grant permission to create or modify a specific resource, the console
displays an encoded message with diagnostic information. You can decode the message using
the DecodeAuthorizationMessage API action for AWS STS, or the decode-authorization-message
command in the AWS CLI.

Examples
• Example: Read-only access (p. 1024)
• Example: Use the EC2 launch wizard (p. 1025)
• Example: Work with volumes (p. 1028)
• Example: Work with security groups (p. 1029)
• Example: Work with Elastic IP addresses (p. 1030)
• Example: Work with Reserved Instances (p. 1031)

For additional information about creating policies for the Amazon EC2 console, see the following AWS
Security Blog post: Granting Users Permission to Work in the Amazon EC2 Console.

Example: Read-only access

To allow users to view all resources in the Amazon EC2 console, you can use the same policy as the
following example: Example: Read-only access (p. 987). Users cannot perform any actions on those
resources or create new resources, unless another statement grants them permission to do so.

View instances, AMIs, and snapshots

Alternatively, you can provide read-only access to a subset of resources. To do this, replace the *
wildcard in the ec2:Describe API action with specific ec2:Describe actions for each resource. The
following policy allows users to view all instances, AMIs, and snapshots in the Amazon EC2 console.
The ec2:DescribeTags action allows users to view public AMIs. The console requires the tagging
information to display public AMIs; however, you can remove this action to allow users to view only
private AMIs.

{
"Version": "2012-10-17",
"Statement": [{
"Effect": "Allow",
"Action": [
"ec2:DescribeInstances",
"ec2:DescribeImages",
"ec2:DescribeTags",
"ec2:DescribeSnapshots"
],
"Resource": "*"
}
]
}

Note
The Amazon EC2 ec2:Describe* API actions do not support resource-level permissions, so
you cannot control which individual resources users can view in the console. Therefore, the *
wildcard is necessary in the Resource element of the above statement. For more information
about which ARNs you can use with which Amazon EC2 API actions, see Actions, resources, and
condition keys for Amazon EC2.

View instances and CloudWatch metrics

1024
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

The following policy allows users to view instances in the Amazon EC2 console, as well as CloudWatch
alarms and metrics in the Monitoring tab of the Instances page. The Amazon EC2 console uses the
CloudWatch API to display the alarms and metrics, so you must grant users permission to use the
cloudwatch:DescribeAlarms and cloudwatch:GetMetricStatistics actions.

{
"Version": "2012-10-17",
"Statement": [{
"Effect": "Allow",
"Action": [
"ec2:DescribeInstances",
"cloudwatch:DescribeAlarms",
"cloudwatch:GetMetricStatistics"
],
"Resource": "*"
}
]
}

Example: Use the EC2 launch wizard

The Amazon EC2 launch wizard is a series of screens with options to configure and launch an instance.
Your policy must include permission to use the API actions that allow users to work with the wizard's
options. If your policy does not include permission to use those actions, some items in the wizard cannot
load properly, and users cannot complete a launch.

Basic launch wizard access

To complete a launch successfully, users must be given permission to use the ec2:RunInstances API
action, and at least the following API actions:

• ec2:DescribeImages: To view and select an AMI.

• ec2:DescribeInstanceTypes: To view and select an instance type.
• ec2:DescribeVpcs: To view the available network options.
• ec2:DescribeSubnets: To view all available subnets for the chosen VPC.
• ec2:DescribeSecurityGroups or ec2:CreateSecurityGroup: To view and select an existing
security group, or to create a new one.
• ec2:DescribeKeyPairs or ec2:CreateKeyPair: To select an existing key pair, or to create a new
one.
• ec2:AuthorizeSecurityGroupIngress: To add inbound rules.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:DescribeInstances",
"ec2:DescribeImages",
"ec2:DescribeInstanceTypes",
"ec2:DescribeKeyPairs",
"ec2:DescribeVpcs",
"ec2:DescribeSubnets",
"ec2:DescribeSecurityGroups",
"ec2:CreateSecurityGroup",
"ec2:AuthorizeSecurityGroupIngress",
"ec2:CreateKeyPair"
],
"Resource": "*"

1025
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

},
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": "*"
}
]
}

You can add API actions to your policy to provide more options for users, for example:

• ec2:DescribeAvailabilityZones: To view and select a specific Availability Zone.

• ec2:DescribeNetworkInterfaces: To view and select existing network interfaces for the selected
subnet.
• To add outbound rules to VPC security groups, users must be granted permission to use the
ec2:AuthorizeSecurityGroupEgress API action. To modify or delete existing rules, users must be
granted permission to use the relevant ec2:RevokeSecurityGroup* API action.
• ec2:CreateTags: To tag the resources that are created by RunInstances. For more information, see
Grant permission to tag resources during creation (p. 984). If users do not have permission to use this
action and they attempt to apply tags on the tagging page of the launch wizard, the launch fails.
Important
Be careful about granting users permission to use the ec2:CreateTags action, because
doing so limits your ability to use the ec2:ResourceTag condition key to restrict their use
of other resources. If you grant users permission to use the ec2:CreateTags action, they
can change a resource's tag in order to bypass those restrictions. For more information, see
Control access to EC2 resources using resource tags (p. 986).
• To use Systems Manager parameters when selecting an AMI, you must add
ssm:DescribeParameters and ssm:GetParameters to your policy. ssm:DescribeParameters
grants your IAM users the permission to view and select Systems Manager parameters.
ssm:GetParameters grants your IAM users the permission to get the values of the Systems
Manager parameters. You can also restrict access to specific Systems Manager parameters. For more
information, see Restrict access to specific Systems Manager parameters later in this section.

Currently, the Amazon EC2 Describe* API actions do not support resource-level permissions, so you
cannot restrict which individual resources users can view in the launch wizard. However, you can apply
resource-level permissions on the ec2:RunInstances API action to restrict which resources users can
use to launch an instance. The launch fails if users select options that they are not authorized to use.

Restrict access to a specific instance type, subnet, and Region

The following policy allows users to launch t2.micro instances using AMIs owned by Amazon, and only
into a specific subnet (subnet-1a2b3c4d). Users can only launch in the sa-east-1 Region. If users select
a different Region, or select a different instance type, AMI, or subnet in the launch wizard, the launch
fails.

The first statement grants users permission to view the options in the launch wizard or to create new
ones, as explained in the example above. The second statement grants users permission to use the
network interface, volume, key pair, security group, and subnet resources for the ec2:RunInstances
action, which are required to launch an instance into a VPC. For more information about using the
ec2:RunInstances action, see Launch instances (RunInstances) (p. 999). The third and fourth
statements grant users permission to use the instance and AMI resources respectively, but only if the
instance is a t2.micro instance, and only if the AMI is owned by Amazon.

{
"Version": "2012-10-17",
"Statement": [{
"Effect": "Allow",

1026
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Action": [
"ec2:DescribeInstances",
"ec2:DescribeImages",
"ec2:DescribeInstanceTypes",
"ec2:DescribeKeyPairs",
"ec2:CreateKeyPair",
"ec2:DescribeVpcs",
"ec2:DescribeSubnets",
"ec2:DescribeSecurityGroups",
"ec2:CreateSecurityGroup",
"ec2:AuthorizeSecurityGroupIngress"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action":"ec2:RunInstances",
"Resource": [
"arn:aws:ec2:sa-east-1:111122223333:network-interface/*",
"arn:aws:ec2:sa-east-1:111122223333:volume/*",
"arn:aws:ec2:sa-east-1:111122223333:key-pair/*",
"arn:aws:ec2:sa-east-1:111122223333:security-group/*",
"arn:aws:ec2:sa-east-1:111122223333:subnet/subnet-1a2b3c4d"
]
},
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:sa-east-1:111122223333:instance/*"
],
"Condition": {
"StringEquals": {
"ec2:InstanceType": "t2.micro"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:RunInstances",
"Resource": [
"arn:aws:ec2:sa-east-1::image/ami-*"
],
"Condition": {
"StringEquals": {
"ec2:Owner": "amazon"
}
}
}
]
}

Restrict access to specific Systems Manager parameters

The following policy grants access to use Systems Manager parameters with a specific name.

The first statement grants users the permission to view Systems Manager parameters when selecting an
AMI in the launch wizard. The second statement grants users the permission to only use parameters that
are named prod-*.

{
"Version": "2012-10-17",
"Statement": [{
"Effect": "Allow",

1027
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

"Action": [
"ssm:DescribeParameters"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": [
"ssm:GetParameters"
],
"Resource": "arn:aws:ssm:us-east-2:123456123:parameter/prod-*"
}
]
}

Example: Work with volumes

The following policy grants users permission to view and create volumes, and attach and detach volumes
to specific instances.

Users can attach any volume to instances that have the tag "purpose=test", and also detach volumes
from those instances. To attach a volume using the Amazon EC2 console, it is helpful for users to have
permission to use the ec2:DescribeInstances action, as this allows them to select an instance from a
pre-populated list in the Attach Volume dialog box. However, this also allows users to view all instances
on the Instances page in the console, so you can omit this action.

In the first statement, the ec2:DescribeAvailabilityZones action is necessary to ensure that a user
can select an Availability Zone when creating a volume.

Users cannot tag the volumes that they create (either during or after volume creation).

{
"Version": "2012-10-17",
"Statement": [{
"Effect": "Allow",
"Action": [
"ec2:DescribeVolumes",
"ec2:DescribeAvailabilityZones",
"ec2:CreateVolume",
"ec2:DescribeInstances"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": [
"ec2:AttachVolume",
"ec2:DetachVolume"
],
"Resource": "arn:aws:ec2:region:111122223333:instance/*",
"Condition": {
"StringEquals": {
"ec2:ResourceTag/purpose": "test"
}
}
},
{
"Effect": "Allow",
"Action": [
"ec2:AttachVolume",
"ec2:DetachVolume"
],
"Resource": "arn:aws:ec2:region:111122223333:volume/*"
}

1028
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

]
}

Example: Work with security groups

View security groups and add and remove rules

The following policy grants users permission to view security groups in the Amazon EC2 console, to add
and remove inbound and outbound rules, and to modify rule descriptions for existing security groups
that have the tag Department=Test.

In the first statement, the ec2:DescribeTags action allows users to view tags in the console, which
makes it easier for users to identify the security groups that they are allowed to modify.

{
"Version": "2012-10-17",
"Statement": [{
"Effect": "Allow",
"Action": [
"ec2:DescribeSecurityGroups",
"ec2:DescribeTags"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": [
"ec2:AuthorizeSecurityGroupIngress",
"ec2:RevokeSecurityGroupIngress",
"ec2:AuthorizeSecurityGroupEgress",
"ec2:RevokeSecurityGroupEgress",
"ec2:UpdateSecurityGroupRuleDescriptionsIngress",
"ec2:UpdateSecurityGroupRuleDescriptionsEgress"
],
"Resource": [
"arn:aws:ec2:region:111122223333:security-group/*"
],
"Condition": {
"StringEquals": {
"ec2:ResourceTag/Department": "Test"
}
}
}
]
}

Work with the Create Security Group dialog box

You can create a policy that allows users to work with the Create Security Group dialog box in the
Amazon EC2 console. To use this dialog box, users must be granted permission to use at the least the
following API actions:

• ec2:CreateSecurityGroup: To create a new security group.

• ec2:DescribeVpcs: To view a list of existing VPCs in the VPC list.

With these permissions, users can create a new security group successfully, but they cannot add any rules
to it. To work with rules in the Create Security Group dialog box, you can add the following API actions
to your policy:

• ec2:AuthorizeSecurityGroupIngress: To add inbound rules.

• ec2:AuthorizeSecurityGroupEgress: To add outbound rules to VPC security groups.

1029
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

• ec2:RevokeSecurityGroupIngress: To modify or delete existing inbound rules. This is useful to

allow users to use the Copy to new feature in the console. This feature opens the Create Security
Group dialog box and populates it with the same rules as the security group that was selected.
• ec2:RevokeSecurityGroupEgress: To modify or delete outbound rules for VPC security groups.
This is useful to allow users to modify or delete the default outbound rule that allows all outbound
traffic.
• ec2:DeleteSecurityGroup: To cater for when invalid rules cannot be saved. The console first
creates the security group, and then adds the specified rules. If the rules are invalid, the action fails,
and the console attempts to delete the security group. The user remains in the Create Security Group
dialog box so that they can correct the invalid rule and try to create the security group again. This API
action is not required, but if a user is not granted permission to use it and attempts to create a security
group with invalid rules, the security group is created without any rules, and the user must add them
afterward.
• ec2:UpdateSecurityGroupRuleDescriptionsIngress: To add or update descriptions of ingress
(inbound) security group rules.
• ec2:UpdateSecurityGroupRuleDescriptionsEgress: To add or update descriptions of egress
(outbound) security group rules.

The following policy grants users permission to use the Create Security Group dialog box, and to create
inbound and outbound rules for security groups that are associated with a specific VPC (vpc-1a2b3c4d).
Users can create security groups for EC2-Classic or another VPC, but they cannot add any rules to them.
Similarly, users cannot add any rules to any existing security group that's not associated with VPC
vpc-1a2b3c4d. Users are also granted permission to view all security groups in the console. This makes
it easier for users to identify the security groups to which they can add inbound rules. This policy also
grants users permission to delete security groups that are associated with VPC vpc-1a2b3c4d.

{
"Version": "2012-10-17",
"Statement": [{
"Effect": "Allow",
"Action": [
"ec2:DescribeSecurityGroups",
"ec2:CreateSecurityGroup",
"ec2:DescribeVpcs"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": [
"ec2:DeleteSecurityGroup",
"ec2:AuthorizeSecurityGroupIngress",
"ec2:AuthorizeSecurityGroupEgress"
],
"Resource": "arn:aws:ec2:region:111122223333:security-group/*",
"Condition":{
"ArnEquals": {
"ec2:Vpc": "arn:aws:ec2:region:111122223333:vpc/vpc-1a2b3c4d"
}
}
}
]
}

Example: Work with Elastic IP addresses

To allow users to view Elastic IP addresses in the Amazon EC2 console, you must grant users permission
to use the ec2:DescribeAddresses action.

1030
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM policies

To allow users to work with Elastic IP addresses, you can add the following actions to your policy.

• ec2:AllocateAddress: To allocate an Elastic IP address.

• ec2:ReleaseAddress: To release an Elastic IP address.
• ec2:AssociateAddress: To associate an Elastic IP address with an instance or a network interface.
• ec2:DescribeNetworkInterfaces and ec2:DescribeInstances: To work with the Associate
address screen. The screen displays the available instances or network interfaces to which you can
associate an Elastic IP address.
• ec2:DisassociateAddress: To disassociate an Elastic IP address from an instance or a network
interface.

The following policy allows users to view, allocate, and associate Elastic IP addresses with instances.
Users cannot associate Elastic IP addresses with network interfaces, disassociate Elastic IP addresses, or
release them.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:DescribeAddresses",
"ec2:AllocateAddress",
"ec2:DescribeInstances",
"ec2:AssociateAddress"
],
"Resource": "*"
}
]
}

Example: Work with Reserved Instances

The following policy can be attached to an IAM user. It gives the user access to view and modify Reserved
Instances in your account, as well as purchase new Reserved Instances in the AWS Management Console.

This policy allows users to view all the Reserved Instances, as well as On-Demand Instances, in the
account. It's not possible to set resource-level permissions for individual Reserved Instances.

{
"Version": "2012-10-17",
"Statement": [{
"Effect": "Allow",
"Action": [
"ec2:DescribeReservedInstances",
"ec2:ModifyReservedInstances",
"ec2:PurchaseReservedInstancesOffering",
"ec2:DescribeInstances",
"ec2:DescribeInstanceTypes",
"ec2:DescribeAvailabilityZones",
"ec2:DescribeReservedInstancesOfferings"
],
"Resource": "*"
}
]
}

The ec2:DescribeAvailabilityZones action is necessary to ensure that the Amazon EC2 console
can display information about the Availability Zones in which you can purchase Reserved Instances. The

1031
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM roles

ec2:DescribeInstances action is not required, but ensures that the user can view the instances in the
account and purchase reservations to match the correct specifications.

You can adjust the API actions to limit user access, for example removing ec2:DescribeInstances
and ec2:DescribeAvailabilityZones means the user has read-only access.

IAM roles for Amazon EC2

Applications must sign their API requests with AWS credentials. Therefore, if you are an application
developer, you need a strategy for managing credentials for your applications that run on EC2 instances.
For example, you can securely distribute your AWS credentials to the instances, enabling the applications
on those instances to use your credentials to sign requests, while protecting your credentials from other
users. However, it's challenging to securely distribute credentials to each instance, especially those that
AWS creates on your behalf, such as Spot Instances or instances in Auto Scaling groups. You must also be
able to update the credentials on each instance when you rotate your AWS credentials.

We designed IAM roles so that your applications can securely make API requests from your instances,
without requiring you to manage the security credentials that the applications use. Instead of creating
and distributing your AWS credentials, you can delegate permission to make API requests using IAM roles
as follows:

1. Create an IAM role.

2. Define which accounts or AWS services can assume the role.
3. Define which API actions and resources the application can use after assuming the role.
4. Specify the role when you launch your instance, or attach the role to an existing instance.
5. Have the application retrieve a set of temporary credentials and use them.

For example, you can use IAM roles to grant permissions to applications running on your instances that
need to use a bucket in Amazon S3. You can specify permissions for IAM roles by creating a policy in
JSON format. These are similar to the policies that you create for IAM users. If you change a role, the
change is propagated to all instances.

When creating IAM roles, associate least privilege IAM policies that restrict access to the specific API calls
the application requires.

You cannot attach multiple IAM roles to a single instance, but you can attach a single IAM role to
multiple instances. For more information about creating and using IAM roles, see Roles in the IAM User
Guide.

You can apply resource-level permissions to your IAM policies to control the users' ability to attach,
replace, or detach IAM roles for an instance. For more information, see Supported resource-level
permissions for Amazon EC2 API actions (p. 981) and the following example: Example: Work with IAM
roles (p. 1018).

Contents
• Instance profiles (p. 1032)
• Retrieve security credentials from instance metadata (p. 1033)
• Grant an IAM user permission to pass an IAM role to an instance (p. 1034)
• Work with IAM roles (p. 1034)

Instance profiles
Amazon EC2 uses an instance profile as a container for an IAM role. When you create an IAM role using
the IAM console, the console creates an instance profile automatically and gives it the same name as the

1032
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM roles

role to which it corresponds. If you use the Amazon EC2 console to launch an instance with an IAM role
or to attach an IAM role to an instance, you choose the role based on a list of instance profile names.

If you use the AWS CLI, API, or an AWS SDK to create a role, you create the role and instance profile as
separate actions, with potentially different names. If you then use the AWS CLI, API, or an AWS SDK to
launch an instance with an IAM role or to attach an IAM role to an instance, specify the instance profile
name.

An instance profile can contain only one IAM role. This limit cannot be increased.

For more information, see Instance Profiles in the IAM User Guide.

Retrieve security credentials from instance metadata

An application on the instance retrieves the security credentials provided by the role from the instance
metadata item iam/security-credentials/role-name. The application is granted the permissions
for the actions and resources that you've defined for the role through the security credentials associated
with the role. These security credentials are temporary and we rotate them automatically. We make new
credentials available at least five minutes before the expiration of the old credentials.
Warning
If you use services that use instance metadata with IAM roles, ensure that you don't expose your
credentials when the services make HTTP calls on your behalf. The types of services that could
expose your credentials include HTTP proxies, HTML/CSS validator services, and XML processors
that support XML inclusion.

The following command retrieves the security credentials for an IAM role named s3access.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" -v http://169.254.169.254/latest/meta-
data/iam/security-credentials/s3access

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/iam/security-credentials/

s3access

The following is example output.

{
"Code" : "Success",
"LastUpdated" : "2012-04-26T16:39:16Z",
"Type" : "AWS-HMAC",
"AccessKeyId" : "ASIAIOSFODNN7EXAMPLE",
"SecretAccessKey" : "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY",
"Token" : "token",
"Expiration" : "2017-05-17T15:09:54Z"
}

For applications, AWS CLI, and Tools for Windows PowerShell commands that run on the instance, you
do not have to explicitly get the temporary security credentials—the AWS SDKs, AWS CLI, and Tools for
Windows PowerShell automatically get the credentials from the EC2 instance metadata service and use
them. To make a call outside of the instance using temporary security credentials (for example, to test
IAM policies), you must provide the access key, secret key, and the session token. For more information,
see Using Temporary Security Credentials to Request Access to AWS Resources in the IAM User Guide.

1033
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM roles

For more information about instance metadata, see Instance metadata and user data (p. 704). For
information about the instance metadata IP address, see Retrieve instance metadata (p. 711).

Grant an IAM user permission to pass an IAM role to an instance

To enable an IAM user to launch an instance with an IAM role or to attach or replace an IAM role for an
existing instance, you must grant the user permission to pass the role to the instance.

The following IAM policy grants users permission to launch instances (ec2:RunInstances)
with an IAM role, or to attach or replace an IAM role for an existing instance
(ec2:AssociateIamInstanceProfile and ec2:ReplaceIamInstanceProfileAssociation).

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:RunInstances",
"ec2:AssociateIamInstanceProfile",
"ec2:ReplaceIamInstanceProfileAssociation"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": "iam:PassRole",
"Resource": "*"
}
]
}

This policy grants IAM users access to all your roles by specifying the resource as "*" in the policy.
However, consider whether users who launch instances with your roles (ones that exist or that you create
later on) might be granted permissions that they don't need or shouldn't have.

Work with IAM roles

You can create an IAM role and attach it to an instance during or after launch. You can also replace or
detach an IAM role for an instance.

Contents
• Create an IAM role (p. 1034)
• Launch an instance with an IAM role (p. 1036)
• Attach an IAM role to an instance (p. 1037)
• Replace an IAM role (p. 1038)
• Detach an IAM role (p. 1039)

Create an IAM role

You must create an IAM role before you can launch an instance with that role or attach it to an instance.

To create an IAM role using the IAM console

1. Open the IAM console at https://console.aws.amazon.com/iam/.

2. In the navigation pane, choose Roles, Create role.

1034
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM roles

3. On the Select role type page, choose EC2 and the EC2 use case. Choose Next: Permissions.
4. On the Attach permissions policy page, select an AWS managed policy that grants your instances
access to the resources that they need.
5. On the Review page, enter a name for the role and choose Create role.

Alternatively, you can use the AWS CLI to create an IAM role. The following example creates an IAM role
with a policy that allows the role to use an Amazon S3 bucket.

To create an IAM role and instance profile (AWS CLI)

1. Create the following trust policy and save it in a text file named ec2-role-trust-policy.json.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Principal": { "Service": "ec2.amazonaws.com"},
"Action": "sts:AssumeRole"
}
]
}

2. Create the s3access role and specify the trust policy that you created using the create-role
command.

aws iam create-role --role-name s3access --assume-role-policy-document file://ec2-role-

trust-policy.json
{
"Role": {
"AssumeRolePolicyDocument": {
"Version": "2012-10-17",
"Statement": [
{
"Action": "sts:AssumeRole",
"Effect": "Allow",
"Principal": {
"Service": "ec2.amazonaws.com"
}
}
]
},
"RoleId": "AROAIIZKPBKS2LEXAMPLE",
"CreateDate": "2013-12-12T23:46:37.247Z",
"RoleName": "s3access",
"Path": "/",
"Arn": "arn:aws:iam::123456789012:role/s3access"
}
}

3. Create an access policy and save it in a text file named ec2-role-access-policy.json. For
example, this policy grants administrative permissions for Amazon S3 to applications running on the
instance.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": ["s3:*"],
"Resource": ["*"]

1035
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM roles

}
]
}

4. Attach the access policy to the role using the put-role-policy command.

aws iam put-role-policy --role-name s3access --policy-name S3-Permissions --policy-

document file://ec2-role-access-policy.json

5. Create an instance profile named s3access-profile using the create-instance-profile command.

aws iam create-instance-profile --instance-profile-name s3access-profile

{
"InstanceProfile": {
"InstanceProfileId": "AIPAJTLBPJLEGREXAMPLE",
"Roles": [],
"CreateDate": "2013-12-12T23:53:34.093Z",
"InstanceProfileName": "s3access-profile",
"Path": "/",
"Arn": "arn:aws:iam::123456789012:instance-profile/s3access-profile"
}
}

6. Add the s3access role to the s3access-profile instance profile.

aws iam add-role-to-instance-profile --instance-profile-name s3access-profile --role-

name s3access

Alternatively, you can use the following AWS Tools for Windows PowerShell commands:

• New-IAMRole
• Register-IAMRolePolicy
• New-IAMInstanceProfile

Launch an instance with an IAM role

After you've created an IAM role, you can launch an instance, and associate that role with the instance
during launch.
Important
After you create an IAM role, it might take several seconds for the permissions to propagate. If
your first attempt to launch an instance with a role fails, wait a few seconds before trying again.
For more information, see Troubleshooting Working with Roles in the IAM User Guide.

To launch an instance with an IAM role (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. On the dashboard, choose Launch instance.
3. Select an AMI and instance type and then choose Next: Configure Instance Details.
4. On the Configure Instance Details page, for IAM role, select the IAM role that you created.
Note
The IAM role list displays the name of the instance profile that you created when you
created your IAM role. If you created your IAM role using the console, the instance profile
was created for you and given the same name as the role. If you created your IAM role using
the AWS CLI, API, or an AWS SDK, you may have named your instance profile differently.

1036
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM roles

5. Configure any other details, then follow the instructions through the rest of the wizard, or choose
Review and Launch to accept default settings and go directly to the Review Instance Launch page.
6. Review your settings, then choose Launch to choose a key pair and launch your instance.
7. If you are using the Amazon EC2 API actions in your application, retrieve the AWS security
credentials made available on the instance and use them to sign the requests. The AWS SDK does
this for you.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-

aws-ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" –v http://169.254.169.254/latest/
meta-data/iam/security-credentials/role_name

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/iam/security-

credentials/role_name

Alternatively, you can use the AWS CLI to associate a role with an instance during launch. You must
specify the instance profile in the command.

To launch an instance with an IAM role (AWS CLI)

1. Use the run-instances command to launch an instance using the instance profile. The following
example shows how to launch an instance with the instance profile.

aws ec2 run-instances \

--image-id ami-11aa22bb \
--iam-instance-profile Name="s3access-profile" \
--key-name my-key-pair \
--security-groups my-security-group \
--subnet-id subnet-1a2b3c4d

Alternatively, use the New-EC2Instance Tools for Windows PowerShell command.

2. If you are using the Amazon EC2 API actions in your application, retrieve the AWS security
credentials made available on the instance and use them to sign the requests. The AWS SDK does
this for you.

curl http://169.254.169.254/latest/meta-data/iam/security-credentials/role_name

Attach an IAM role to an instance

To attach an IAM role to an instance that has no role, the instance can be in the stopped or running
state.

New console

To attach an IAM role to an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance, choose Actions, Security, Modify IAM role.

1037
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM roles

4. Select the IAM role to attach to your instance, and choose Save.

Old console

To attach an IAM role to an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance, choose Actions, Instance Settings, Attach/Replace IAM role.
4. Select the IAM role to attach to your instance, and choose Apply.

To attach an IAM role to an instance (AWS CLI)

1. If required, describe your instances to get the ID of the instance to which to attach the role.

aws ec2 describe-instances

2. Use the associate-iam-instance-profile command to attach the IAM role to the instance by specifying
the instance profile. You can use the Amazon Resource Name (ARN) of the instance profile, or you
can use its name.

aws ec2 associate-iam-instance-profile \

--instance-id i-1234567890abcdef0 \
--iam-instance-profile Name="TestRole-1"

{
"IamInstanceProfileAssociation": {
"InstanceId": "i-1234567890abcdef0",
"State": "associating",
"AssociationId": "iip-assoc-0dbd8529a48294120",
"IamInstanceProfile": {
"Id": "AIPAJLNLDX3AMYZNWYYAY",
"Arn": "arn:aws:iam::123456789012:instance-profile/TestRole-1"
}
}
}

Alternatively, use the following Tools for Windows PowerShell commands:

• Get-EC2Instance
• Register-EC2IamInstanceProfile

Replace an IAM role

To replace the IAM role on an instance that already has an attached IAM role, the instance must be in the
running state. You can do this if you want to change the IAM role for an instance without detaching
the existing one first. For example, you can do this to ensure that API actions performed by applications
running on the instance are not interrupted.

New console

To replace an IAM role for an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.

1038
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM roles

3. Select the instance, choose Actions, Security, Modify IAM role.

4. Select the IAM role to attach to your instance, and choose Save.

Old console

To replace an IAM role for an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance, choose Actions, Instance Settings, Attach/Replace IAM role.
4. Select the IAM role to attach to your instance, and choose Apply.

To replace an IAM role for an instance (AWS CLI)

1. If required, describe your IAM instance profile associations to get the association ID for the IAM
instance profile to replace.

aws ec2 describe-iam-instance-profile-associations

2. Use the replace-iam-instance-profile-association command to replace the IAM instance profile by

specifying the association ID for the existing instance profile and the ARN or name of the instance
profile that should replace it.

aws ec2 replace-iam-instance-profile-association \

--association-id iip-assoc-0044d817db6c0a4ba \
--iam-instance-profile Name="TestRole-2"

{
"IamInstanceProfileAssociation": {
"InstanceId": "i-087711ddaf98f9489",
"State": "associating",
"AssociationId": "iip-assoc-09654be48e33b91e0",
"IamInstanceProfile": {
"Id": "AIPAJCJEDKX7QYHWYK7GS",
"Arn": "arn:aws:iam::123456789012:instance-profile/TestRole-2"
}
}
}

Alternatively, use the following Tools for Windows PowerShell commands:

• Get-EC2IamInstanceProfileAssociation
• Set-EC2IamInstanceProfileAssociation

Detach an IAM role

You can detach an IAM role from a running or stopped instance.

New console

To detach an IAM role from an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.

1039
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
IAM roles

3. Select the instance, choose Actions, Security, Modify IAM role.

4. For IAM role, choose No IAM Role. Choose Save.
5. In the confirmation dialog box, enter Detach, and then choose Detach.

Old console

To detach an IAM role from an instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance, choose Actions, Instance Settings, Attach/Replace IAM role.
4. For IAM role, choose No Role. Choose Apply.
5. In the confirmation dialog box, choose Yes, Detach.

To detach an IAM role from an instance (AWS CLI)

1. If required, use describe-iam-instance-profile-associations to describe your IAM instance profile

associations and get the association ID for the IAM instance profile to detach.

aws ec2 describe-iam-instance-profile-associations

{
"IamInstanceProfileAssociations": [
{
"InstanceId": "i-088ce778fbfeb4361",
"State": "associated",
"AssociationId": "iip-assoc-0044d817db6c0a4ba",
"IamInstanceProfile": {
"Id": "AIPAJEDNCAA64SSD265D6",
"Arn": "arn:aws:iam::123456789012:instance-profile/TestRole-2"
}
}
]
}

2. Use the disassociate-iam-instance-profile command to detach the IAM instance profile using its
association ID.

aws ec2 disassociate-iam-instance-profile --association-id iip-assoc-0044d817db6c0a4ba

{
"IamInstanceProfileAssociation": {
"InstanceId": "i-087711ddaf98f9489",
"State": "disassociating",
"AssociationId": "iip-assoc-0044d817db6c0a4ba",
"IamInstanceProfile": {
"Id": "AIPAJEDNCAA64SSD265D6",
"Arn": "arn:aws:iam::123456789012:instance-profile/TestRole-2"
}
}
}

Alternatively, use the following Tools for Windows PowerShell commands:

• Get-EC2IamInstanceProfileAssociation
• Unregister-EC2IamInstanceProfile

1040
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Network access

Authorize inbound traffic for your Linux instances

Security groups enable you to control traffic to your instance, including the kind of traffic that can reach
your instance. For example, you can allow computers from only your home network to access your
instance using SSH. If your instance is a web server, you can allow all IP addresses to access your instance
using HTTP or HTTPS, so that external users can browse the content on your web server.

Your default security groups and newly created security groups include default rules that do not
enable you to access your instance from the internet. For more information, see Default security
groups (p. 1062) and Custom security groups (p. 1063). To enable network access to your instance, you
must allow inbound traffic to your instance. To open a port for inbound traffic, add a rule to a security
group that you associated with your instance when you launched it.

To connect to your instance, you must set up a rule to authorize SSH traffic from your computer's public
IPv4 address. To allow SSH traffic from additional IP address ranges, add another rule for each range you
need to authorize.

If you've enabled your VPC for IPv6 and launched your instance with an IPv6 address, you can connect to
your instance using its IPv6 address instead of a public IPv4 address. Your local computer must have an
IPv6 address and must be configured to use IPv6.

If you need to enable network access to a Windows instance, see Authorizing inbound traffic for your
Windows instances in the Amazon EC2 User Guide for Windows Instances.

Before you start

Decide who requires access to your instance; for example, a single host or a specific network that you
trust such as your local computer's public IPv4 address. The security group editor in the Amazon EC2
console can automatically detect the public IPv4 address of your local computer for you. Alternatively,
you can use the search phrase "what is my IP address" in an internet browser, or use the following
service: Check IP. If you are connecting through an ISP or from behind your firewall without a static IP
address, you need to find out the range of IP addresses used by client computers.
Warning
If you use 0.0.0.0/0, you enable all IPv4 addresses to access your instance using SSH. If you
use ::/0, you enable all IPv6 address to access your instance. This is acceptable for a short time
in a test environment, but it's unsafe for production environments. In production, you authorize
only a specific IP address or range of addresses to access your instance.

Decide whether you'll support SSH access to your instances using EC2 Instance Connect. If you will not
use EC2 Instance Connect, consider uninstalling it or denying the following action in your IAM policies:
ec2-instance-connect:SendSSHPublicKey. For more information, see Uninstall EC2 Instance
Connect (p. 613) and Configure IAM Permissions for EC2 Instance Connect (p. 608).

Add a rule for inbound SSH traffic to a Linux instance

Security groups act as a firewall for associated instances, controlling both inbound and outbound traffic
at the instance level. You must add rules to a security group to enable you to connect to your Linux
instance from your IP address using SSH.

New console

To add a rule to a security group for inbound SSH traffic over IPv4 (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select your instance and, in bottom half of the screen, choose the Security tab. Security groups
lists the security groups that are associated with the instance. Inbound rules displays a list of
the inbound rules that are in effect for the instance.

1041
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Network access

4. For the security group to which you'll add the new rule, choose the security group ID link to
open the security group.
5. On the Inbound rules tab, choose Edit inbound rules.
6. On the Edit inbound rules page, do the following:

a. Choose Add rule.

b. For Type, choose SSH.
c. For Source, choose My IP to automatically populate the field with the public IPv4 address
of your local computer.

Alternatively, for Source, choose Custom and enter the public IPv4 address of your
computer or network in CIDR notation. For example, if your IPv4 address is 203.0.113.25,
enter 203.0.113.25/32 to list this single IPv4 address in CIDR notation. If your company
allocates addresses from a range, enter the entire range, such as 203.0.113.0/24.

For information about finding your IP address, see Before you start (p. 1041).
d. Choose Save rules.

Old console

To add a rule to a security group for inbound SSH traffic over IPv4 (console)

1. In the navigation pane of the Amazon EC2 console, choose Instances. Select your instance and
look at the Description tab; Security groups lists the security groups that are associated with
the instance. Choose view inbound rules to display a list of the rules that are in effect for the
instance.
2. In the navigation pane, choose Security Groups. Select one of the security groups associated
with your instance.
3. In the details pane, on the Inbound tab, choose Edit. In the dialog, choose Add Rule, and then
choose SSH from the Type list.
4. In the Source field, choose My IP to automatically populate the field with the public IPv4
address of your local computer. Alternatively, choose Custom and specify the public IPv4
address of your computer or network in CIDR notation. For example, if your IPv4 address
is 203.0.113.25, specify 203.0.113.25/32 to list this single IPv4 address in CIDR
notation. If your company allocates addresses from a range, specify the entire range, such as
203.0.113.0/24.

For information about finding your IP address, see Before you start (p. 1041).
5. Choose Save.

If you launched an instance with an IPv6 address and want to connect to your instance using its IPv6
address, you must add rules that allow inbound IPv6 traffic over SSH.

New console

To add a rule to a security group for inbound SSH traffic over IPv6 (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select your instance and, in bottom half of the screen, choose the Security tab. Security groups
lists the security groups that are associated with the instance. Inbound rules displays a list of
the inbound rules that are in effect for the instance.
4. For the security group to which you'll add the new rule, choose the security group ID link to
open the security group.

1042
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Network access

5. On the Inbound rules tab, choose Edit inbound rules.

6. On the Edit inbound rules page, do the following:

a. Choose Add rule.

b. For Type, choose SSH.
c. For Source, choose Custom and enter the IPv6 address of your computer in CIDR notation.
For example, if your IPv6 address is 2001:db8:1234:1a00:9691:9503:25ad:1761,
enter 2001:db8:1234:1a00:9691:9503:25ad:1761/128 to list the single IP address
in CIDR notation. If your company allocates addresses from a range, enter the entire range,
such as 2001:db8:1234:1a00::/64.
d. Choose Save rules.

Old console

To add a rule to a security group for inbound SSH traffic over IPv6 (console)

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups. Select the security group for your instance.
3. Choose Inbound, Edit, Add Rule.
4. For Type, choose SSH.
5. In the Source field, specify the IPv6 address of your computer in CIDR notation. For
example, if your IPv6 address is 2001:db8:1234:1a00:9691:9503:25ad:1761, specify
2001:db8:1234:1a00:9691:9503:25ad:1761/128 to list the single IP address in CIDR
notation. If your company allocates addresses from a range, specify the entire range, such as
2001:db8:1234:1a00::/64.
6. Choose Save.

Note
Be sure to run the following commands on your local system, not on the instance itself. For
more information about these command line interfaces, see Access Amazon EC2 (p. 3).

To add a rule to a security group using the command line

1. Find the security group that is associated with your instance using one of the following commands:

• describe-instance-attribute (AWS CLI)

aws ec2 describe-instance-attribute --instance-id instance_id --attribute groupSet

• Get-EC2InstanceAttribute (AWS Tools for Windows PowerShell)

PS C:\> (Get-EC2InstanceAttribute -InstanceId instance_id -Attribute groupSet).Groups

Both commands return a security group ID, which you use in the next step.
2. Add the rule to the security group using one of the following commands:

• authorize-security-group-ingress (AWS CLI)

aws ec2 authorize-security-group-ingress --group-id security_group_id --protocol tcp

--port 22 --cidr cidr_ip_range

• Grant-EC2SecurityGroupIngress (AWS Tools for Windows PowerShell)

1043
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Key pairs

The Grant-EC2SecurityGroupIngress command needs an IpPermission parameter, which

describes the protocol, port range, and IP address range to be used for the security group rule. The
following command creates the IpPermission parameter:

PS C:\> $ip1 = @{ IpProtocol="tcp"; FromPort="22"; ToPort="22";

IpRanges="cidr_ip_range" }

PS C:\> Grant-EC2SecurityGroupIngress -GroupId security_group_id -IpPermission

@($ip1)

Assign a security group to an instance

You can assign a security group to an instance when you launch the instance. When you add or remove
rules, those changes are automatically applied to all instances to which you've assigned the security
group.

After you launch an instance, you can change its security groups. For more information, see Changing an
instance's security groups in the Amazon VPC User Guide.

Amazon EC2 key pairs and Linux instances

A key pair, consisting of a private key and a public key, is a set of security credentials that you use to
prove your identity when connecting to an instance. Amazon EC2 stores the public key, and you store
the private key. You use the private key, instead of a password, to securely access your instances. Anyone
who possesses your private keys can connect to your instances, so it's important that you store your
private keys in a secure place.

When you launch an instance, you are prompted for a key pair (p. 531). If you plan to connect to the
instance using SSH, you must specify a key pair. You can choose an existing key pair or create a new one.
When your instance boots for the first time, the content of the public key that you specified at launch
is placed on your Linux instance in an entry within ~/.ssh/authorized_keys. When you connect
to your Linux instance using SSH, to log in you must specify the private key that corresponds to the
public key content. For more information about connecting to your instance, see Connect to your Linux
instance (p. 598). For more information about key pairs and Windows instances, see Amazon EC2 key
pairs and Windows instances in the Amazon EC2 User Guide for Windows Instances

Because Amazon EC2 doesn't keep a copy of your private key, there is no way to recover a private key if
you lose it. However, there can still be a way to connect to instances for which you've lost the private key.
For more information, see Connect to your Linux instance if you lose your private key (p. 1053).

The keys that Amazon EC2 uses are 2048-bit SSH-2 RSA keys. You can have up to 5,000 key pairs per
Region.

Contents
• Create or import a key pair (p. 1045)
• Tag a key pair (p. 1048)
• Retrieve the public key for your key pair (p. 1050)
• Retrieve the public key for your key pair through instance metadata (p. 1050)
• Locate the public key on an instance (p. 1051)
• Identify the key pair that was specified at launch (p. 1051)
• (Optional) Verify your key pair's fingerprint (p. 1052)
• Add or replace a key pair for your instance (p. 1052)

1044
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create or import a key pair

• Connect to your Linux instance if you lose your private key (p. 1053)
• Delete your key pair (p. 1057)

Create or import a key pair

You can use Amazon EC2 to create a new key pair, or you can import an existing key pair.

Options
• Option 1: Create a key pair using Amazon EC2 (p. 1045)
• Option 2: Import your own public key to Amazon EC2 (p. 1046)

Option 1: Create a key pair using Amazon EC2

You can create a key pair using one of the following methods.

New console

To create your key pair

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, under NETWORK & SECURITY, choose Key Pairs.
3. Choose Create key pair.
4. For Name, enter a descriptive name for the key pair. Amazon EC2 associates the public key with
the name that you specify as the key name. A key name can include up to 255 ASCII characters.
It can’t include leading or trailing spaces.
5. For File format, choose the format in which to save the private key. To save the private key in
a format that can be used with OpenSSH, choose pem. To save the private key in a format that
can be used with PuTTY, choose ppk.
6. Choose Create key pair.
7. The private key file is automatically downloaded by your browser. The base file name is the
name you specified as the name of your key pair, and the file name extension is determined by
the file format you chose. Save the private key file in a safe place.
Important
This is the only chance for you to save the private key file.
8. If you will use an SSH client on a macOS or Linux computer to connect to your Linux instance,
use the following command to set the permissions of your private key file so that only you can
read it.

chmod 400 my-key-pair.pem

If you do not set these permissions, then you cannot connect to your instance using this key pair.
For more information, see Error: Unprotected private key file (p. 1336).

Old console

To create your key pair

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, under NETWORK & SECURITY, choose Key Pairs.
3. Choose Create Key Pair.

1045
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create or import a key pair

4. For Key pair name, enter a descriptive name for the key pair, and then choose Create. A key
name can include up to 255 ASCII characters. It can’t include leading or trailing spaces.
5. The private key file is automatically downloaded by your browser. The base file name is the
name you specified as the name of your key pair, and the file name extension is .pem. Save the
private key file in a safe place.
Important
This is the only chance for you to save the private key file.
6. If you will use an SSH client on a macOS or Linux computer to connect to your Linux instance,
use the following command to set the permissions of your private key file so that only you can
read it.

chmod 400 my-key-pair.pem

If you do not set these permissions, then you cannot connect to your instance using this key pair.
For more information, see Error: Unprotected private key file (p. 1336).

AWS CLI

To create your key pair

1. Use the create-key-pair AWS CLI command as follows to generate the key and save it to a .pem
file.

aws ec2 create-key-pair --key-name my-key-pair --query "KeyMaterial" --output text

> my-key-pair.pem

2. If you will use an SSH client on a macOS or Linux computer to connect to your Linux instance,
use the following command to set the permissions of your private key file so that only you can
read it.

chmod 400 my-key-pair.pem

If you do not set these permissions, then you cannot connect to your instance using this key pair.
For more information, see Error: Unprotected private key file (p. 1336).

PowerShell

To create your key pair

Use the New-EC2KeyPair AWS Tools for Windows PowerShell command as follows to generate the
key and save it to a .pem file.

PS C:\> (New-EC2KeyPair -KeyName "my-key-pair").KeyMaterial | Out-File -Encoding ascii

-FilePath C:\path\my-key-pair.pem

Option 2: Import your own public key to Amazon EC2

Instead of using Amazon EC2 to create your key pair, you can create an RSA key pair using a third-party
tool and then import the public key to Amazon EC2.

Requirements for key pairs

• The following formats are supported:

1046
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create or import a key pair

• OpenSSH public key format (the format in ~/.ssh/authorized_keys). If you connect using SSH
while using the EC2 Instance Connect API, the SSH2 format is also supported.
• Base64 encoded DER format
• SSH public key file format as specified in RFC4716
• SSH private key file format must be PEM
• Create an RSA key. Amazon EC2 does not accept DSA keys.
• The supported lengths are 1024, 2048, and 4096. If you connect using SSH while using the EC2
Instance Connect API, the supported lengths are 2048 and 4096.

To create a key pair using a third-party tool

1. Generate a key pair with a third-party tool of your choice. For example, you can use ssh-keygen (a
tool provided with the standard OpenSSH installation). Alternatively, Java, Ruby, Python, and many
other programming languages provide standard libraries that you can use to create an RSA key pair.
Important
The private key must be in the PEM format. For example, use ssh-keygen -m PEM to
generate the OpenSSH key in the PEM format.
2. Save the public key to a local file. For example, ~/.ssh/my-key-pair.pub. The file name
extension for this file is not important.
3. Save the private key to a local file that has the .pem extension. For example, ~/.ssh/my-key-
pair.pem.
Important
Save the private key file in a safe place. You'll need to provide the name of your public key
when you launch an instance and the corresponding private key each time you connect to
the instance.

After you have created the key pair, use one of the following methods to import your public key to
Amazon EC2.

New console

To import the public key

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Key Pairs.
3. Choose Import key pair.
4. For Name, enter a descriptive name for the public key. The name can include up to 255 ASCII
characters. It can’t include leading or trailing spaces.
Note
When you connect to your instance from the EC2 console, the console suggests this
name for the name of your private key file.
5. Either choose Browse to navigate to and select your public key, or paste the contents of your
public key into the Public key contents field.
6. Choose Import key pair.
7. Verify that the public key that you imported appears in the list of key pairs.

Old console

To import the public key

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

1047
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tag a key pair

2. In the navigation pane, under NETWORK & SECURITY, choose Key Pairs.
3. Choose Import Key Pair.
4. In the Import Key Pair dialog box, choose Browse, and select the public key file that you saved
previously. Enter a name for the key pair in the Key pair name field, and choose Import. The
name can include up to 255 ASCII characters. It can’t include leading or trailing spaces.
5. Verify that the key pair you imported appears in the list of key pairs.

AWS CLI

To import the public key

Use the import-key-pair AWS CLI command.

To verify that the key pair was imported successfully

Use the describe-key-pairs AWS CLI command.

PowerShell

To import the public key

Use the Import-EC2KeyPair AWS Tools for Windows PowerShell command.

To verify that the key pair was imported successfully

Use the Get-EC2KeyPair AWS Tools for Windows PowerShell command.

Tag a key pair

To help categorize and manage your existing key pairs, you can tag them with custom metadata. For
more information about how tags work, see Tag your Amazon EC2 resources (p. 1311).

You can view, add, and delete tags using the new console and the command line tools.

New console

To view, add, or delete a tag for an existing key pair

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Key Pairs.
3. Select a key pair, and then choose Actions, Manage tags.
4. The Manage tags page displays any tags that are assigned to the key pair.

• To add a tag, choose Add tag, and then enter the tag key and value. You can add up to 50
tags per key pair. For more information, see Tag restrictions (p. 1315).
• To delete a tag, choose Remove next to the tag to delete.
5. Choose Save.

AWS CLI

To view key pair tags

Use the describe-tags AWS CLI command. In the following example, you describe the tags for all of
your key pairs.

$ aws ec2 describe-tags --filters "Name=resource-type,Values=key-pair"

1048
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tag a key pair

{
"Tags": [
{
"Key": "Environment",
"ResourceId": "key-0123456789EXAMPLE",
"ResourceType": "key-pair",
"Value": "Production"
},
{
"Key": "Environment",
"ResourceId": "key-9876543210EXAMPLE",
"ResourceType": "key-pair",
"Value": "Production"
}]
}

To describe the tags for a specific key pair

Use the describe-key-pairs AWS CLI command.

$ aws ec2 describe-key-pairs --key-pair-ids key-0123456789EXAMPLE

{
"KeyPairs": [
{
"KeyName": "MyKeyPair",
"KeyFingerprint":
"1f:51:ae:28:bf:89:e9:d8:1f:25:5d:37:2d:7d:b8:ca:9f:f5:f1:6f",
"KeyPairId": "key-0123456789EXAMPLE",
"Tags": [
{
"Key": "Environment",
"Value": "Production"
}]
}]
}

To tag an existing key pair

Use the create-tags AWS CLI command. In the following example, the existing key pair is tagged with
Key=Cost-Center and Value=CC-123.

$ aws ec2 create-tags --resources key-0123456789EXAMPLE --tags Key=Cost-

Center,Value=CC-123

To delete a tag from a key pair

Use the delete-tags AWS CLI command. For examples, see Examples in the AWS CLI Command
Reference.
PowerShell

To view key pair tags

Use the Get-EC2Tag command.

To describe the tags for a specific key pair

Use the Get-EC2KeyPair command.

To tag an existing key pair

1049
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Retrieve the public key for your key pair

Use the New-EC2Tag command.

To delete a tag from a key pair

Use the Remove-EC2Tag command.

Retrieve the public key for your key pair

On your local Linux or macOS computer, you can use the ssh-keygen command to retrieve the public key
for your key pair. Specify the path where you downloaded your private key (the .pem file).

ssh-keygen -y -f /path_to_key_pair/my-key-pair.pem

The command returns the public key, as shown in the following example.

ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAABAQClKsfkNkuSevGj3eYhCe53pcjqP3maAhDFcvBS7O6V
hz2ItxCih+PnDSUaw+WNQn/mZphTk/a/gU8jEzoOWbkM4yxyb/wB96xbiFveSFJuOp/d6RJhJOI0iBXr
lsLnBItntckiJ7FbtxJMXLvvwJryDUilBMTjYtwB+QhYXUMOzce5Pjz5/i8SeJtjnV3iAoG/cQk+0FzZ
qaeJAAHco+CY/5WrUBkrHmFJr6HcXkvJdWPkYQS3xqC0+FmUZofz221CBt5IMucxXPkX4rWi+z7wB3Rb
BQoQzd8v7yeb7OzlPnWOyN0qFU0XA246RA8QFYiCNYwI3f05p6KLxEXAMPLE

If the command fails, run the following command to ensure that you've changed the permissions on your
key pair file so that only you can view it.

chmod 400 my-key-pair.pem

Retrieve the public key for your key pair through

instance metadata
The public key that you specified when you launched an instance is also available to you through its
instance metadata. To view the public key that you specified when launching the instance, use the
following command from your instance:

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" –v http://169.254.169.254/latest/meta-
data/public-keys/0/openssh-key

The following is an example output.

ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAABAQClKsfkNkuSevGj3eYhCe53pcjqP3maAhDFcvBS7O6V
hz2ItxCih+PnDSUaw+WNQn/mZphTk/a/gU8jEzoOWbkM4yxyb/wB96xbiFveSFJuOp/d6RJhJOI0iBXr
lsLnBItntckiJ7FbtxJMXLvvwJryDUilBMTjYtwB+QhYXUMOzce5Pjz5/i8SeJtjnV3iAoG/cQk+0FzZ
qaeJAAHco+CY/5WrUBkrHmFJr6HcXkvJdWPkYQS3xqC0+FmUZofz221CBt5IMucxXPkX4rWi+z7wB3Rb
BQoQzd8v7yeb7OzlPnWOyN0qFU0XA246RA8QFYiCNYwI3f05p6KLxEXAMPLE my-key-pair

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/public-keys/0/openssh-key

The following is an example output.

ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAABAQClKsfkNkuSevGj3eYhCe53pcjqP3maAhDFcvBS7O6V

1050
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Locate the public key on an instance

hz2ItxCih+PnDSUaw+WNQn/mZphTk/a/gU8jEzoOWbkM4yxyb/wB96xbiFveSFJuOp/d6RJhJOI0iBXr
lsLnBItntckiJ7FbtxJMXLvvwJryDUilBMTjYtwB+QhYXUMOzce5Pjz5/i8SeJtjnV3iAoG/cQk+0FzZ
qaeJAAHco+CY/5WrUBkrHmFJr6HcXkvJdWPkYQS3xqC0+FmUZofz221CBt5IMucxXPkX4rWi+z7wB3Rb
BQoQzd8v7yeb7OzlPnWOyN0qFU0XA246RA8QFYiCNYwI3f05p6KLxEXAMPLE my-key-pair

If you change the key pair that you use to connect to the instance, we don't update the instance
metadata to show the new public key. Instead, the instance metadata continues to show the public key
for the key pair that you specified when you launched the instance. For more information, see Retrieve
instance metadata (p. 711).

Alternatively, on a Linux instance, the public key content is placed in an entry within ~/.ssh/
authorized_keys. You can open this file in an editor. The following is an example entry for the key
pair named my-key-pair. It consists of the public key followed by the name of the key pair.

ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAABAQClKsfkNkuSevGj3eYhCe53pcjqP3maAhDFcvBS7O6V
hz2ItxCih+PnDSUaw+WNQn/mZphTk/a/gU8jEzoOWbkM4yxyb/wB96xbiFveSFJuOp/d6RJhJOI0iBXr
lsLnBItntckiJ7FbtxJMXLvvwJryDUilBMTjYtwB+QhYXUMOzce5Pjz5/i8SeJtjnV3iAoG/cQk+0FzZ
qaeJAAHco+CY/5WrUBkrHmFJr6HcXkvJdWPkYQS3xqC0+FmUZofz221CBt5IMucxXPkX4rWi+z7wB3Rb
BQoQzd8v7yeb7OzlPnWOyN0qFU0XA246RA8QFYiCNYwI3f05p6KLxEXAMPLE my-key-pair

Locate the public key on an instance

When you launch an instance, you are prompted for a key pair (p. 531). If you plan to connect to the
instance using SSH, you must specify a key pair. You can choose an existing key pair or create a new one.
When your instance boots for the first time, the content of the public key that you specified at launch is
placed on your Linux instance in an entry within ~/.ssh/authorized_keys.

To locate the public key on an instance

1. Connect to your instance. For more information, see Connect to your Linux instance (p. 598).
2. In the terminal window, open the authorized_keys file using your favorite text editor (such as vim
or nano).

[ec2-user ~]$ nano ~/.ssh/authorized_keys

The authorized_keys file opens, displaying the public key, as shown in the following example.

ssh-rsa
AAAAB3NzaC1yc2EAAAADAQABAAABAQClKsfkNkuSevGj3eYhCe53pcjqP3maAhDFcvBS7O6Vhz2ItxCih
+PnDSUaw+WNQn/mZphTk/a/gU8jEzoOWbkM4yxyb/wB96xbiFveSFJuOp/
d6RJhJOI0iBXrlsLnBItntckiJ7FbtxJMXLvvwJryDUilBMTjYtwB+QhYXUMOzce5Pjz5/i8SeJtjnV3iAoG/
cQk+0FzZqaeJAAHco+CY/5WrUBkrHmFJr6HcXkvJdWPkYQS3xqC0+FmUZofz221CBt5IMucxXPkX4rWi
+z7wB3RbBQoQzd8v7yeb7OzlPnWOyN0qFU0XA246RA8QFYiCNYwI3f05p6KLxEXAMPLE

Identify the key pair that was specified at launch

When you launch an instance, you are prompted for a key pair (p. 531). If you plan to connect to the
instance using SSH, you must specify a key pair.

New console

To identify the key pair that was specified at launch

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

1051
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
(Optional) Verify your key pair's fingerprint

2. In the navigation pane, choose Instances, and then select your instance.
3. On the Details tab, under Instance details, the Key pair name field displays the name of the
key pair that you specified when you launched the instance. The value of the Key pair name
does not change even if you change the public key on the instance, or add key pairs.

Old console

To identify the key pair that was specified at launch

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and then select your instance.
3. On the Description tab, the Key pair name field displays the name of the key pair that you
specified when you launched the instance. The value of the Key pair name does not change
even if you change the public key on the instance, or add key pairs.

(Optional) Verify your key pair's fingerprint

On the Key Pairs page in the Amazon EC2 console, the Fingerprint column displays the fingerprints
generated from your key pairs. AWS calculates the fingerprint differently depending on whether the key
pair was generated by AWS or a third-party tool. If you created the key pair using AWS, the fingerprint is
calculated using an SHA-1 hash function. If you created the key pair with a third-party tool and uploaded
the public key to AWS, or if you generated a new public key from an existing AWS-created private key
and uploaded it to AWS, the fingerprint is calculated using an MD5 hash function.

You can use the SSH2 fingerprint that's displayed on the Key Pairs page to verify that the private key
you have on your local machine matches the public key stored in AWS. From the computer where you
downloaded the private key file, generate an SSH2 fingerprint from the private key file. The output
should match the fingerprint that's displayed in the console.

If you created your key pair using AWS, you can use the OpenSSL tools to generate a fingerprint as
shown in the following example.

$ openssl pkcs8 -in path_to_private_key -inform PEM -outform DER -topk8 -nocrypt | openssl
sha1 -c

If you created a key pair using a third-party tool and uploaded the public key to AWS, you can use the
OpenSSL tools to generate the fingerprint as shown in the following example.

$ openssl rsa -in path_to_private_key -pubout -outform DER | openssl md5 -c

If you created an OpenSSH key pair using OpenSSH 7.8 or later and uploaded the public key to AWS, you
can use ssh-keygen to generate the fingerprint as shown in the following example.

$ ssh-keygen -ef path_to_private_key -m PEM | openssl rsa -RSAPublicKey_in -outform DER |

openssl md5 -c

Add or replace a key pair for your instance

You can change the key pair that is used to access the default system account of your instance. For
example, if a user in your organization requires access to the system user account using a separate key
pair, you can add that key pair to your instance. Or, if someone has a copy of the .pem file and you want
to prevent them from connecting to your instance (for example, if they've left your organization), you
can replace the key pair with a new one.

1052
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect to your Linux instance if you lose your private key

To add or replace a key pair, you must be able to connect to your instance. If you've lost your existing
private key or you launched your instance without a key pair, you won't be able connect to your instance
and therefore won't be able to add or replace a key pair. If you've lost your existing private key, you
might be able to retrieve it. For more information, see Connect to your Linux instance if you lose your
private key (p. 1053). If you launched your instance without a key pair, you won't be able to connect to
the instance unless you chose an AMI that is configured to allow users another way to log in.
Note
These procedures are for modifying the key pair for the default user account, such as ec2-user.
For more information about adding user accounts to your instance, see Manage user accounts on
your Amazon Linux instance (p. 656).

To add or replace a key pair

1. Create a new key pair using the Amazon EC2 console (p. 1045) or a third-party tool (p. 1046).
2. Retrieve the public key from your new key pair. For more information, see Retrieve the public key for
your key pair (p. 1050).
3. Connect to your instance using your existing private key file.
4. Using a text editor of your choice, open the .ssh/authorized_keys file on the instance. Paste the
public key information from your new key pair underneath the existing public key information. Save
the file.
5. Disconnect from your instance, and test that you can connect to your instance using the new private
key file.
6. (Optional) If you're replacing an existing key pair, connect to your instance and delete the public key
information for the original key pair from the .ssh/authorized_keys file.

Note
If you're using an Auto Scaling group, ensure that the key pair you're replacing is not specified in
your launch template or launch configuration. Amazon EC2 Auto Scaling launches a replacement
instance if it detects an unhealthy instance; however, the instance launch fails if the key pair
cannot be found.

Connect to your Linux instance if you lose your

private key
If you lose the private key for an EBS-backed instance, you can regain access to your instance. You must
stop the instance, detach its root volume and attach it to another instance as a data volume, modify the
authorized_keys file with a new public key, move the volume back to the original instance, and restart
the instance. For more information about launching, connecting to, and stopping instances, see Instance
lifecycle (p. 520).

This procedure is not supported for instances with instance-store backed root volumes. To determine the
root device type of your instance, open the Amazon EC2 console, choose Instances, select the instance,
and check the value of Root device type in the details pane. The value is either ebs or instance
store. If the root device is an instance store volume, you cannot use this procedure to regain access to
your instance; you must have the private key to connect to the instance.

Steps for connecting to an EBS-backed instance with a different key pair

• Step 1: Create a new key pair (p. 1054)
• Step 2: Get information about the original instance and its root volume (p. 1054)
• Step 3: Stop the original instance (p. 1054)
• Step 4: Launch a temporary instance (p. 1055)
• Step 5: Detach the root volume from the original instance and attach it to the temporary
instance (p. 1055)

1053
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect to your Linux instance if you lose your private key

• Step 6: Add the new public key to authorized_keys on the original volume mounted to the temporary
instance (p. 1055)
• Step 7: Unmount and detach the original volume from the temporary instance, and reattach it to the
original instance (p. 1057)
• Step 8: Connect to the original instance using the new key pair (p. 1057)
• Step 9: Clean up (p. 1057)

Step 1: Create a new key pair

Create a new key pair using either the Amazon EC2 console or a third-party tool. If you want to name
your new key pair exactly the same as the lost private key, you must first delete the existing key pair. For
information about creating a new key pair, see Option 1: Create a key pair using Amazon EC2 (p. 1045)
or Option 2: Import your own public key to Amazon EC2 (p. 1046).

Step 2: Get information about the original instance and its root
volume
Make note of the following information because you'll need it to complete this procedure.

New console

To get information about your original instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Instances in the navigation pane, and then select the instance that you'd like to connect
to. (We'll refer to this as the original instance.)
3. On the Details tab, make note of the instance ID and AMI ID.
4. On the Networking tab, make note of the Availability Zone.
5. On the Storage tab, under Root device name, make note of the device name for the root
volume (for example, /dev/xvda). Then, under Block devices, find this device name and make
note of the volume ID (for example, vol-0a1234b5678c910de).

Old console

To get information about your original instance

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Instances in the navigation pane, and then select the instance that you'd like to connect
to. (We'll refer to this as the original instance.)
3. From the Description tab, save the following information that you need to complete this
procedure.

• Write down the instance ID, AMI ID, and Availability Zone of the original instance.
• In the Root device field, take note of the device name for the root volume (for example, /
dev/sda1 or /dev/xvda). Choose the link and write down the volume ID in the EBS ID field
(vol-xxxxxxxxxxxxxxxxx).

Step 3: Stop the original instance

Choose Instance state, Stop instance. If this option is disabled, either the instance is already stopped or
its root device is an instance store volume.

1054
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect to your Linux instance if you lose your private key

Warning
When you stop an instance, the data on any instance store volumes is erased. To keep data from
instance store volumes, be sure to back it up to persistent storage.

Step 4: Launch a temporary instance

Choose Launch instances, and then use the launch wizard to launch a temporary instance with the
following options:

• On the Choose an AMI page, select the same AMI that you used to launch the original instance. If
this AMI is unavailable, you can create an AMI that you can use from the stopped instance. For more
information, see Create an Amazon EBS-backed Linux AMI (p. 125).
• On the Choose an Instance Type page, leave the default instance type that the wizard selects for you.
• On the Configure Instance Details page, specify the same Availability Zone as the original instance. If
you're launching an instance in a VPC, select a subnet in this Availability Zone.
• On the Add Tags page, add the tag Name=Temporary to the instance to indicate that this is a
temporary instance.
• On the Review page, choose Launch. Create a new key pair, download it to a safe location on your
computer, and then choose Launch Instances.

Step 5: Detach the root volume from the original instance and
attach it to the temporary instance
1. In the navigation pane, choose Volumes and select the root device volume for the original instance
(you made note of its volume ID in a previous step). Choose Actions, Detach Volume, and then
select Yes, Detach. Wait for the state of the volume to become available. (You might need to
choose the Refresh icon.)
2. With the volume still selected, choose Actions, and then select Attach Volume. Select the instance
ID of the temporary instance, make note of the device name specified under Device (for example, /
dev/sdf), and then choose Attach.
Note
If you launched your original instance from an AWS Marketplace AMI and your volume
contains AWS Marketplace codes, you must first stop the temporary instance before you can
attach the volume.

Step 6: Add the new public key to authorized_keys on the

original volume mounted to the temporary instance
1. Connect to the temporary instance.
2. From the temporary instance, mount the volume that you attached to the instance so that you can
access its file system. For example, if the device name is /dev/sdf, use the following commands to
mount the volume as /mnt/tempvol.
Note
The device name might appear differently on your instance. For example, devices mounted
as /dev/sdf might show up as /dev/xvdf on the instance. Some versions of Red Hat (or
its variants, such as CentOS) might even increment the trailing letter by 4 characters, where
/dev/sdf becomes /dev/xvdk.

a. Use the lsblk command to determine if the volume is partitioned.

[ec2-user ~]$ lsblk

1055
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect to your Linux instance if you lose your private key

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT

xvda 202:0 0 8G 0 disk
##xvda1 202:1 0 8G 0 part /
xvdf 202:80 0 101G 0 disk
##xvdf1 202:81 0 101G 0 part
xvdg 202:96 0 30G 0 disk

In the preceding example, /dev/xvda and /dev/xvdf are partitioned volumes, and /dev/
xvdg is not. If your volume is partitioned, you mount the partition (/dev/xvdf1) instead of
the raw device (/dev/xvdf) in the next steps.
b. Create a temporary directory to mount the volume.

[ec2-user ~]$ sudo mkdir /mnt/tempvol

c. Mount the volume (or partition) at the temporary mount point, using the volume name or
device name that you identified earlier. The required command depends on your operating
system's file system.

• Amazon Linux, Ubuntu, and Debian

[ec2-user ~]$ sudo mount /dev/xvdf1 /mnt/tempvol

• Amazon Linux 2, CentOS, SUSE Linux 12, and RHEL 7.x

[ec2-user ~]$ sudo mount -o nouuid /dev/xvdf1 /mnt/tempvol

Note
If you get an error stating that the file system is corrupt, run the following command to use
the fsck utility to check the file system and repair any issues:

[ec2-user ~]$ sudo fsck /dev/xvdf1

3. From the temporary instance, use the following command to update authorized_keys on the
mounted volume with the new public key from the authorized_keys for the temporary instance.
Important
The following examples use the Amazon Linux user name ec2-user. You might need to
substitute a different user name, such as ubuntu for Ubuntu instances.

[ec2-user ~]$ cp .ssh/authorized_keys /mnt/tempvol/home/ec2-user/.ssh/authorized_keys

If this copy succeeded, you can go to the next step.

(Optional) Otherwise, if you don't have permission to edit files in /mnt/tempvol, you must update
the file using sudo and then check the permissions on the file to verify that you are able to log into
the original instance. Use the following command to check the permissions on the file.

[ec2-user ~]$ sudo ls -l /mnt/tempvol/home/ec2-user/.ssh

total 4
-rw------- 1 222 500 398 Sep 13 22:54 authorized_keys

In this example output, 222 is the user ID and 500 is the group ID. Next, use sudo to re-run the copy
command that failed.

1056
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Delete your key pair

[ec2-user ~]$ sudo cp .ssh/authorized_keys /mnt/tempvol/home/ec2-user/.ssh/

authorized_keys

Run the following command again to determine whether the permissions changed.

[ec2-user ~]$ sudo ls -l /mnt/tempvol/home/ec2-user/.ssh

If the user ID and group ID have changed, use the following command to restore them.

[ec2-user ~]$ sudo chown 222:500 /mnt/tempvol/home/ec2-user/.ssh/authorized_keys

Step 7: Unmount and detach the original volume from the

temporary instance, and reattach it to the original instance
1. From the temporary instance, unmount the volume that you attached so that you can reattach it to
the original instance. For example, use the following command to unmount the volume at /mnt/
tempvol.

[ec2-user ~]$ sudo umount /mnt/tempvol

2. Detach the volume from the temporary instance (you unmounted it in the previous step): From the
Amazon EC2 console, select the root device volume for the original instance (you made note of the
volume ID in a previous step), choose Actions, Detach Volume, and then choose Yes, Detach. Wait
for the state of the volume to become available. (You might need to choose the Refresh icon.)
3. Reattach the volume to the original instance: With the volume still selected, choose Actions, Attach
Volume. Select the instance ID of the original instance, specify the device name that you noted
earlier in Step 2 (p. 1054) for the original root device attachment (/dev/sda1 or /dev/xvda), and
then choose Attach.
Important
If you don't specify the same device name as the original attachment, you cannot start the
original instance. Amazon EC2 expects the root device volume at sda1 or /dev/xvda.

Step 8: Connect to the original instance using the new key pair
Select the original instance, choose Instance state, Start instance. After the instance enters the
running state, you can connect to it using the private key file for your new key pair.
Note
If the name of your new key pair and corresponding private key file is different from the name
of the original key pair, ensure that you specify the name of the new private key file when you
connect to your instance.

Step 9: Clean up
(Optional) You can terminate the temporary instance if you have no further use for it. Select the
temporary instance, and choose Instance state, Terminate instance.

Delete your key pair

When you delete a key pair, you are only deleting the Amazon EC2 copy of the public key. Deleting a
key pair doesn't affect the private key on your computer or the public key on any instances that already

1057
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Security groups

launched using that key pair. You can't launch a new instance using a deleted key pair, but you can
continue to connect to any instances that you launched using a deleted key pair, as long as you still have
the private key (.pem) file.

If you're using an Auto Scaling group (for example, in an Elastic Beanstalk environment), ensure that the
key pair you're deleting is not specified in your launch configuration. Amazon EC2 Auto Scaling launches
a replacement instance if it detects an unhealthy instance; however, the instance launch fails if the key
pair cannot be found.

You can delete a key pair using one of the following methods.

New console

To delete your key pair

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Key Pairs.
3. Select the key pair to delete and choose Delete.
4. In the confirmation field, enter Delete and then choose Delete.

Old console

To delete your key pair

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, under NETWORK & SECURITY, choose Key Pairs.
3. Select the key pair and choose Delete.
4. When prompted, choose Yes.

AWS CLI

To delete your key pair

Use the delete-key-pair AWS CLI command.

PowerShell

To delete your key pair

Use the Remove-EC2KeyPair AWS Tools for Windows PowerShell command.

If you create a Linux AMI from an instance, and then use the AMI to launch a new instance in a different
Region or account, the new instance includes the public key from the original instance. This enables
you to connect to the new instance using the same private key file as your original instance. You can
remove this public key from your instance by removing its entry from the .ssh/authorized_keys
file using a text editor of your choice. For more information about managing users on your instance
and providing remote access using a specific key pair, see Manage user accounts on your Amazon Linux
instance (p. 656).

Amazon EC2 security groups for Linux instances

A security group acts as a virtual firewall for your EC2 instances to control incoming and outgoing traffic.
Inbound rules control the incoming traffic to your instance, and outbound rules control the outgoing

1058
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Security groups

traffic from your instance. When you launch an instance, you can specify one or more security groups.
If you don't specify a security group, Amazon EC2 uses the default security group. You can add rules
to each security group that allow traffic to or from its associated instances. You can modify the rules
for a security group at any time. New and modified rules are automatically applied to all instances that
are associated with the security group. When Amazon EC2 decides whether to allow traffic to reach an
instance, it evaluates all of the rules from all of the security groups that are associated with the instance.

When you launch an instance in a VPC, you must specify a security group that's created for that VPC.
After you launch an instance, you can change its security groups. Security groups are associated with
network interfaces. Changing an instance's security groups changes the security groups associated
with the primary network interface (eth0). For more information, see Changing an instance's security
groups in the Amazon VPC User Guide. You can also change the security groups associated with any other
network interface. For more information, see Modify network interface attributes (p. 858).

Security is a shared responsibility between AWS and you. For more information, see Security in Amazon
EC2 (p. 972). AWS provides security groups as one of the tools for securing your instances, and you
need to configure them to meet your security needs. If you have requirements that aren't fully met by
security groups, you can maintain your own firewall on any of your instances in addition to using security
groups.

To allow traffic to a Windows instance, see Amazon EC2 security groups for Windows instances in the
Amazon EC2 User Guide for Windows Instances.

Contents
• Security group rules (p. 1060)
• Connection tracking (p. 1061)
• Throttling (p. 1062)
• Default and custom security groups (p. 1062)
• Default security groups (p. 1062)
• Custom security groups (p. 1063)
• Work with security groups (p. 1063)
• Create a security group (p. 1064)
• Copy a security group (p. 1065)
• View your security groups (p. 1065)
• Add rules to a security group (p. 1066)
• Update security group rules (p. 1068)
• Delete rules from a security group (p. 1069)
• Delete a security group (p. 1070)
• Assign a security group to an instance (p. 1071)
• Change an instance's security group (p. 1071)
• Security group rules for different use cases (p. 1072)
• Web server rules (p. 1072)
• Database server rules (p. 1073)
• Rules to connect to instances from your computer (p. 1074)
• Rules to connect to instances from an instance with the same security group (p. 1075)
• Rules for ping/ICMP (p. 1075)
• DNS server rules (p. 1075)
• Amazon EFS rules (p. 1076)
• Elastic Load Balancing rules (p. 1076)
• VPC peering rules (p. 1077) 1059
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Security group rules

Security group rules

The rules of a security group control the inbound traffic that's allowed to reach the instances that are
associated with the security group. The rules also control the outbound traffic that's allowed to leave
them.

The following are the characteristics of security group rules:

• By default, security groups allow all outbound traffic.

• Security group rules are always permissive; you can't create rules that deny access.
• Security group rules enable you to filter traffic based on protocols and port numbers.
• Security groups are stateful—if you send a request from your instance, the response traffic for that
request is allowed to flow in regardless of inbound security group rules. For VPC security groups, this
also means that responses to allowed inbound traffic are allowed to flow out, regardless of outbound
rules. For more information, see Connection tracking (p. 1061).
• You can add and remove rules at any time. Your changes are automatically applied to the instances
that are associated with the security group.

The effect of some rule changes can depend on how the traffic is tracked. For more information, see
Connection tracking (p. 1061).
• When you associate multiple security groups with an instance, the rules from each security group
are effectively aggregated to create one set of rules. Amazon EC2 uses this set of rules to determine
whether to allow access.

You can assign multiple security groups to an instance. Therefore, an instance can have hundreds of
rules that apply. This might cause problems when you access the instance. We recommend that you
condense your rules as much as possible.

For each rule, you specify the following:

• Name: The name for the security group (for example, my-security-group).

A name can be up to 255 characters in length. Allowed characters are a-z, A-Z, 0-9, spaces, and ._-:/
()#,@[]+=;{}!$*. When the name contains trailing spaces, we trim the spaces when we save the name.
For example, if you enter "Test Security Group " for the name, we store it as "Test Security Group".
• Protocol: The protocol to allow. The most common protocols are 6 (TCP), 17 (UDP), and 1 (ICMP).
• Port range: For TCP, UDP, or a custom protocol, the range of ports to allow. You can specify a single
port number (for example, 22), or range of port numbers (for example, 7000-8000).
• ICMP type and code: For ICMP, the ICMP type and code.
• Source or destination: The source (inbound rules) or destination (outbound rules) for the traffic.
Specify one of these options:
• An individual IPv4 address. You must use the /32 prefix length; for example, 203.0.113.1/32.
• An individual IPv6 address. You must use the /128 prefix length; for example,
2001:db8:1234:1a00::123/128.
• A range of IPv4 addresses, in CIDR block notation; for example, 203.0.113.0/24.
• A range of IPv6 addresses, in CIDR block notation; for example, 2001:db8:1234:1a00::/64.
• A prefix list ID, for example, pl-1234abc1234abc123. For more information, see Prefix lists in the
Amazon VPC User Guide.
• Another security group. This allows instances that are associated with the specified security group
to access instances associated with this security group. Choosing this option does not add rules
from the source security group to this security group. You can specify one of the following security
groups:
• The current security group

1060
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connection tracking

• A different security group for the same VPC

• A different security group for a peer VPC in a VPC peering connection
• (Optional) Description: You can add a description for the rule, which can help you identify it later. A
description can be up to 255 characters in length. Allowed characters are a-z, A-Z, 0-9, spaces, and ._-:/
()#,@[]+=;{}!$*.

When you specify a security group as the source or destination for a rule, the rule affects all instances
that are associated with the security group. Incoming traffic is allowed based on the private IP addresses
of the instances that are associated with the source security group (and not the public IP or Elastic IP
addresses). For more information about IP addresses, see Amazon EC2 instance IP addressing (p. 809). If
your security group rule references a security group in a peer VPC, and the referenced security group or
VPC peering connection is deleted, the rule is marked as stale. For more information, see Working with
Stale Security Group Rules in the Amazon VPC Peering Guide.

If there is more than one rule for a specific port, Amazon EC2 applies the most permissive rule. For
example, if you have a rule that allows access to TCP port 22 (SSH) from IP address 203.0.113.1, and
another rule that allows access to TCP port 22 from everyone, everyone has access to TCP port 22.

Connection tracking
Your security groups use connection tracking to track information about traffic to and from the instance.
Rules are applied based on the connection state of the traffic to determine if the traffic is allowed or
denied. This approach allows security groups to be stateful. This means that responses to inbound traffic
are allowed to flow out of the instance regardless of outbound security group rules, and vice versa. For
example, if you initiate an ICMP ping command to your instance from your home computer, and your
inbound security group rules allow ICMP traffic, information about the connection (including the port
information) is tracked. Response traffic from the instance for the ping command is not tracked as a
new request, but rather as an established connection and is allowed to flow out of the instance, even if
your outbound security group rules restrict outbound ICMP traffic.

Not all flows of traffic are tracked. If a security group rule permits TCP or UDP flows for all traffic
(0.0.0.0/0 or ::/0) and there is a corresponding rule in the other direction that permits all response
traffic (0.0.0.0/0 or ::/0) for all ports (0-65535), then that flow of traffic is not tracked. The response
traffic is therefore allowed to flow based on the inbound or outbound rule that permits the response
traffic, and not on tracking information.

In the following example, the security group has specific inbound rules for TCP and ICMP traffic, and
outbound rules that allow all outbound IPv4 and IPv6 traffic.

Inbound rules

Protocol type Port number Source IP

TCP 22 (SSH) 203.0.113.1/32

TCP 80 (HTTP) 0.0.0.0/0

TCP 80 (HTTP) ::/0

ICMP All 0.0.0.0/0

Outbound rules

Protocol type Port number Destination IP

All All 0.0.0.0/0

All All ::/0

1061
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Default and custom security groups

• TCP traffic on port 22 (SSH) to and from the instance is tracked, because the inbound rule allows traffic
from 203.0.113.1/32 only, and not all IP addresses (0.0.0.0/0).
• TCP traffic on port 80 (HTTP) to and from the instance is not tracked, because both the inbound and
outbound rules allow all traffic (0.0.0.0/0 or ::/0).
• ICMP traffic is always tracked, regardless of rules.
• If you remove the outbound rule from the security group, all traffic to and from the instance is tracked,
including traffic on port 80 (HTTP).

An untracked flow of traffic is immediately interrupted if the rule that enables the flow is removed or
modified. For example, if you have an open (0.0.0.0/0) outbound rule, and you remove a rule that allows
all (0.0.0.0/0) inbound SSH (TCP port 22) traffic to the instance (or modify it such that the connection
would no longer be permitted), your existing SSH connections to the instance are immediately dropped.
The connection was not previously being tracked, so the change will break the connection. On the other
hand, if you have a narrower inbound rule that initially allows the SSH connection (meaning that the
connection was tracked), but change that rule to no longer allow new connections from the address of
the current SSH client, the existing connection will not be broken by changing the rule.

For protocols other than TCP, UDP, or ICMP, only the IP address and protocol number is tracked. If your
instance sends traffic to another host (host B), and host B initiates the same type of traffic to your
instance in a separate request within 600 seconds of the original request or response, your instance
accepts it regardless of inbound security group rules. Your instance accepts it because it’s regarded as
response traffic.

To ensure that traffic is immediately interrupted when you remove a security group rule, or to ensure
that all inbound traffic is subject to firewall rules, you can use a network ACL for your subnet. Network
ACLs are stateless and therefore do not automatically allow response traffic. For more information, see
Network ACLs in the Amazon VPC User Guide.

Throttling
There is a maximum number of connections that can be tracked per instance. After the maximum is
reached, any packets that are sent or received are lost because a new connection cannot be established.
When this happens, applications that send and receive packets cannot communicate properly. For more
information, see Monitor network performance for your instance (p. 884).

The conntrack_allowance_exceeded metric shows the number of lost packets. For more
information, see View the network performance metrics for your Linux instance (p. 885).

Default and custom security groups

Your AWS account automatically has a default security group for the default VPC in each Region. If you
don't specify a security group when you launch an instance, the instance is automatically associated with
the default security group for the VPC. If you don't want your instances to use the default security group,
you can create your own custom security groups and specify them when you launch your instances.

Topics
• Default security groups (p. 1062)
• Custom security groups (p. 1063)

Default security groups

Your AWS account automatically has a default security group for the default VPC in each Region. If you
don't specify a security group when you launch an instance, the instance is automatically associated with
the default security group for the VPC.

1062
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with security groups

A default security group is named default, and it has an ID assigned by AWS. The following are the
default rules for each default security group:

• Allows all inbound traffic from other instances associated with the default security group. The security
group specifies itself as a source security group in its inbound rules.
• Allows all outbound traffic from the instance.

You can add or remove inbound and outbound rules for any default security group.

You can't delete a default security group. If you try to delete a default security group, you see the
following error: Client.CannotDelete: the specified group: "sg-51530134" name:
"default" cannot be deleted by a user.

Custom security groups

If you don't want your instances to use the default security group, you can create your own security
groups and specify them when you launch your instances. You can create multiple security groups to
reflect the different roles that your instances play; for example, a web server or a database server.

When you create a security group, you must provide it with a name and a description. Security group
names and descriptions can be up to 255 characters in length, and are limited to the following
characters:

a-z, A-Z, 0-9, spaces, and ._-:/()#,@[]+=&;{}!$*

A security group name cannot start with sg-. A security group name must be unique for the VPC.

The following are the default rules for a security group that you create:

• Allows no inbound traffic

• Allows all outbound traffic

After you've created a security group, you can change its inbound rules to reflect the type of inbound
traffic that you want to reach the associated instances. You can also change its outbound rules.

For more information about the rules you can add to a security group, see Security group rules for
different use cases (p. 1072).

Work with security groups

You can assign a security group to an instance when you launch the instance. When you add or remove
rules, those changes are automatically applied to all instances to which you've assigned the security
group. For more information, see Assign a security group to an instance (p. 1071).

After you launch an instance, you can change its security groups. For more information, see Change an
instance's security group (p. 1071).

You can create, view, update, and delete security groups and security group rules using the Amazon EC2
console and the command line tools.

Tasks
• Create a security group (p. 1064)
• Copy a security group (p. 1065)
• View your security groups (p. 1065)

1063
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with security groups

• Add rules to a security group (p. 1066)

• Update security group rules (p. 1068)
• Delete rules from a security group (p. 1069)
• Delete a security group (p. 1070)
• Assign a security group to an instance (p. 1071)
• Change an instance's security group (p. 1071)

Create a security group

You can create a custom security group using one of the following methods. You must specify the VPC
for which you're creating the security group.

New console

To create a security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Choose Create security group.
4. In the Basic details section, do the following.

a. Enter a descriptive name and brief description for the security group. They can't be edited
after the security group is created. The name and description can be up to 255 characters
long. The valid characters are a-z, A-Z, 0-9, spaces, and ._-:/()#,@[]+=&;{}!
$*.
b. For VPC, choose the VPC in which to create the security group. The security group can only
be used in the VPC in which it is created.
5. You can add security group rules now, or you can add them at any time after you have created
the security group. For more information about adding security group rules, see Add rules to a
security group (p. 1066).
6. Choose Create.

Old console

To create a security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Choose Create Security Group.
4. Specify a name and description for the security group.
5. For VPC, choose the ID of the VPC.
6. You can start adding rules, or you can choose Create to create the security group now (you can
always add rules later). For more information about adding rules, see Add rules to a security
group (p. 1066).

Command line

To create a security group

Use one of the following commands:

1064
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with security groups

• create-security-group (AWS CLI)

• New-EC2SecurityGroup (AWS Tools for Windows PowerShell)

Copy a security group

You can create a new security group by creating a copy of an existing one. When you copy a security
group, the copy is created with the same inbound and outbound rules as the original security group. If
the original security group is in a VPC, the copy is created in the same VPC unless you specify a different
one.

The copy receives a new unique security group ID and you must give it a name. You can also add a
description.

You can't copy a security group from one Region to another Region.

You can create a copy of a security group using one of the following methods.

New console

To copy a security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Select the security group to copy and choose Actions, Copy to new security group.
4. Specify a name and optional description, and change the VPC and security group rules if
needed.
5. Choose Create.

Old console

To copy a security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Select the security group you want to copy, choose Actions, Copy to new.
4. The Create Security Group dialog opens, and is populated with the rules from the existing
security group. Specify a name and description for your new security group. For VPC, choose the
ID of the VPC. When you are done, choose Create.

View your security groups

You can view information about your security groups using one of the following methods.

New console

To view your security groups

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Your security groups are listed. To view the details for a specific security group, including its
inbound and outbound rules, choose its ID in the Security group ID column.

1065
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with security groups

Old console

To view your security groups

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. (Optional) Select VPC ID from the filter list, then choose the ID of the VPC.
4. Select a security group. General information is displayed on the Description tab, inbound rules
on the Inbound tab, outbound rules on the Outbound tab, and tags on the Tags tab.

Command line

To view your security groups

Use one of the following commands.

• describe-security-groups (AWS CLI)

• Get-EC2SecurityGroup (AWS Tools for Windows PowerShell)

Add rules to a security group

When you add a rule to a security group, the new rule is automatically applied to any instances that are
associated with the security group. There might be a short delay before the rule is applied. For more
information about choosing security group rules for specific types of access, see Security group rules for
different use cases (p. 1072). For security group rule quotas, see Amazon VPC quotas in the Amazon VPC
User Guide.

You can add rules to a security group using one of the following methods.

New console

To add an inbound rule to a security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. In the list, select the security group and choose Actions, Edit inbound rules.
4. Choose Add rule and do the following.

a. For Type, choose the type of protocol to allow.

• If you choose a custom TCP or UDP protocol, you must manually enter the port range to
allow.
• If you choose a custom ICMP protocol, you must choose the ICMP type name from
Protocol, and, if applicable, the code name from Port range.
• If you choose any other type, the protocol and port range are configured automatically.
b. For Source, do one of the following.

• Choose Custom and then enter an IP address in CIDR notation, a CIDR block, another
security group, or a prefix list from which to allow inbound traffic.
• Choose Anywhere to allow all inbound traffic of the specified protocol to reach your
instance. This option automatically adds the 0.0.0.0/0 IPv4 CIDR block as an allowed
source. This is acceptable for a short time in a test environment, but it's unsafe for
production environments. In production, authorize only a specific IP address or range of
addresses to access your instance.

1066
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with security groups

If your security group is in a VPC that's enabled for IPv6, this option automatically adds a
second rule for IPv6 traffic (::/0).
• Choose My IP to allow inbound traffic from only your local computer's public IPv4
address.
c. For Description, optionally specify a brief description for the rule.
5. Choose Preview changes, Save rules.

To add an outbound rule to a security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. In the list, select the security group and choose Actions, Edit outbound rules.
4. Choose Add rule and do the following.

a. For Type, choose the type of protocol to allow.

• If you choose a custom TCP or UDP protocol, you must manually enter the port range to
allow.
• If you choose a custom ICMP protocol, you must choose the ICMP type name from
Protocol, and, if applicable, the code name from Port range.
• If you choose any other type, the protocol and port range are configured automatically.
b. For Destination, do one of the following.

• Choose Custom and then enter an IP address in CIDR notation, a CIDR block, another
security group, or a prefix list for which to allow outbound traffic.
• Choose Anywhere to allow outbound traffic to all IP addresses. This option automatically
adds the 0.0.0.0/0 IPv4 CIDR block as an allowed source.

If your security group is in a VPC that's enabled for IPv6, this option automatically adds a
second rule for IPv6 traffic (::/0).
• Choose My IP to allow outbound traffic only to your local computer's public IPv4 address.
c. For Description, optionally specify a brief description for the rule.
5. Choose Preview changes, Confirm.

Old console

To add rules to a security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups and select the security group.
3. On the Inbound tab, choose Edit.
4. In the dialog, choose Add Rule and do the following:

• For Type, select the protocol.

• If you select a custom TCP or UDP protocol, specify the port range in Port Range.
• If you select a custom ICMP protocol, choose the ICMP type name from Protocol, and, if
applicable, the code name from Port Range.
• For Source, choose one of the following:
• Custom: in the provided field, you must specify an IP address in CIDR notation, a CIDR
block, or another security group.

1067
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with security groups

• Anywhere: automatically adds the 0.0.0.0/0 IPv4 CIDR block. This option enables all
traffic of the specified type to reach your instance. This is acceptable for a short time in a
test environment, but it's unsafe for production environments. In production, authorize only
a specific IP address or range of addresses to access your instance.

If your security group is in a VPC that's enabled for IPv6, the Anywhere option creates two
rules—one for IPv4 traffic (0.0.0.0/0) and one for IPv6 traffic (::/0).
• My IP: automatically adds the public IPv4 address of your local computer.
• For Description, you can optionally specify a description for the rule.

For more information about the types of rules that you can add, see Security group rules for
different use cases (p. 1072).
5. Choose Save.
6. You can also specify outbound rules. On the Outbound tab, choose Edit, Add Rule, and do the
following:

• For Type, select the protocol.

• If you select a custom TCP or UDP protocol, specify the port range in Port Range.
• If you select a custom ICMP protocol, choose the ICMP type name from Protocol, and, if
applicable, the code name from Port Range.
• For Destination, choose one of the following:
• Custom: in the provided field, you must specify an IP address in CIDR notation, a CIDR
block, or another security group.
• Anywhere: automatically adds the 0.0.0.0/0 IPv4 CIDR block. This option enables
outbound traffic to all IP addresses.

If your security group is in a VPC that's enabled for IPv6, the Anywhere option creates two
rules—one for IPv4 traffic (0.0.0.0/0) and one for IPv6 traffic (::/0).
• My IP: automatically adds the IP address of your local computer.
• For Description, you can optionally specify a description for the rule.
7. Choose Save.

Command line

To add rules to a security group

Use one of the following commands.

• authorize-security-group-ingress (AWS CLI)

• Grant-EC2SecurityGroupIngress (AWS Tools for Windows PowerShell)

To add one or more egress rules to a security group

Use one of the following commands.

• authorize-security-group-egress (AWS CLI)

• Grant-EC2SecurityGroupEgress (AWS Tools for Windows PowerShell)

Update security group rules

You can update a security group rule using one of the following methods. The updated rule is
automatically applied to any instances that are associated with the security group.

1068
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with security groups

New console

When you modify the protocol, port range, or source or destination of an existing security group rule
using the console, the console deletes the existing rule and adds a new one for you.

To update a security group rule

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Select the security group to update, choose Actions, and then choose Edit inbound rules to
update a rule for inbound traffic or Edit outbound rules to update a rule for outbound traffic.
4. Update the rule as required and then choose Preview changes, Confirm.

Old console

When you modify the protocol, port range, or source or destination of an existing security group rule
using the console, the console deletes the existing rule and adds a new one for you.

To update a security group rule

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Select the security group to update, and choose the Inbound tab to update a rule for inbound
traffic or the Outbound tab to update a rule for outbound traffic.
4. Choose Edit.
5. Modify the rule entry as required and choose Save.

Command line

You cannot modify the protocol, port range, or source or destination of an existing rule using the
Amazon EC2 API or a command line tools. Instead, you must delete the existing rule and add a new
rule. You can, however, update the description of an existing rule.

To update the description for an existing inbound rule

Use one of the following commands.

• update-security-group-rule-descriptions-ingress (AWS CLI)

• Update-EC2SecurityGroupRuleIngressDescription (AWS Tools for Windows PowerShell)

To update the description for an existing outbound rule

Use one of the following commands.

• update-security-group-rule-descriptions-egress (AWS CLI)

• Update-EC2SecurityGroupRuleEgressDescription (AWS Tools for Windows PowerShell)

Delete rules from a security group

When you delete a rule from a security group, the change is automatically applied to any instances
associated with the security group.

You can delete rules from a security group using one of the following methods.

1069
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with security groups

New console

To delete a security group rule

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Select the security group to update, choose Actions, and then choose Edit inbound rules to
remove an inbound rule or Edit outbound rules to remove an outbound rule.
4. Choose the Delete button to the right of the rule to delete.
5. Choose Preview changes, Confirm.

Old console

To delete a security group rule

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Select a security group.
4. On the Inbound tab (for inbound rules) or Outbound tab (for outbound rules), choose Edit.
Choose Delete (a cross icon) next to each rule to delete.
5. Choose Save.

Command line

To remove one or more ingress rules from a security group

Use one of the following commands.

• revoke-security-group-ingress (AWS CLI)

• Revoke-EC2SecurityGroupIngress (AWS Tools for Windows PowerShell)

To remove one or more egress rules from a security group

Use one of the following commands.

• revoke-security-group-egress (AWS CLI)

• Revoke-EC2SecurityGroupEgress (AWS Tools for Windows PowerShell)

Delete a security group

You can't delete a security group that is associated with an instance. You can't delete the default security
group. You can't delete a security group that is referenced by a rule in another security group in the same
VPC. If your security group is referenced by one of its own rules, you must delete the rule before you can
delete the security group.

New console

To delete a security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Select the security group to delete and choose Actions, Delete security group, Delete.

1070
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with security groups

Old console

To delete a security group

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Security Groups.
3. Select a security group and choose Actions, Delete Security Group.
4. Choose Yes, Delete.

Command line

To delete a security group

Use one of the following commands.

• delete-security-group (AWS CLI)

• Remove-EC2SecurityGroup (AWS Tools for Windows PowerShell)

Assign a security group to an instance

You can assign one or more security groups to an instance when you launch the instance. You can also
specify one or more security groups in a launch template. The security groups will be assigned to all
instances that are launched using the launch template.

• To assign a security group to an instance when you launch the instance, see Step 6: Configure Security
Group (p. 531).
• To specify a security group in a launch template, see Step 6 of Create a new launch template using
parameters you define (p. 534).

Change an instance's security group

After you launch an instance, you can change its security groups by adding or removing security groups.
You can change the security groups when the instance is in the running or stopped state.

New console

To change the security groups for an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select your instance, and then choose Actions, Security, Change security groups.
4. For Associated security groups, select a security group from the list and choose Add security
group.

To remove an already associated security group, choose Remove for that security group.
5. Choose Save.

Old console

To change the security groups for an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.

1071
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Security group rules for different use cases

3. Select your instance, and then choose Actions, Networking, Change Security Groups.
4. To add one or more security groups, select its check box.

To remove an already associated security group, clear its check box.

5. Choose Assign Security Groups.

Command line

To change the security groups for an instance using the command line

Use one of the following commands.

• modify-instance-attribute (AWS CLI)

• Edit-EC2InstanceAttribute (AWS Tools for Windows PowerShell)

Security group rules for different use cases

You can create a security group and add rules that reflect the role of the instance that's associated with
the security group. For example, an instance that's configured as a web server needs security group rules
that allow inbound HTTP and HTTPS access. Likewise, a database instance needs rules that allow access
for the type of database, such as access over port 3306 for MySQL.

The following are examples of the kinds of rules that you can add to security groups for specific kinds of
access.

Examples
• Web server rules (p. 1072)
• Database server rules (p. 1073)
• Rules to connect to instances from your computer (p. 1074)
• Rules to connect to instances from an instance with the same security group (p. 1075)
• Rules for ping/ICMP (p. 1075)
• DNS server rules (p. 1075)
• Amazon EFS rules (p. 1076)
• Elastic Load Balancing rules (p. 1076)
• VPC peering rules (p. 1077)

Web server rules

The following inbound rules allow HTTP and HTTPS access from any IP address. If your VPC is enabled
for IPv6, you can add rules to control inbound HTTP and HTTPS traffic from IPv6 addresses.

Protocol type Protocol number Port Source IP Notes

TCP 6 80 (HTTP) 0.0.0.0/0 Allows inbound

HTTP access from
any IPv4 address

TCP 6 443 (HTTPS) 0.0.0.0/0 Allows inbound

HTTPS access
from any IPv4
address

1072
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Security group rules for different use cases

Protocol type Protocol number Port Source IP Notes

TCP 6 80 (HTTP) ::/0 Allows inbound

HTTP access from
any IPv6 address

TCP 6 443 (HTTPS) ::/0 Allows inbound

HTTPS access
from any IPv6
address

Database server rules

The following inbound rules are examples of rules you might add for database access, depending on
what type of database you're running on your instance. For more information about Amazon RDS
instances, see the Amazon RDS User Guide.

For the source IP, specify one of the following:

• A specific IP address or range of IP addresses (in CIDR block notation) in your local network
• A security group ID for a group of instances that access the database

Protocol type Protocol number Port Notes

TCP 6 1433 (MS SQL) The default port to

access a Microsoft SQL
Server database, for
example, on an Amazon
RDS instance

TCP 6 3306 (MYSQL/Aurora) The default port to

access a MySQL or
Aurora database, for
example, on an Amazon
RDS instance

TCP 6 5439 (Redshift) The default port to

access an Amazon
Redshift cluster
database.

TCP 6 5432 (PostgreSQL) The default port to

access a PostgreSQL
database, for example,
on an Amazon RDS
instance

TCP 6 1521 (Oracle) The default port

to access an Oracle
database, for example,
on an Amazon RDS
instance

1073
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Security group rules for different use cases

You can optionally restrict outbound traffic from your database servers. For example, you might want
to allow access to the internet for software updates, but restrict all other kinds of traffic. You must first
remove the default outbound rule that allows all outbound traffic.

Protocol type Protocol number Port Destination IP Notes

TCP 6 80 (HTTP) 0.0.0.0/0 Allows outbound

HTTP access to
any IPv4 address

TCP 6 443 (HTTPS) 0.0.0.0/0 Allows outbound

HTTPS access to
any IPv4 address

TCP 6 80 (HTTP) ::/0 (IPv6-enabled

VPC only) Allows
outbound HTTP
access to any IPv6
address

TCP 6 443 (HTTPS) ::/0 (IPv6-enabled

VPC only) Allows
outbound HTTPS
access to any IPv6
address

Rules to connect to instances from your computer

To connect to your instance, your security group must have inbound rules that allow SSH access (for
Linux instances) or RDP access (for Windows instances).

Protocol type Protocol number Port Source IP

TCP 6 22 (SSH) The public IPv4 address

of your computer, or a
range of IP addresses
(in CIDR block notation)
in your local network. If
your VPC is enabled for
IPv6 and your instance
has an IPv6 address,
you can enter an IPv6
address or range.

TCP 6 3389 (RDP) The public IPv4 address

of your computer, or a
range of IP addresses
(in CIDR block notation)
in your local network. If
your VPC is enabled for
IPv6 and your instance
has an IPv6 address,
you can enter an IPv6
address or range.

1074
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Security group rules for different use cases

Rules to connect to instances from an instance with the same

security group
To allow instances that are associated with the same security group to communicate with each other, you
must explicitly add rules for this.

The following table describes the inbound rule for a security group that enables associated instances to
communicate with each other. The rule allows all types of traffic.

Protocol type Protocol number Ports Source IP

-1 (All) -1 (All) -1 (All) The ID of the security

group

Rules for ping/ICMP

The ping command is a type of ICMP traffic. To ping your instance, you must add the following inbound
ICMP rule.

Protocol type Protocol number ICMP type ICMP code Source IP

ICMP 1 8 (Echo) N/A The public IPv4

address of your
computer, or a
range of IPv4
addresses (in CIDR
block notation) in
your local network

To use the ping6 command to ping the IPv6 address for your instance, you must add the following
inbound ICMPv6 rule.

Protocol type Protocol number ICMP type ICMP code Source IP

ICMPv6 58 128 (Echo) 0 The IPv6 address

of your computer,
or a range of IPv6
addresses (in CIDR
block notation) in
your local network

DNS server rules

If you've set up your EC2 instance as a DNS server, you must ensure that TCP and UDP traffic can reach
your DNS server over port 53.

For the source IP, specify one of the following:

• An IP address or range of IP addresses (in CIDR block notation) in a network

1075
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Security group rules for different use cases

• The ID of a security group for the set of instances in your network that require access to the DNS
server

Protocol type Protocol number Port

TCP 6 53

UDP 17 53

Amazon EFS rules

If you're using an Amazon EFS file system with your Amazon EC2 instances, the security group that you
associate with your Amazon EFS mount targets must allow traffic over the NFS protocol.

Protocol type Protocol number Ports Source IP Notes

TCP 6 2049 (NFS) The ID of the Allows inbound

security group. NFS access
from resources
(including the
mount target)
associated with
this security
group.

To mount an Amazon EFS file system on your Amazon EC2 instance, you must connect to your instance.
Therefore, the security group associated with your instance must have rules that allow inbound SSH from
your local computer or local network.

Protocol type Protocol number Ports Source IP Notes

TCP 6 22 (SSH) The IP address Allows inbound

range of your SSH access
local computer, from your local
or the range of IP computer.
addresses (in CIDR
block notation) for
your network.

Elastic Load Balancing rules

If you're using a load balancer, the security group associated with your load balancer must have rules
that allow communication with your instances or targets.

Inbound

Protocol type Protocol number Port Source IP Notes

TCP 6 The listener port For an Internet- Allow inbound

facing load- traffic on the load

1076
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Update management

balancer: 0.0.0.0/0 balancer listener

(all IPv4 addresses) port.

For an internal
load-balancer: the
IPv4 CIDR block of
the VPC

Outbound

Protocol type Protocol number Port Destination IP Notes

TCP 6 The instance The ID of the Allow outbound

listener port instance security traffic to instances
group on the instance
listener port.

TCP 6 The health check The ID of the Allow outbound

port instance security traffic to instances
group on the health
check port.

The security group rules for your instances must allow the load balancer to communicate with your
instances on both the listener port and the health check port.

Inbound

Protocol type Protocol number Port Source IP Notes

TCP 6 The instance The ID of the load Allow traffic from

listener port balancer security the load balancer
group on the instance
listener port.

TCP 6 The health check The ID of the load Allow traffic from
port balancer security the load balancer
group on the health
check port.

For more information, see Configure security groups for your Classic Load Balancer in the User Guide
for Classic Load Balancers, and Security groups for your Application Load Balancer in the User Guide for
Application Load Balancers.

VPC peering rules

You can update the inbound or outbound rules for your VPC security groups to reference security groups
in the peered VPC. Doing so allows traffic to flow to and from instances that are associated with the
referenced security group in the peered VPC. For more information about how to configure security
groups for VPC peering, see Updating your security groups to reference peer VPC groups.

Update management in Amazon EC2

We recommend that you regularly patch, update, and secure the operating system and applications
on your EC2 instances. You can use AWS Systems Manager Patch Manager to automate the process of

1077
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Compliance validation

installing security-related updates for both the operating system and applications. Alternatively, you can
use any automatic update services or recommended processes for installing updates that are provided by
the application vendor.

Compliance validation for Amazon EC2

Third-party auditors assess the security and compliance of Amazon EC2 as part of multiple AWS
compliance programs. These include SOC, PCI, FedRAMP, HIPAA, and others.

For a list of AWS services in scope of specific compliance programs, see AWS Services in Scope by
Compliance Program. For general information, see AWS Compliance Programs.

You can download third-party audit reports using AWS Artifact. For more information, see Downloading
Reports in AWS Artifact.

Your compliance responsibility when using Amazon EC2 is determined by the sensitivity of your data,
your company's compliance objectives, and applicable laws and regulations. AWS provides the following
resources to help with compliance:

• Security and Compliance Quick Start Guides – These deployment guides discuss architectural
considerations and provide steps for deploying security- and compliance-focused baseline
environments on AWS.
• Architecting for HIPAA Security and Compliance Whitepaper – This whitepaper describes how
companies can use AWS to create HIPAA-compliant applications.
• AWS Compliance Resources – This collection of workbooks and guides might apply to your industry
and location.
• Evaluating Resources with Rules in the AWS Config Developer Guide – AWS Config; assesses how well
your resource configurations comply with internal practices, industry guidelines, and regulations.
• AWS Security Hub – This AWS service provides a comprehensive view of your security state within AWS
that helps you check your compliance with security industry standards and best practices.

1078
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Storage
Amazon EC2 provides you with flexible, cost effective, and easy-to-use data storage options for your
instances. Each option has a unique combination of performance and durability. These storage options
can be used independently or in combination to suit your requirements.

After reading this section, you should have a good understanding about how you can use the data
storage options supported by Amazon EC2 to meet your specific requirements. These storage options
include the following:

• Amazon Elastic Block Store (p. 1080)

• Amazon EC2 instance store (p. 1268)
• Use Amazon EFS with Amazon EC2 (p. 1287)
• Use Amazon S3 with Amazon EC2 (p. 1285)

The following figure shows the relationship between these storage options and your instance.

Amazon EBS

Amazon EBS provides durable, block-level storage volumes that you can attach to a running instance.
You can use Amazon EBS as a primary storage device for data that requires frequent and granular
updates. For example, Amazon EBS is the recommended storage option when you run a database on an
instance.

An EBS volume behaves like a raw, unformatted, external block device that you can attach to a single
instance. The volume persists independently from the running life of an instance. After an EBS volume
is attached to an instance, you can use it like any other physical hard drive. As illustrated in the previous

1079
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Amazon EBS

figure, multiple volumes can be attached to an instance. You can also detach an EBS volume from one
instance and attach it to another instance. You can dynamically change the configuration of a volume
attached to an instance. EBS volumes can also be created as encrypted volumes using the Amazon EBS
encryption feature. For more information, see Amazon EBS encryption (p. 1176).

To keep a backup copy of your data, you can create a snapshot of an EBS volume, which is stored in
Amazon S3. You can create an EBS volume from a snapshot, and attach it to another instance. For more
information, see Amazon Elastic Block Store (p. 1080).

Amazon EC2 instance store

Many instances can access storage from disks that are physically attached to the host computer. This
disk storage is referred to as instance store. Instance store provides temporary block-level storage for
instances. The data on an instance store volume persists only during the life of the associated instance;
if you stop, hibernate, or terminate an instance, any data on instance store volumes is lost. For more
information, see Amazon EC2 instance store (p. 1268).

Amazon EFS file system

Amazon EFS provides scalable file storage for use with Amazon EC2. You can create an EFS file system
and configure your instances to mount the file system. You can use an EFS file system as a common data
source for workloads and applications running on multiple instances. For more information, see Use
Amazon EFS with Amazon EC2 (p. 1287).

Amazon S3

Amazon S3 provides access to reliable and inexpensive data storage infrastructure. It is designed to
make web-scale computing easier by enabling you to store and retrieve any amount of data, at any time,
from within Amazon EC2 or anywhere on the web. For example, you can use Amazon S3 to store backup
copies of your data and applications. Amazon EC2 uses Amazon S3 to store EBS snapshots and instance
store-backed AMIs. For more information, see Use Amazon S3 with Amazon EC2 (p. 1285).

Adding storage

Every time you launch an instance from an AMI, a root storage device is created for that instance. The
root storage device contains all the information necessary to boot the instance. You can specify storage
volumes in addition to the root device volume when you create an AMI or launch an instance using block
device mapping. For more information, see Block device mapping (p. 1294).

You can also attach EBS volumes to a running instance. For more information, see Attach an Amazon EBS
volume to an instance (p. 1106).

Storage pricing

For information about storage pricing, open AWS Pricing, scroll down to Services Pricing, choose
Storage, and then choose the storage option to open that storage option's pricing page. For information
about estimating the cost of storage, see the AWS Pricing Calculator.

Amazon Elastic Block Store (Amazon EBS)

Amazon Elastic Block Store (Amazon EBS) provides block level storage volumes for use with EC2
instances. EBS volumes behave like raw, unformatted block devices. You can mount these volumes
as devices on your instances. EBS volumes that are attached to an instance are exposed as storage
volumes that persist independently from the life of the instance. You can create a file system on top
of these volumes, or use them in any way you would use a block device (such as a hard drive). You can
dynamically change the configuration of a volume attached to an instance.

1080
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Features of Amazon EBS

We recommend Amazon EBS for data that must be quickly accessible and requires long-term persistence.
EBS volumes are particularly well-suited for use as the primary storage for file systems, databases, or for
any applications that require fine granular updates and access to raw, unformatted, block-level storage.
Amazon EBS is well suited to both database-style applications that rely on random reads and writes, and
to throughput-intensive applications that perform long, continuous reads and writes.

With Amazon EBS, you pay only for what you use. For more information about Amazon EBS pricing, see
the Projecting Costs section of the Amazon Elastic Block Store page.

Contents
• Features of Amazon EBS (p. 1081)
• Amazon EBS volumes (p. 1082)
• Amazon EBS snapshots (p. 1126)
• Amazon EBS data services (p. 1163)
• Amazon EBS and NVMe on Linux instances (p. 1214)
• Amazon EBS–optimized instances (p. 1217)
• Amazon EBS volume performance on Linux instances (p. 1237)
• Amazon CloudWatch metrics for Amazon EBS (p. 1252)
• Amazon CloudWatch Events for Amazon EBS (p. 1258)
• Amazon EBS quotas (p. 1268)

Features of Amazon EBS

• EBS volumes are created in a specific Availability Zone, and can then be attached to any instances
in that same Availability Zone. To make a volume available outside of the Availability Zone, you can
create a snapshot and restore that snapshot to a new volume anywhere in that Region. You can copy
snapshots to other Regions and then restore them to new volumes there, making it easier to leverage
multiple AWS Regions for geographical expansion, data center migration, and disaster recovery.
• Amazon EBS provides the following volume types: General Purpose SSD, Provisioned IOPS SSD,
Throughput Optimized HDD, and Cold HDD. For more information, see EBS volume types (p. 1084).

The following is a summary of performance and use cases for each volume type.
• General Purpose SSD volumes (gp2 and gp3) balance price and performance for a wide variety of
transactional workloads. These volumes are ideal for a use cases such as boot volumes, medium-size
single instance databases, and development and test environments.
• Provisioned IOPS SSD volumes (io1 and io2) support up to 64,000 IOPS and 1,000 MiB/s of
throughput. This enables you to predictably scale to tens of thousands of IOPS per EC2 instance.
• Throughput Optimized HDD volumes (st1) provide low-cost magnetic storage that defines
performance in terms of throughput rather than IOPS. These volumes are ideal for large, sequential
workloads such as Amazon EMR, ETL, data warehouses, and log processing.
• Cold HDD volumes (sc1) provide low-cost magnetic storage that defines performance in terms of
throughput rather than IOPS. These volumes are ideal for large, sequential, cold-data workloads.
If you require infrequent access to your data and are looking to save costs, these volumes provides
inexpensive block storage.
• You can create your EBS volumes as encrypted volumes, in order to meet a wide range of data-at-rest
encryption requirements for regulated/audited data and applications. When you create an encrypted
EBS volume and attach it to a supported instance type, data stored at rest on the volume, disk I/O, and
snapshots created from the volume are all encrypted. The encryption occurs on the servers that host
EC2 instances, providing encryption of data-in-transit from EC2 instances to EBS storage. For more
information, see Amazon EBS encryption (p. 1176).
• You can create point-in-time snapshots of EBS volumes, which are persisted to Amazon S3. Snapshots
protect data for long-term durability, and they can be used as the starting point for new EBS volumes.

1081
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

The same snapshot can be used to instantiate as many volumes as you wish. These snapshots can be
copied across AWS Regions. For more information, see Amazon EBS snapshots (p. 1126).
• Performance metrics, such as bandwidth, throughput, latency, and average queue length, are available
through the AWS Management Console. These metrics, provided by Amazon CloudWatch, allow you to
monitor the performance of your volumes to make sure that you are providing enough performance
for your applications without paying for resources you don't need. For more information, see Amazon
EBS volume performance on Linux instances (p. 1237).

Amazon EBS volumes

An Amazon EBS volume is a durable, block-level storage device that you can attach to your instances.
After you attach a volume to an instance, you can use it as you would use a physical hard drive. EBS
volumes are flexible. For current-generation volumes attached to current-generation instance types, you
can dynamically increase size, modify the provisioned IOPS capacity, and change volume type on live
production volumes.

You can use EBS volumes as primary storage for data that requires frequent updates, such as the system
drive for an instance or storage for a database application. You can also use them for throughput-
intensive applications that perform continuous disk scans. EBS volumes persist independently from the
running life of an EC2 instance.

You can attach multiple EBS volumes to a single instance. The volume and instance must be in the same
Availability Zone. Depending on the volume and instance types, you can use Multi-Attach (p. 1107) to
mount a volume to multiple instances at the same time.

Amazon EBS provides the following volume types: General Purpose SSD (gp2 and gp3), Provisioned IOPS
SSD (io1 and io2), Throughput Optimized HDD (st1), Cold HDD (sc1), and Magnetic (standard). They
differ in performance characteristics and price, allowing you to tailor your storage performance and cost
to the needs of your applications. For more information, see Amazon EBS volume types (p. 1084).

Your account has a limit on the number of EBS volumes that you can use, and the total storage available
to you. For more information about these limits, and how to request an increase in your limits, see
Amazon EC2 service quotas (p. 1323).

For more information about pricing, see Amazon EBS Pricing.

Contents
• Benefits of using EBS volumes (p. 1082)
• Amazon EBS volume types (p. 1084)
• Constraints on the size and configuration of an EBS volume (p. 1101)
• Create an Amazon EBS volume (p. 1104)
• Attach an Amazon EBS volume to an instance (p. 1106)
• Attach a volume to multiple instances with Amazon EBS Multi-Attach (p. 1107)
• Make an Amazon EBS volume available for use on Linux (p. 1111)
• View information about an Amazon EBS volume (p. 1114)
• Replace an Amazon EBS volume using a previous snapshot (p. 1116)
• Monitor the status of your volumes (p. 1116)
• Detach an Amazon EBS volume from a Linux instance (p. 1123)
• Delete an Amazon EBS volume (p. 1126)

Benefits of using EBS volumes

EBS volumes provide benefits that are not provided by instance store volumes.

1082
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Data availability
When you create an EBS volume, it is automatically replicated within its Availability Zone to prevent data
loss due to failure of any single hardware component. You can attach an EBS volume to any EC2 instance
in the same Availability Zone. After you attach a volume, it appears as a native block device similar to
a hard drive or other physical device. At that point, the instance can interact with the volume just as it
would with a local drive. You can connect to the instance and format the EBS volume with a file system,
such as ext3, and then install applications.

If you attach multiple volumes to a device that you have named, you can stripe data across the volumes
for increased I/O and throughput performance.

You can attach io1 and io2 EBS volumes to up to 16 Nitro-based instances. For more information, see
Attach a volume to multiple instances with Amazon EBS Multi-Attach (p. 1107). Otherwise, you can
attach an EBS volume to a single instance.

You can get monitoring data for your EBS volumes, including root device volumes for EBS-backed
instances, at no additional charge. For more information about monitoring metrics, see Amazon
CloudWatch metrics for Amazon EBS (p. 1252). For information about tracking the status of your
volumes, see Amazon CloudWatch Events for Amazon EBS (p. 1258).

Data persistence
An EBS volume is off-instance storage that can persist independently from the life of an instance. You
continue to pay for the volume usage as long as the data persists.

EBS volumes that are attached to a running instance can automatically detach from the instance with
their data intact when the instance is terminated if you uncheck the Delete on Termination check
box when you configure EBS volumes for your instance on the EC2 console. The volume can then be
reattached to a new instance, enabling quick recovery. If the check box for Delete on Termination
is checked, the volume(s) will delete upon termination of the EC2 instance. If you are using an EBS-
backed instance, you can stop and restart that instance without affecting the data stored in the attached
volume. The volume remains attached throughout the stop-start cycle. This enables you to process
and store the data on your volume indefinitely, only using the processing and storage resources when
required. The data persists on the volume until the volume is deleted explicitly. The physical block
storage used by deleted EBS volumes is overwritten with zeroes before it is allocated to another account.
If you are dealing with sensitive data, you should consider encrypting your data manually or storing
the data on a volume protected by Amazon EBS encryption. For more information, see Amazon EBS
encryption (p. 1176).

By default, the root EBS volume that is created and attached to an instance at launch is deleted
when that instance is terminated. You can modify this behavior by changing the value of the flag
DeleteOnTermination to false when you launch the instance. This modified value causes the
volume to persist even after the instance is terminated, and enables you to attach the volume to another
instance.

By default, additional EBS volumes that are created and attached to an instance at launch are not
deleted when that instance is terminated. You can modify this behavior by changing the value of the flag
DeleteOnTermination to true when you launch the instance. This modified value causes the volumes
to be deleted when the instance is terminated.

Data encryption
For simplified data encryption, you can create encrypted EBS volumes with the Amazon EBS encryption
feature. All EBS volume types support encryption. You can use encrypted EBS volumes to meet a wide
range of data-at-rest encryption requirements for regulated/audited data and applications. Amazon EBS
encryption uses 256-bit Advanced Encryption Standard algorithms (AES-256) and an Amazon-managed
key infrastructure. The encryption occurs on the server that hosts the EC2 instance, providing encryption

1083
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

of data-in-transit from the EC2 instance to Amazon EBS storage. For more information, see Amazon EBS
encryption (p. 1176).

Amazon EBS encryption uses AWS Key Management Service (AWS KMS) master keys when creating
encrypted volumes and any snapshots created from your encrypted volumes. The first time you create
an encrypted EBS volume in a region, a default master key is created for you automatically. This key
is used for Amazon EBS encryption unless you select a customer master key (CMK) that you created
separately using AWS KMS. Creating your own CMK gives you more flexibility, including the ability to
create, rotate, disable, define access controls, and audit the encryption keys used to protect your data.
For more information, see the AWS Key Management Service Developer Guide.

Snapshots
Amazon EBS provides the ability to create snapshots (backups) of any EBS volume and write a copy of
the data in the volume to Amazon S3, where it is stored redundantly in multiple Availability Zones. The
volume does not need to be attached to a running instance in order to take a snapshot. As you continue
to write data to a volume, you can periodically create a snapshot of the volume to use as a baseline for
new volumes. These snapshots can be used to create multiple new EBS volumes or move volumes across
Availability Zones. Snapshots of encrypted EBS volumes are automatically encrypted.

When you create a new volume from a snapshot, it's an exact copy of the original volume at the time
the snapshot was taken. EBS volumes that are created from encrypted snapshots are automatically
encrypted. By optionally specifying a different Availability Zone, you can use this functionality to create a
duplicate volume in that zone. The snapshots can be shared with specific AWS accounts or made public.
When you create snapshots, you incur charges in Amazon S3 based on the volume's total size. For a
successive snapshot of the volume, you are only charged for any additional data beyond the volume's
original size.

Snapshots are incremental backups, meaning that only the blocks on the volume that have changed
after your most recent snapshot are saved. If you have a volume with 100 GiB of data, but only 5 GiB of
data have changed since your last snapshot, only the 5 GiB of modified data is written to Amazon S3.
Even though snapshots are saved incrementally, the snapshot deletion process is designed so that you
need to retain only the most recent snapshot.

To help categorize and manage your volumes and snapshots, you can tag them with metadata of your
choice. For more information, see Tag your Amazon EC2 resources (p. 1311).

To back up your volumes automatically, you can use Amazon Data Lifecycle Manager (p. 1186) or AWS
Backup.

Flexibility
EBS volumes support live configuration changes while in production. You can modify volume type,
volume size, and IOPS capacity without service interruptions. For more information, see Amazon EBS
Elastic Volumes (p. 1164).

Amazon EBS volume types

Amazon EBS provides the following volume types, which differ in performance characteristics and price,
so that you can tailor your storage performance and cost to the needs of your applications. The volumes
types fall into these categories:

• Solid state drives (SSD) (p. 1085) — Optimized for transactional workloads involving frequent read/
write operations with small I/O size, where the dominant performance attribute is IOPS.
• Hard disk drives (HDD) (p. 1086) — Optimized for large streaming workloads where the dominant
performance attribute is throughput.
• Previous generation (p. 1087) — Hard disk drives that can be used for workloads with small datasets
where data is accessed infrequently and performance is not of primary importance. We recommend
that you consider a current generation volume type instead.

1084
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

There are several factors that can affect the performance of EBS volumes, such as instance configuration,
I/O characteristics, and workload demand. To fully use the IOPS provisioned on an EBS volume, use EBS-
optimized instances (p. 1217). For more information about getting the most out of your EBS volumes,
see Amazon EBS volume performance on Linux instances (p. 1237).

For more information about pricing, see Amazon EBS Pricing.

Solid state drives (SSD)

The SSD-backed volumes provided by Amazon EBS fall into these categories:

• General Purpose SSD — Provides a balance of price and performance. We recommend these volumes
for most workloads.
• Provisioned IOPS SSD — Provides high performance for mission-critical, low-latency, or high-
throughput workloads.

The following is a summary of the use cases and characteristics of SSD-backed volumes. For information
about the maximum IOPS and throughput per instance, see Amazon EBS–optimized instances (p. 1217).

  General Purpose SSD Provisioned IOPS SSD

Volume gp3 gp2 io2 Block io2 io1

type Express ‡

Durability99.8% - 99.9% 99.8% - 99.9% 99.999% durability (0.001% annual 99.8% - 99.9%
durability (0.1% durability (0.1% failure rate) durability (0.1%
- 0.2% annual - 0.2% annual - 0.2% annual
failure rate) failure rate) failure rate)

Use • Low-latency interactive apps Workloads that • Workloads that require sustained
cases • Development and test require sub- IOPS performance or more than
environments millisecond 16,000 IOPS
latency, and • I/O-intensive database workloads
sustained IOPS
performance
or more than
64,000 IOPS or
1,000 MiB/s of
throughput

Volume 1 GiB - 16 TiB 4 GiB - 64 TiB 4 GiB - 16 TiB

size

Max 16,000 256,000 64,000 †

IOPS
per
volume
(16 KiB
I/O)

Max 1,000 MiB/s 250 MiB/s * 4,000 MiB/s 1,000 MiB/s †

throughput
per
volume

Amazon Not supported Not supported Supported

EBS

1085
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

  General Purpose SSD Provisioned IOPS SSD

Multi-
attach

Boot Supported
volume

* The throughput limit is between 128 MiB/s and 250 MiB/s, depending on the volume size. Volumes
smaller than or equal to 170 GiB deliver a maximum throughput of 128 MiB/s. Volumes larger than 170
GiB but smaller than 334 GiB deliver a maximum throughput of 250 MiB/s if burst credits are available.
Volumes larger than or equal to 334 GiB deliver 250 MiB/s regardless of burst credits. Older gp2 volumes
might not reach full performance unless you modify the volume (p. 1164).

† Maximum IOPS and throughput are guaranteed only on Instances built on the Nitro System (p. 209)
provisioned with more than 32,000 IOPS. Other instances guarantee up to 32,000 IOPS and 500 MiB/s.
Older io1 volumes might not reach full performance unless you modify the volume (p. 1164).

‡ io2 Block Express volumes are available as an opt-in preview only. For more information, see io2 Block
Express volumes (In preview) (p. 1092).

Hard disk drives (HDD)

The HDD-backed volumes provided by Amazon EBS fall into these categories:

• Throughput Optimized HDD — A low-cost HDD designed for frequently accessed, throughput-
intensive workloads.
• Cold HDD — The lowest-cost HDD design for less frequently accessed workloads.

The following is a summary of the use cases and characteristics of HDD-backed volumes. For information
about the maximum IOPS and throughput per instance, see Amazon EBS–optimized instances (p. 1217).

  Throughput Optimized HDD Cold HDD

Volume type st1 sc1

Durability 99.8% - 99.9% durability (0.1% - 0.2% 99.8% - 99.9% durability (0.1% - 0.2%
annual failure rate) annual failure rate)

Use cases • Big data • Throughput-oriented storage for

• Data warehouses data that is infrequently accessed
• Log processing • Scenarios where the lowest storage
cost is important

Volume size 125 GiB - 16 TiB 125 GiB - 16 TiB

Max IOPS per 500 250

volume (1 MiB I/
O)

Max throughput 500 MiB/s 250 MiB/s

per volume

Amazon EBS Not supported Not supported

Multi-attach

Boot volume Not supported Not supported

1086
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Previous generation volume types

The following table describes previous-generation EBS volume types. If you need higher performance
or performance consistency than previous-generation volumes can provide, we recommend that you
consider using General Purpose SSD (gp2 and gp3) or other current volume types. For more information,
see Previous Generation Volumes.

  Magnetic

Volume type standard

Use cases Workloads where data is infrequently accessed

Volume size 1 GiB-1 TiB

Max IOPS per volume 40–200

Max throughput per volume 40–90 MiB/s

Boot volume Supported

General Purpose SSD volumes (gp3)

General Purpose SSD (gp3) volumes offer cost-effective storage that is ideal for a broad range of
workloads. These volumes deliver a consistent baseline rate of 3,000 IOPS and 125 MiB/s, included with
the price of storage. You can provision additional IOPS (up to 16,000) and throughput (up to 1,000 MiB/
s) for an additional cost.

The maximum ratio of provisioned IOPS to provisioned volume size is 500 IOPS per GiB. The maximum
ratio of provisioned throughput to provisioned IOPS is .25 MiB/s per IOPS. The following volume
configurations support provisioning either maximum IOPS or maximum throughput:

• 32 GiB or larger: 500 IOPS/GiB x 32 GiB = 16,000 IOPS

• 8 GiB or larger and 4,000 IOPS or higher: 4,000 IOPS x 0.25 MiB/s/IOPS = 1,000 MiB/s

General Purpose SSD volumes (gp2)

General Purpose SSD (gp2) volumes offer cost-effective storage that is ideal for a broad range of
workloads. These volumes deliver single-digit millisecond latencies and the ability to burst to 3,000
IOPS for extended periods of time. Between a minimum of 100 IOPS (at 33.33 GiB and below) and a
maximum of 16,000 IOPS (at 5,334 GiB and above), baseline performance scales linearly at 3 IOPS per
GiB of volume size. AWS designs gp2 volumes to deliver their provisioned performance 99% of the time.
A gp2 volume can range in size from 1 GiB to 16 TiB.

I/O Credits and burst performance

The performance of gp2 volumes is tied to volume size, which determines the baseline performance
level of the volume and how quickly it accumulates I/O credits; larger volumes have higher baseline
performance levels and accumulate I/O credits faster. I/O credits represent the available bandwidth
that your gp2 volume can use to burst large amounts of I/O when more than the baseline performance
is needed. The more credits your volume has for I/O, the more time it can burst beyond its baseline
performance level and the better it performs when more performance is needed. The following diagram
shows the burst-bucket behavior for gp2.

1087
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Each volume receives an initial I/O credit balance of 5.4 million I/O credits, which is enough to sustain
the maximum burst performance of 3,000 IOPS for at least 30 minutes. This initial credit balance is
designed to provide a fast initial boot cycle for boot volumes and to provide a good bootstrapping
experience for other applications. Volumes earn I/O credits at the baseline performance rate of 3 IOPS
per GiB of volume size. For example, a 100 GiB gp2 volume has a baseline performance of 300 IOPS.

1088
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

When your volume requires more than the baseline performance I/O level, it draws on I/O credits in the
credit balance to burst to the required performance level, up to a maximum of 3,000 IOPS. When your
volume uses fewer I/O credits than it earns in a second, unused I/O credits are added to the I/O credit
balance. The maximum I/O credit balance for a volume is equal to the initial credit balance (5.4 million I/
O credits).

When the baseline performance of a volume is higher than maximum burst performance, I/O credits are
never spent. If the volume is attached to an instance built on the Nitro System (p. 209), the burst balance
is not reported. For other instances, the reported burst balance is 100%.

The burst duration of a volume is dependent on the size of the volume, the burst IOPS required, and the
credit balance when the burst begins. This is shown in the following equation:

(Credit balance)
Burst duration = ------------------------------------
(Burst IOPS) - 3(Volume size in GiB)

The following table lists several volume sizes and the associated baseline performance of the volume
(which is also the rate at which it accumulates I/O credits), the burst duration at the 3,000 IOPS

1089
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

maximum (when starting with a full credit balance), and the time in seconds that the volume would take
to refill an empty credit balance.

Volume size (GiB) Baseline performance Burst duration when Seconds to fill empty
(IOPS) driving sustained credit balance when
3,000 IOPS (second) driving no IO

1 100 1,802 54,000

100 300 2,000 18,000

250 750 2,400 7,200

334 (Min. size for max 1,002 2,703 5,389

throughput)

500 1,500 3,600 3,600

750 2,250 7,200 2,400

1,000 3,000 N/A* N/A*

5,334 (Min. size for max 16,000 N/A* N/A*

IOPS)

16,384 (16 TiB, max 16,000 N/A* N/A*

volume size)

* The baseline performance of the volume exceeds the maximum burst performance.

What happens if I empty my I/O credit balance?

If your gp2 volume uses all of its I/O credit balance, the maximum IOPS performance of the volume
remains at the baseline IOPS performance level (the rate at which your volume earns credits) and the
volume's maximum throughput is reduced to the baseline IOPS multiplied by the maximum I/O size.
Throughput can never exceed 250 MiB/s. When I/O demand drops below the baseline level and unused
credits are added to the I/O credit balance, the maximum IOPS performance of the volume again
exceeds the baseline. For example, a 100 GiB gp2 volume with an empty credit balance has a baseline
performance of 300 IOPS and a throughput limit of 75 MiB/s (300 I/O operations per second * 256 KiB
per I/O operation = 75 MiB/s). The larger a volume is, the greater the baseline performance is and the
faster it replenishes the credit balance. For more information about how IOPS are measured, see I/O
characteristics and monitoring (p. 1239).

If you notice that your volume performance is frequently limited to the baseline level (due to an empty I/
O credit balance), you should consider switching to a gp3 volume.

For information about using CloudWatch metrics and alarms to monitor your burst bucket balance, see
Monitor the burst bucket balance for volumes (p. 1101).

Throughput performance

Throughput for a gp2 volume can be calculated using the following formula, up to the throughput limit
of 250 MiB/s:

Throughput in MiB/s = ((Volume size in GiB) × (IOPS per GiB) × (I/O size in KiB))

Assuming V = volume size, I = I/O size, R = I/O rate, and T = throughput, this can be simplified to:

1090
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

T = VIR

The smallest volume size that achieves the maximum throughput is given by:

T
V = -----
I R

250 MiB/s
= ---------------------
(256 KiB)(3 IOPS/GiB)

[(250)(2^20)(Bytes)]/s
= ------------------------------------------
(256)(2^10)(Bytes)([3 IOP/s]/[(2^30)(Bytes)])

(250)(2^20)(2^30)(Bytes)
= ------------------------
(256)(2^10)(3)

= 357,913,941,333 Bytes

= 333# GiB (334 GiB in practice because volumes are provisioned in whole gibibytes)

Provisioned IOPS SSD volumes

Provisioned IOPS SSD (io1 and io2) volumes are designed to meet the needs of I/O-intensive
workloads, particularly database workloads, that are sensitive to storage performance and consistency.
Provisioned IOPS SSD volumes use a consistent IOPS rate, which you specify when you create the
volume, and Amazon EBS delivers the provisioned performance 99.9 percent of the time.

io1 volumes are designed to provide 99.8 to 99.9 percent volume durability with an annual failure
rate (AFR) no higher than 0.2 percent, which translates to a maximum of two volume failures per 1,000
running volumes over a one-year period. io2 volumes are designed to provide 99.999 percent volume
durability with an AFR no higher than 0.001 percent, which translates to a single volume failure per
100,000 running volumes over a one-year period.

Provisioned IOPS SSD io1 volumes are available for all Amazon EC2 instance types. Provisioned IOPS
SSD io2 volumes are available for all EC2 instances types, with the exception of R5b.

Provisioned IOPS SSD volumes can range in size from 4 GiB to 16 TiB. You can provision from 100 IOPS
up to 64,000 IOPS per volume on Instances built on the Nitro System (p. 209) and up to 32,000 on other
instances. The maximum ratio of provisioned IOPS to requested volume size (in GiB) is 50:1 for io1
volumes, and 500:1 for io2 volumes. For example, a 100 GiB io1 volume can be provisioned with up
to 5,000 IOPS, while a 100 GiB io2 volume can be provisioned with up to 50,000 IOPS. On a supported
instance type, the following volume sizes allow provisioning up to the 64,000 IOPS maximum:

• io1 volume 1,280 GiB in size or greater (50 × 1,280 GiB = 64,000 IOPS)
• io2 volume 128 GiB in size or greater (500 × 128 GiB = 64,000 IOPS)

Provisioned IOPS SSD volumes provisioned with up to 32,000 IOPS support a maximum I/O size of 256
KiB and yield as much as 500 MiB/s of throughput. With the I/O size at the maximum, peak throughput
is reached at 2,000 IOPS. A volume provisioned with more than 32,000 IOPS (up to the cap of 64,000
IOPS) supports a maximum I/O size of 16 KiB and yields as much as 1,000 MiB/s of throughput. The
following graph illustrates these performance characteristics:

1091
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Your per-I/O latency experience depends on the provisioned IOPS and on your workload profile. For the
best I/O latency experience, ensure that you provision IOPS to meet the I/O profile of your workload.

io2 Block Express volumes (In preview)

Note
You can now opt in to the io2 Block Express volumes preview in us-east-1, us-west-2,
us-east-2, eu-central-1, ap-southeast-1, and ap-northeast-1. After you opt in, all
io2 volumes that you create in your account in the opted-in Regions will be io2 Block Express
volumes. Carefully review the feature limitations that apply to the preview before opting in.
You can opt in to the preview only with an account that does not have any existing io2 volumes.

EBS Block Express is the next generation of Amazon EBS storage server architecture. It has been built for
the purpose of meeting the performance requirements of the most demanding I/O intensive applications
that run on Nitro-based Amazon EC2 instances.

Block Express architecture increases performance and scale. Block Express servers communicate with
Nitro-based instances using the Scalable Reliable Datagram (SRD) networking protocol. This interface
is implemented in the Nitro Card dedicated for Amazon EBS I/O function on the host hardware of the
instance. It minimizes I/O delay and latency variation (network jitter), which provides faster and more
consistent performance for your applications. For more information, see io2 Block Express volumes.

io2 Block Express volumes are suited for workloads that benefit from a single volume that provides sub-
millisecond latency, and supports higher IOPS, higher throughput, and larger capacity than io2 volumes.

Topics
• Performance (p. 1093)
• Limitations during preview (p. 1093)

1092
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

• Opt in to the preview (p. 1094)

• Opt out of the preview (p. 1094)
• Quotas (p. 1094)
• Pricing and billing (p. 1094)

Performance

With io2 Block Express volumes, you can provision volumes with:

• Sub-millisecond average latency

• Storage capacity up to 64 TiB (65,536 GiB)
• Provisioned IOPS up to 256,000, with an IOPS:GiB ratio of 1,000:1. Maximum IOPS can be provisioned
with volumes 256 GiB in size and larger (1,000 IOPS x 256 GiB = 256,000 IOPS).
• Volume throughput up to 4,000 MiB/s. Throughput scales proportionally up to 0.256 MiB/s per
provisioned IOPS. Maximum throughput can be achieved at 16,000 IOPS or higher.

Limitations during preview

io2 Block Express volumes have the following limitations during the preview:

• io2 Block Express volumes are supported with R5B instances only. If you attempt to launch an
unsupported instance type with io2 Block Express volumes, the request fails.
• io2 Block Express volumes are available in the following Regions only: us-east-1, us-west-2, us-
east-2, eu-central-1, ap-southeast-1, and ap-northeast-1.
• io2 Block Express volumes are not optimized for boot and could cause longer boot times. We
recommend that you continue to use io1, gp2, or gp3 for boot volumes.
• You can't create encrypted io2 Block Express volumes, or launch instances with encrypted io2
Block Express volumes, from unencrypted snapshots or shared snapshots. To create encrypted io2

1093
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Block Express volumes, or launch instances with encrypted io2 Block Express volumes, you must use
encrypted snapshots or AMIs that you own in your account.
• You can't create encrypted io2 Block Express volumes, or launch instances with encrypted io2 Block
Express volumes using the Amazon EC2 console. You must use the Amazon EC2 API or the AWS CLI.
• io2 Block Express volumes do not fully support fast snapshot restore. While you can restore io2 Block
Express volumes from snapshots that are enabled for fast snapshot restore, you will experience higher
I/O latency while the volumes are being initialized. This is formerly known as pre-warming. To avoid
the higher latency, we recommend that you initialize your io2 Block Express volumes before starting
your I/O operations. For more information, see Initialize Amazon EBS volumes (p. 1241).
• io2 Block Express volumes do not support Elastic Volume operations. You can't change the volume
type or performance after the volume has been created.
• io2 Block Express volumes do not support Multi-Attach.

Opt in to the preview

To opt in to the io2 Block Express volumes preview, visit the io2 Block Express volumes Preview page
and complete the form. You must specify the Regions in which to opt in.

After you opt in, all new io2 volumes that you create in your account in the opted-in Regions will be io2
Block Express volumes. While your account is opted in, you will not be able to create io2 volumes that
are not io2 Block Express volumes in the opted-in Regions. To transition your existing workloads from
your existing volumes, create snapshots of the volumes and then restore them to new io2 Block Express
volumes.

Opt out of the preview

To opt out of the io2 Block Express volumes preview, visit the io2 Block Express volumes Preview page,
leave all of the Region check boxes unselected and complete the rest of the form.

After you opt out, all new io2 volumes that you create in the opted-out Regions will be io2 volumes,
and not io2 Block Express volumes. io2 volumes that you created in the opted-out Regions before
opting out will continue to be io2 Block Express volumes. If you created snapshots of volumes larger
than 16 TiB, you will not be able to restore them to io2 volumes in the opted-out Regions. Before opting
out, we recommend that for io2 Block Express volumes greater than 16 TiB you copy the data to a raid
set that comprises of io2 volumes less than 16 TiB in size and then create a backup of the raid set using
multi-volume snapshots. For more information, see Multi-volume snapshots (p. 1130).

Quotas
io2 Block Express volumes adhere to the same service quotas as io2 volumes. For more information, see
Amazon EBS quotas.

Pricing and billing

io2 volumes and io2 Block Express volumes are billed at the same rate. For more information, see
Amazon EBS pricing.

Usage reports do not distinguish between io2 Block Express volumes and io2 volumes. We recommend
that you use tags to help you identify costs associated with io2 Block Express volumes.

Throughput Optimized HDD volumes

Throughput Optimized HDD (st1) volumes provide low-cost magnetic storage that defines performance
in terms of throughput rather than IOPS. This volume type is a good fit for large, sequential workloads
such as Amazon EMR, ETL, data warehouses, and log processing. Bootable st1 volumes are not
supported.

Throughput Optimized HDD (st1) volumes, though similar to Cold HDD (sc1) volumes, are designed to
support frequently accessed data.

1094
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

This volume type is optimized for workloads involving large, sequential I/O, and we recommend that
customers with workloads performing small, random I/O use gp2. For more information, see Inefficiency
of small read/writes on HDD (p. 1101).

Throughput credits and burst performance

Like gp2, st1 uses a burst-bucket model for performance. Volume size determines the baseline
throughput of your volume, which is the rate at which the volume accumulates throughput credits.
Volume size also determines the burst throughput of your volume, which is the rate at which you can
spend credits when they are available. Larger volumes have higher baseline and burst throughput. The
more credits your volume has, the longer it can drive I/O at the burst level.

The following diagram shows the burst-bucket behavior for st1.

Subject to throughput and throughput-credit caps, the available throughput of an st1 volume is
expressed by the following formula:

(Volume size) x (Credit accumulation rate per TiB) = Throughput

For a 1-TiB st1 volume, burst throughput is limited to 250 MiB/s, the bucket fills with credits at 40 MiB/
s, and it can hold up to 1 TiB-worth of credits.

Larger volumes scale these limits linearly, with throughput capped at a maximum of 500 MiB/s. After the
bucket is depleted, throughput is limited to the baseline rate of 40 MiB/s per TiB.

On volume sizes ranging from 0.125 TiB to 16 TiB, baseline throughput varies from 5 MiB/s to a cap of
500 MiB/s, which is reached at 12.5 TiB as follows:

40 MiB/s
12.5 TiB x ---------- = 500 MiB/s
1 TiB

Burst throughput varies from 31 MiB/s to a cap of 500 MiB/s, which is reached at 2 TiB as follows:

250 MiB/s
2 TiB x ---------- = 500 MiB/s
1 TiB

The following table states the full range of base and burst throughput values for st1:

1095
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Volume size (TiB) ST1 base throughput (MiB/s) ST1 burst throughput (MiB/s)

0.125 5 31

0.5 20 125

1 40 250

2 80 500

3 120 500

4 160 500

5 200 500

6 240 500

7 280 500

8 320 500

9 360 500

10 400 500

11 440 500

12 480 500

12.5 500 500

13 500 500

14 500 500

15 500 500

16 500 500

The following diagram plots the table values:

Note
When you create a snapshot of a Throughput Optimized HDD (st1) volume, performance may
drop as far as the volume's baseline value while the snapshot is in progress.

1096
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

For information about using CloudWatch metrics and alarms to monitor your burst bucket balance, see
Monitor the burst bucket balance for volumes (p. 1101).

Cold HDD volumes

Cold HDD (sc1) volumes provide low-cost magnetic storage that defines performance in terms of
throughput rather than IOPS. With a lower throughput limit than st1, sc1 is a good fit for large,
sequential cold-data workloads. If you require infrequent access to your data and are looking to save
costs, sc1 provides inexpensive block storage. Bootable sc1 volumes are not supported.

Cold HDD (sc1) volumes, though similar to Throughput Optimized HDD (st1) volumes, are designed to
support infrequently accessed data.
Note
This volume type is optimized for workloads involving large, sequential I/O, and we recommend
that customers with workloads performing small, random I/O use gp2. For more information,
see Inefficiency of small read/writes on HDD (p. 1101).

Throughput credits and burst performance

Like gp2, sc1 uses a burst-bucket model for performance. Volume size determines the baseline
throughput of your volume, which is the rate at which the volume accumulates throughput credits.
Volume size also determines the burst throughput of your volume, which is the rate at which you can
spend credits when they are available. Larger volumes have higher baseline and burst throughput. The
more credits your volume has, the longer it can drive I/O at the burst level.

Subject to throughput and throughput-credit caps, the available throughput of an sc1 volume is
expressed by the following formula:

(Volume size) x (Credit accumulation rate per TiB) = Throughput

For a 1-TiB sc1 volume, burst throughput is limited to 80 MiB/s, the bucket fills with credits at 12 MiB/s,
and it can hold up to 1 TiB-worth of credits.

Larger volumes scale these limits linearly, with throughput capped at a maximum of 250 MiB/s. After the
bucket is depleted, throughput is limited to the baseline rate of 12 MiB/s per TiB.

On volume sizes ranging from 0.125 TiB to 16 TiB, baseline throughput varies from 1.5 MiB/s to a
maximum of 192 MiB/s, which is reached at 16 TiB as follows:

12 MiB/s

1097
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

16 TiB x ---------- = 192 MiB/s

1 TiB

Burst throughput varies from 10 MiB/s to a cap of 250 MiB/s, which is reached at 3.125 TiB as follows:

80 MiB/s
3.125 TiB x ----------- = 250 MiB/s
1 TiB

The following table states the full range of base and burst throughput values for sc1:

Volume Size (TiB) SC1 Base Throughput (MiB/s) SC1 Burst Throughput (MiB/s)

0.125 1.5 10

0.5 6 40

1 12 80

2 24 160

3 36 240

3.125 37.5 250

4 48 250

5 60 250

6 72 250

7 84 250

8 96 250

9 108 250

10 120 250

11 132 250

12 144 250

13 156 250

14 168 250

15 180 250

16 192 250

The following diagram plots the table values:

1098
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Note
When you create a snapshot of a Cold HDD (sc1) volume, performance may drop as far as the
volume's baseline value while the snapshot is in progress.

For information about using CloudWatch metrics and alarms to monitor your burst bucket balance, see
Monitor the burst bucket balance for volumes (p. 1101).

Magnetic volumes
Magnetic volumes are backed by magnetic drives and are suited for workloads where data is accessed
infrequently, and scenarios where low-cost storage for small volume sizes is important. These volumes
deliver approximately 100 IOPS on average, with burst capability of up to hundreds of IOPS, and they
can range in size from 1 GiB to 1 TiB.
Note
Magnetic is a previous generation volume type. For new applications, we recommend using one
of the newer volume types. For more information, see Previous Generation Volumes.

For information about using CloudWatch metrics and alarms to monitor your burst bucket balance, see
Monitor the burst bucket balance for volumes (p. 1101).

Performance considerations when using HDD volumes

For optimal throughput results using HDD volumes, plan your workloads with the following
considerations in mind.

Comparing Throughput Optimized HDD and Cold HDD

The st1 and sc1 bucket sizes vary according to volume size, and a full bucket contains enough tokens
for a full volume scan. However, larger st1 and sc1 volumes take longer for the volume scan to
complete due to per-instance and per-volume throughput limits. Volumes attached to smaller instances
are limited to the per-instance throughput rather than the st1 or sc1 throughput limits.

Both st1 and sc1 are designed for performance consistency of 90% of burst throughput 99% of the
time. Non-compliant periods are approximately uniformly distributed, targeting 99% of expected total
throughput each hour.

The following table shows ideal scan times for volumes of various size, assuming full buckets and
sufficient instance throughput.

In general, scan times are expressed by this formula:

Volume size
------------- = Scan time

1099
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Throughput

For example, taking the performance consistency guarantees and other optimizations into account, an
st1 customer with a 5-TiB volume can expect to complete a full volume scan in 2.91 to 3.27 hours.

5 TiB 5 TiB
----------- = ------------------- = 10,486 s = 2.91 hours (optimal)
500 MiB/s 0.00047684 TiB/s

2.91 hours
2.91 hours + -------------- = 3.27 hours (minimum expected)
(0.90)(0.99) <-- From expected performance of 90% of burst 99% of the time

Similarly, an sc1 customer with a 5-TiB volume can expect to complete a full volume scan in 5.83 to 6.54
hours.

5 TiB
------------------- = 20972 s = 5.83 hours (optimal)
0.000238418 TiB/s

5.83 hours
-------------- = 6.54 hours (minimum expected)
(0.90)(0.99)

Volume size (TiB) ST1 scan time with burst SC1 scan time with burst
(hours)* (hours)*

1 1.17 3.64

2 1.17 3.64

3 1.75 3.64

4 2.33 4.66

5 2.91 5.83

6 3.50 6.99

7 4.08 8.16

8 4.66 9.32

9 5.24 10.49

10 5.83 11.65

11 6.41 12.82

12 6.99 13.98

13 7.57 15.15

14 8.16 16.31

15 8.74 17.48

16 9.32 18.64

1100
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

* These scan times assume an average queue depth (rounded to the nearest whole number) of four or
more when performing 1 MiB of sequential I/O.

Therefore if you have a throughput-oriented workload that needs to complete scans quickly (up to 500
MiB/s), or requires several full volume scans a day, use st1. If you are optimizing for cost, your data is
relatively infrequently accessed, and you don’t need more than 250 MiB/s of scanning performance, then
use sc1.

Inefficiency of small read/writes on HDD

The performance model for st1 and sc1 volumes is optimized for sequential I/Os, favoring high-
throughput workloads, offering acceptable performance on workloads with mixed IOPS and throughput,
and discouraging workloads with small, random I/O.

For example, an I/O request of 1 MiB or less counts as a 1 MiB I/O credit. However, if the I/Os are
sequential, they are merged into 1 MiB I/O blocks and count only as a 1 MiB I/O credit.

Limitations on per-instance throughput

Throughput for st1 and sc1 volumes is always determined by the smaller of the following:

• Throughput limits of the volume

• Throughput limits of the instance

As for all Amazon EBS volumes, we recommend that you select an appropriate EBS-optimized EC2
instance in order to avoid network bottlenecks. For more information, see Amazon EBS–optimized
instances (p. 1217).

Monitor the burst bucket balance for volumes

You can monitor the burst-bucket level for gp2, st1, and sc1 volumes using the EBS BurstBalance
metric available in Amazon CloudWatch. This metric shows the percentage of I/O credits (for gp2)
or throughput credits (for st1 and sc1) remaining in the burst bucket. For more information
about the BurstBalance metric and other metrics related to I/O, see I/O characteristics and
monitoring (p. 1239). CloudWatch also allows you to set an alarm that notifies you when the
BurstBalance value falls to a certain level. For more information, see Creating Amazon CloudWatch
Alarms.

Constraints on the size and configuration of an EBS volume

The size of an Amazon EBS volume is constrained by the physics and arithmetic of block data storage,
as well as by the implementation decisions of operating system (OS) and file system designers. AWS
imposes additional limits on volume size to safeguard the reliability of its services.

The following sections describe the most important factors that limit the usable size of an EBS volume
and offer recommendations for configuring your EBS volumes.

Contents
• Storage capacity (p. 1101)
• Service limitations (p. 1102)
• Partitioning schemes (p. 1102)
• Data block sizes (p. 1103)

Storage capacity
The following table summarizes the theoretical and implemented storage capacities for the most
commonly used file systems on Amazon EBS, assuming a 4,096 byte block size.

1101
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Partitioning Max Theoretical Ext4 XFS NTFS Max

scheme addressable max size implemented implemented implemented supported
blocks (blocks × max size* max size** max size by EBS
block size)
32
MBR 2 2 TiB 2 TiB 2 TiB 2 TiB 2 TiB
64
GPT † 2 64 ZiB 1 EiB = 500 TiB 256 TiB 16 TiB
2
1024 TiB
(certified on
(50 TiB RHEL7)
certified on
RHEL7)

* https://ext4.wiki.kernel.org/index.php/Ext4_Howto and https://access.redhat.com/solutions/1532

** https://access.redhat.com/solutions/1532

† io2 Block Express volumes support up to 64 TiB for GPT partitions.

Service limitations
Amazon EBS abstracts the massively distributed storage of a data center into virtual hard disk drives. To
an operating system installed on an EC2 instance, an attached EBS volume appears to be a physical hard
disk drive containing 512-byte disk sectors. The OS manages the allocation of data blocks (or clusters)
onto those virtual sectors through its storage management utilities. The allocation is in conformity with
a volume partitioning scheme, such as master boot record (MBR) or GUID partition table (GPT), and
within the capabilities of the installed file system (ext4, NTFS, and so on).

EBS is not aware of the data contained in its virtual disk sectors; it only ensures the integrity of the
sectors. This means that AWS actions and OS actions are independent of each other. When you are
selecting a volume size, be aware of the capabilities and limits of both, as in the following cases:

• EBS currently supports a maximum volume size of 16 TiB. This means that you can create an EBS
volume as large as 16 TiB, but whether the OS recognizes all of that capacity depends on its own
design characteristics and on how the volume is partitioned.
• Linux boot volumes may use either the MBR or GPT partitioning scheme. MBR supports boot volumes
up to 2047 GiB (2 TiB - 1 GiB). GPT with GRUB 2 supports boot volumes 2 TiB or larger. If your Linux
AMI uses MBR, your boot volume is limited to 2047 GiB, but your non-boot volumes do not have this
limit. For more information, see Make an Amazon EBS volume available for use on Linux (p. 1111).

Partitioning schemes
Among other impacts, the partitioning scheme determines how many logical data blocks can be uniquely
addressed in a single volume. For more information, see Data block sizes (p. 1103). The common
partitioning schemes in use are master boot record (MBR) and GUID partition table (GPT). The important
differences between these schemes can be summarized as follows.

MBR

MBR uses a 32-bit data structure to store block addresses. This means that each data block is mapped
32
with one of 2 possible integers. The maximum addressable size of a volume is given by:

(232 - 1) × Block size = Number of addressable blocks

1102
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

The block size for MBR volumes is conventionally limited to 512 bytes. Therefore:

(232 - 1) × 512 bytes = 2 TiB - 512 bytes

Engineering workarounds to increase this 2-TiB limit for MBR volumes have not met with widespread
industry adoption. Consequently, Linux and Windows never detect an MBR volume as being larger than 2
TiB even if AWS shows its size to be larger.

GPT

GPT uses a 64-bit data structure to store block addresses. This means that each data block is mapped
64
with one of 2 possible integers. The maximum addressable size of a volume is given by:

(264 - 1) × Block size = Number of addressable blocks

The block size for GPT volumes is commonly 4,096 bytes. Therefore:

(264 - 1) × 4,096 bytes

= 264 x 4,096 bytes - 1 x 4,096 bytes
= 264 x 212 bytes - 4,096 bytes
= 270 x 26 bytes - 4,096 bytes
= 64 ZiB - 4,096 bytes

Real-world computer systems don't support anything close to this theoretical maximum. Implemented
file-system size is currently limited to 50 TiB for ext4 and 256 TiB for NTFS—both of which exceed the
16-TiB limit imposed by AWS.

Data block sizes

Data storage on a modern hard drive is managed through logical block addressing, an abstraction layer
that allows the operating system to read and write data in logical blocks without knowing much about
the underlying hardware. The OS relies on the storage device to map the blocks to its physical sectors.
EBS advertises 512-byte sectors to the operating system, which reads and writes data to disk using data
blocks that are a multiple of the sector size.

The industry default size for logical data blocks is currently 4,096 bytes (4 KiB). Because certain
workloads benefit from a smaller or larger block size, file systems support non-default block sizes
that can be specified during formatting. Scenarios in which non-default block sizes should be used are
outside the scope of this topic, but the choice of block size has consequences for the storage capacity of
the volume. The following table shows storage capacity as a function of block size:

Block size Max volume size

4 KiB (default) 16 TiB

8 KiB 32 TiB

16 KiB 64 TiB

32 KiB 128 TiB

64 KiB (maximum) 256 TiB

The EBS-imposed limit on volume size (16 TiB) is currently equal to the maximum size enabled by 4-KiB
data blocks.

1103
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Create an Amazon EBS volume

You can create an Amazon EBS volume and then attach it to any EC2 instance in the same Availability
Zone. If you create an encrypted EBS volume, you can only attach it to supported instance types. For
more information, see Supported instance types (p. 1177).

If you are creating a volume for a high-performance storage scenario, you should make sure to use
a Provisioned IOPS SSD volume (io1 or io2) and attach it to an instance with enough bandwidth to
support your application, such as an EBS-optimized instance. The same advice holds for Throughput
Optimized HDD (st1) and Cold HDD (sc1) volumes. For more information, see Amazon EBS–optimized
instances (p. 1217).

Empty EBS volumes receive their maximum performance the moment that they are available and do not
require initialization (formerly known as pre-warming). However, storage blocks on volumes that were
created from snapshots must be initialized (pulled down from Amazon S3 and written to the volume)
before you can access the block. This preliminary action takes time and can cause a significant increase
in the latency of an I/O operation the first time each block is accessed. Volume performance is achieved
after all blocks have been downloaded and written to the volume. For most applications, amortizing this
cost over the lifetime of the volume is acceptable. To avoid this initial performance hit in a production
environment, you can force immediate initialization of the entire volume or enable fast snapshot restore.
For more information, see Initialize Amazon EBS volumes (p. 1241).

Methods of creating a volume

• Create and attach EBS volumes when you launch instances by specifying a block device mapping. For
more information, see Launch an instance using the Launch Instance Wizard (p. 526) and Block device
mapping (p. 1294).
• Create an empty EBS volume and attach it to a running instance. For more information, see Create an
empty volume (p. 1104) below.
• Create an EBS volume from a previously created snapshot and attach it to a running instance. For more
information, see Create a volume from a snapshot (p. 1105) below.

Create an empty volume

Empty volumes receive their maximum performance the moment that they are available and do not
require initialization.

To create a empty EBS volume using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region in which you would like to create your volume. This
choice is important because some Amazon EC2 resources can be shared between Regions, while
others can't. For more information, see Resource locations (p. 1304).
3. In the navigation pane, choose ELASTIC BLOCK STORE, Volumes.
4. Choose Create Volume.
5. For Volume Type, choose a volume type. For more information, see Amazon EBS volume
types (p. 1084).
6. For Size, enter the size of the volume, in GiB. For more information, see Constraints on the size and
configuration of an EBS volume (p. 1101).
7. For IOPS, enter the maximum number of input/output operations per second (IOPS) that the
volume should provide. You can specify IOPS only for gp3, io1, and io2 volumes.
8. For Throughput, enter the throughput that the volume should provide, in MiB/s. You can specify
throughput only for gp3 volumes.

1104
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

9. For Availability Zone, choose the Availability Zone in which to create the volume. An EBS volume
must be attached to an EC2 instance that is in the same Availability Zone as the volume.
10. (Optional) If the instance type supports EBS encryption and you want to encrypt the volume, select
Encrypt this volume and choose a CMK. If encryption by default is enabled in this Region, EBS
encryption is enabled and the default CMK for EBS encryption is chosen. You can choose a different
CMK from Master Key or paste the full ARN of any key that you can access. For more information,
see Amazon EBS encryption (p. 1176).
11. (Optional) Choose Create additional tags to add tags to the volume. For each tag, provide a tag key
and a tag value. For more information, see Tag your Amazon EC2 resources (p. 1311).
12. Choose Create Volume. The volume is ready for use when the volume status is Available.
13. To use your new volume, attach it to an instance, format it, and mount it. For more information, see
Attach an Amazon EBS volume to an instance (p. 1106).

To create an empty EBS volume using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• create-volume (AWS CLI)

• New-EC2Volume (AWS Tools for Windows PowerShell)

Create a volume from a snapshot

Volumes created from snapshots load lazily in the background. This means that there is no need to
wait for all of the data to transfer from Amazon S3 to your EBS volume before the instance can start
accessing an attached volume and all its data. If your instance accesses data that hasn't yet been
loaded, the volume immediately downloads the requested data from Amazon S3, and then continues
loading the rest of the volume data in the background. Volume performance is achieved after all
blocks are downloaded and written to the volume. To avoid the initial performance hit in a production
environment, see Initialize Amazon EBS volumes (p. 1241).

New EBS volumes that are created from encrypted snapshots are automatically encrypted. You can also
encrypt a volume on-the-fly while restoring it from an unencrypted snapshot. Encrypted volumes can
only be attached to instance types that support EBS encryption. For more information, see Supported
instance types (p. 1177).

Use the following procedure to create a volume from a snapshot.

To create an EBS volume from a snapshot using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region that your snapshot is located in.

To use the snapshot to create a volume in a different region, copy your snapshot to the new Region
and then use it to create a volume in that Region. For more information, see Copy an Amazon EBS
snapshot (p. 1134).
3. In the navigation pane, choose ELASTIC BLOCK STORE, Volumes.
4. Choose Create Volume.
5. For Volume Type, choose a volume type. For more information, see Amazon EBS volume
types (p. 1084).
6. For Snapshot ID, start typing the ID or description of the snapshot from which you are restoring the
volume, and choose it from the list of suggested options.

1105
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

7. (Optional) Select Encrypt this volume to change the encryption state of your volume. This is
optional if encryption by default (p. 1178) is enabled. Select a CMK from Master Key to specify a
CMK other than the default CMK for EBS encryption.
8. For Size, verify that the default size of the snapshot meets your needs or enter the size of the
volume, in GiB.

If you specify both a volume size and a snapshot, the size must be equal to or greater than the
snapshot size. When you select a volume type and a snapshot, the minimum and maximum sizes
for the volume are shown next to Size. For more information, see Constraints on the size and
configuration of an EBS volume (p. 1101).
9. For IOPS, enter the maximum number of input/output operations per second (IOPS) that the
volume should provide. You can specify IOPS only for gp3, io1, and io2 volumes.
10. For Throughput, enter the throughput that the volume should provide, in MiB/s. You can specify
throughput only for gp3 volumes.
11. For Availability Zone, choose the Availability Zone in which to create the volume. An EBS volume
must be attached to an EC2 instance that is in the same Availability Zone as the volume.
12. (Optional) Choose Create additional tags to add tags to the volume. For each tag, provide a tag key
and a tag value.
13. Choose Create Volume.
14. To use your new volume, attach it to an instance and mount it. For more information, see Attach an
Amazon EBS volume to an instance (p. 1106).
15. If you created a volume that is larger than the snapshot, you must extend the file system on
the volume to take advantage of the extra space. For more information, see Amazon EBS Elastic
Volumes (p. 1164).

To create an EBS volume from a snapshot using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• create-volume (AWS CLI)

• New-EC2Volume (AWS Tools for Windows PowerShell)

Attach an Amazon EBS volume to an instance

You can attach an available EBS volume to one or more of your instances that is in the same Availability
Zone as the volume.

For information about adding EBS volumes to your instance at launch, see Instance block device
mapping (p. 1299).

Prerequisites

• Determine how many volumes you can attach to your instance. For more information, see Instance
volume limits (p. 1291).
• Determine whether you can attach your volume to multiple instances and enable Multi-Attach. For
more information, see Attach a volume to multiple instances with Amazon EBS Multi-Attach (p. 1107).
• If a volume is encrypted, it can only be attached to an instance that supports Amazon EBS encryption.
For more information, see Supported instance types (p. 1177).
• If a volume has an AWS Marketplace product code:
• The volume can only be attached to a stopped instance.
• You must be subscribed to the AWS Marketplace code that is on the volume.

1106
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

• The configuration (instance type, operating system) of the instance must support that specific AWS
Marketplace code. For example, you cannot take a volume from a Windows instance and attach it to
a Linux instance.
• AWS Marketplace product codes are copied from the volume to the instance.

To attach an EBS volume to an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic Block Store, Volumes.
3. Select an available volume and choose Actions, Attach Volume.
4. For Instance, start typing the name or ID of the instance. Select the instance from the list of options
(only instances that are in the same Availability Zone as the volume are displayed).
5. For Device, you can keep the suggested device name, or type a different supported device name. For
more information, see Name devices on Linux instances (p. 1292).
6. Choose Attach.
7. Connect to your instance and mount the volume. For more information, see Make an Amazon EBS
volume available for use on Linux (p. 1111).

To attach an EBS volume to an instance using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• attach-volume (AWS CLI)

• Add-EC2Volume (AWS Tools for Windows PowerShell)

Attach a volume to multiple instances with Amazon EBS Multi-

Attach
Amazon EBS Multi-Attach enables you to attach a single Provisioned IOPS SSD (io1 or io2) volume to
up to 16 Nitro-based instances that are in the same Availability Zone. You can attach multiple Multi-
Attach enabled volumes to an instance or set of instances. Each instance to which the volume is attached
has full read and write permission to the shared volume. Multi-Attach makes it easier for you to achieve
higher application availability in clustered Linux applications that manage concurrent write operations.

Contents
• Considerations and limitations (p. 319)
• Performance (p. 1108)
• Work with Multi-Attach (p. 1109)
• Monitor a Multi-Attach enabled volume (p. 1111)
• Pricing and billing (p. 1111)

Considerations and limitations

• Multi-Attach enabled volumes can be attached to up to 16 Linux instances built on the Nitro
System (p. 209) that are in the same Availability Zone. You can attach a volume that is Multi-Attach
enabled to Windows instances, but the operating system does not recognize the data on the volume
that is shared between the instances, which can result in data inconsistency.
• Multi-Attach is supported exclusively on Provisioned IOPS SSD volumes (p. 1091).

1107
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

• Multi-Attach is available in the following Regions only:

• For io2 volumes—us-east-2, eu-central-1, and ap-south-1.
• For io1 volumes—us-east-1, us-west-2, eu-west-1, and ap-northeast-2.
• Multi-Attach enabled volumes can't be attached to R5b instances.
• Multi-Attach enabled volumes do not support I/O fencing. I/O fencing protocols control write access
in a shared storage environment to maintain data consistency. Your applications must provide write
ordering for the attached instances to maintain data consistency.
• Multi-Attach enabled volumes can't be created as boot volumes.
• Multi-Attach enabled volumes can be attached to one block device mapping per instance.
• Multi-Attach can't be enabled during instance launch using either the Amazon EC2 console or
RunInstances API.
• Multi-Attach enabled volumes that have an issue at the Amazon EBS infrastructure layer are
unavailable to all attached instances. Issues at the Amazon EC2 or networking layer might impact only
some attached instances.
• The following table shows volume modification support for Multi-Attach enabled io1 and io2
volumes after creation.

  io2 volumes io1 volumes

Modify volume ✓* ✗
type

Modify volume ✓ ✗
size

Modify ✓ ✗
provisioned IOPS

Enable Multi- ✓ ✗
Attach

Disable Multi- ✓ ** ✗
Attach

* To change the volume type, first disable Multi-Attach.

** You can't disable Multi-Attach if the volume is attached to more than one instance.

Performance
Each attached instance is able to drive its maximum IOPS performance up to the volume's maximum
provisioned performance. However, the aggregate performance of all of the attached instances can't
exceed the volume's maximum provisioned performance. If the attached instances' demand for IOPS is
higher than the volume's Provisioned IOPS, the volume will not exceed its provisioned performance.

For example, say you create an io2 Multi-Attach enabled volume with 50,000 Provisioned IOPS and you
attach it to an m5.8xlarge instance and a c5.12xlarge instance. The m5.8xlarge and c5.12xlarge
instances support a maximum of 30,000 and 40,000 IOPS respectively. Each instance can drive its
maximum IOPS as it is less than the volume's Provisioned IOPS of 50,000. However, if both instances
drive I/O to the volume simultaneously, their combined IOPS can't exceed the volume's provisioned
performance of 50,000 IOPS. The volume will not exceed 50,000 IOPS.

To achieve consistent performance, it is best practice to balance I/O driven from attached instances
across the sectors of a Multi-Attach enabled volume.

1108
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Work with Multi-Attach

Multi-Attach enabled volumes can be managed in much the same way that you would manage any other
Amazon EBS volume. However, in order to use the Multi-Attach functionality, you must enable it for the
volume. When you create a new volume, Multi-Attach is disabled by default.

Contents
• Enable Multi-Attach (p. 1109)
• Disable Multi-Attach (p. 1110)
• Attach a volume to instances (p. 1110)
• Delete on termination (p. 1110)

Enable Multi-Attach

You can enable Multi-Attach for io1 and io2 volumes during creation.

Use one of the following methods to enable Multi-Attach for an io1 or io2 volume during creation.

Console

To enable Multi-Attach during volume creation

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Volumes.
3. Choose Create Volume.
4. For Volume Type, choose Provisioned IOPS SSD (io1) or Provisioned IOPS SSD (io2).
5. For Size and IOPS, choose the required volume size and the number of IOPS to provision.
6. For Availability Zone, choose the same Availability Zone that the instances are in.
7. For Multi-Attach, choose Enable.
8. Choose Create Volume.

Command line

To enable Multi-Attach during volume creation

Use the create-volume command and specify the --multi-attach-enabled parameter.

$ aws ec2 create-volume --volume-type io2 --multi-attach-enabled --size 100 --iops 2000
--region us-west-2 --availability-zone us-west-2b

You can also enable Multi-Attach for io2 volumes after they have been created.
Note
You can't enable Multi-Attach for io1 volumes after creation.

Use one of the following methods to enable Multi-Attach for an Amazon EBS volume after it has been
created.

Console

To enable Multi-Attach after creation

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

1109
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

2. In the navigation pane, choose Volumes.

3. Select the volume and choose Actions, Modify Volume.
4. For Multi-Attach, choose Enable.
5. Choose Modify.

Command line

To enable Multi-Attach after creation

Use the modify-volume command and specify the --multi-attach-enabled parameter.

$ aws ec2 modify-volume --volume-id vol-1234567890abcdef0 --multi-attach-enabled

Disable Multi-Attach

You can disable Multi-Attach for an io2 volume only if it is attached to no more than one instance.
Note
You can't disable Multi-Attach for io1 volumes after creation.

Use one of the following methods to disable Multi-Attach for an io2 volume.

Console

To disable Multi-Attach

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Volumes.
3. Select the volume and choose Actions, Modify Volume.
4. For Multi-Attach, clear Enable.
5. Choose Modify.

Command line

To disable Multi-Attach after creation

Use the modify-volume command and specify the –no-multi-attach-enabled parameter.

$ aws ec2 modify-volume --volume-id vol-1234567890abcdef0 -–no-multi-attach-enabled

Attach a volume to instances

You attach a Multi-Attach enabled volume to an instance in the same way that you attach any other EBS
volume. For more information, see Attach an Amazon EBS volume to an instance (p. 1106).

Delete on termination

Multi-Attach enabled volumes are deleted on instance termination if the last attached instance is
terminated and if that instance is configured to delete the volume on termination. If the volume is
attached to multiple instances that have different delete on termination settings in their volume block
device mappings, the last attached instance's block device mapping setting determines the delete on
termination behavior.

1110
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

To ensure predictable delete on termination behavior, enable or disable delete on termination for all of
the instances to which the volume is attached.

By default, when a volume is attached to an instance, the delete on termination setting for the block
device mapping is set to false. If you want to turn on delete on termination for a Multi-Attach enabled
volume, modify the block device mapping.

If you want the volume to be deleted when the attached instances are terminated, enable delete on
termination in the block device mapping for all of the attached instances. If you want to retain the
volume after the attached instances have been terminated, disable delete on termination in the block
device mapping for all of the attached instances. For more information, see Preserve Amazon EBS
volumes on instance termination (p. 646).

You can modify an instance's delete on termination setting at launch or after it has launched. If you
enable or disable delete on termination during instance launch, the settings apply only to volumes that
are attached at launch. If you attach a volume to an instance after launch, you must explicitly set the
delete on termination behavior for that volume.

You can modify an instance's delete on termination setting using the command line tools only.

To modify the delete on termination setting for an existing instance

Use the modify-instance-attribute command and specify the DeleteOnTermination attribute in the
--block-device-mappings option.

aws ec2 modify-instance-attribute --instance-id i-1234567890abcdef0 --block-device-mappings

file://mapping.json

Specify the following in mapping.json.

[
{
"DeviceName": "/dev/sdf",
"Ebs": {
"DeleteOnTermination": true|false
}
}
]

Monitor a Multi-Attach enabled volume

You can monitor a Multi-Attach enabled volume using the CloudWatch Metrics for Amazon EBS volumes.
For more information, see Amazon CloudWatch metrics for Amazon EBS (p. 1252).

Data is aggregated across all of the attached instances. You can't monitor metrics for individual attached
instances.

Pricing and billing

There are no additional charges for using Amazon EBS Multi-Attach. You are billed the standard charges
that apply to Provisioned IOPS SSD (io1 and io2) volumes. For more information, see Amazon EBS
pricing.

Make an Amazon EBS volume available for use on Linux

After you attach an Amazon EBS volume to your instance, it is exposed as a block device. You can format
the volume with any file system and then mount it. After you make the EBS volume available for use, you
can access it in the same ways that you access any other volume. Any data written to this file system is
written to the EBS volume and is transparent to applications using the device.

1111
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

You can take snapshots of your EBS volume for backup purposes or to use as a baseline when you create
another volume. For more information, see Amazon EBS snapshots (p. 1126).

You can get directions for volumes on a Windows instance from Making a Volume Available for Use on
Windows in the Amazon EC2 User Guide for Windows Instances.

Format and mount an attached volume

Suppose that you have an EC2 instance with an EBS volume for the root device, /dev/xvda, and
that you have just attached an empty EBS volume to the instance using /dev/sdf. Use the following
procedure to make the newly attached volume available for use.

To format and mount an EBS volume on Linux

1. Connect to your instance using SSH. For more information, see Connect to your Linux
instance (p. 598).
2. The device could be attached to the instance with a different device name than you specified in the
block device mapping. For more information, see Name devices on Linux instances (p. 1292). Use the
lsblk command to view your available disk devices and their mount points (if applicable) to help you
determine the correct device name to use. The output of lsblk removes the /dev/ prefix from full
device paths.

The following is example output for an instance built on the Nitro System (p. 209), which exposes
EBS volumes as NVMe block devices. The root device is /dev/nvme0n1. The attached volume is /
dev/nvme1n1, which is not yet mounted.

[ec2-user ~]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
nvme1n1 259:0 0 10G 0 disk
nvme0n1 259:1 0 8G 0 disk
-nvme0n1p1 259:2 0 8G 0 part /
-nvme0n1p128 259:3 0 1M 0 part

The following is example output for a T2 instance. The root device is /dev/xvda. The attached
volume is /dev/xvdf, which is not yet mounted.

[ec2-user ~]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
xvda 202:0 0 8G 0 disk
-xvda1 202:1 0 8G 0 part /
xvdf 202:80 0 10G 0 disk

3. Determine whether there is a file system on the volume. New volumes are raw block devices, and
you must create a file system on them before you can mount and use them. Volumes that were
created from snapshots likely have a file system on them already; if you create a new file system on
top of an existing file system, the operation overwrites your data.

Use the file -s command to get information about a device, such as its file system type. If the output
shows simply data, as in the following example output, there is no file system on the device and
you must create one.

[ec2-user ~]$ sudo file -s /dev/xvdf

/dev/xvdf: data

If the device has a file system, the command shows information about the file system type. For
example, the following output shows a root device with the XFS file system.

[ec2-user ~]$ sudo file -s /dev/xvda1

1112
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

/dev/xvda1: SGI XFS filesystem data (blksz 4096, inosz 512, v2 dirs)

4. (Conditional) If you discovered that there is a file system on the device in the previous step, skip this
step. If you have an empty volume, use the mkfs -t command to create a file system on the volume.
Warning
Do not use this command if you're mounting a volume that already has data on it (for
example, a volume that was created from a snapshot). Otherwise, you'll format the volume
and delete the existing data.

[ec2-user ~]$ sudo mkfs -t xfs /dev/xvdf

If you get an error that mkfs.xfs is not found, use the following command to install the XFS tools
and then repeat the previous command:

[ec2-user ~]$ sudo yum install xfsprogs

5. Use the mkdir command to create a mount point directory for the volume. The mount point is
where the volume is located in the file system tree and where you read and write files to after you
mount the volume. The following example creates a directory named /data.

[ec2-user ~]$ sudo mkdir /data

6. Use the following command to mount the volume at the directory you created in the previous step.

[ec2-user ~]$ sudo mount /dev/xvdf /data

7. Review the file permissions of your new volume mount to make sure that your users and
applications can write to the volume. For more information about file permissions, see File security
at The Linux Documentation Project.
8. The mount point is not automatically preserved after rebooting your instance. To automatically
mount this EBS volume after reboot, see Automatically mount an attached volume after
reboot (p. 1113).

Automatically mount an attached volume after reboot

To mount an attached EBS volume on every system reboot, add an entry for the device to the /etc/
fstab file.

You can use the device name, such as /dev/xvdf, in /etc/fstab, but we recommend using the
device's 128-bit universally unique identifier (UUID) instead. Device names can change, but the UUID
persists throughout the life of the partition. By using the UUID, you reduce the chances that the system
becomes unbootable after a hardware reconfiguration. For more information, see Identify the EBS
device (p. 1215).

To mount an attached volume automatically after reboot

1. (Optional) Create a backup of your /etc/fstab file that you can use if you accidentally destroy or
delete this file while editing it.

[ec2-user ~]$ sudo cp /etc/fstab /etc/fstab.orig

2. Use the blkid command to find the UUID of the device.

[ec2-user ~]$ sudo blkid

/dev/xvda1: LABEL="/" UUID="ca774df7-756d-4261-a3f1-76038323e572" TYPE="xfs"
PARTLABEL="Linux" PARTUUID="02dcd367-e87c-4f2e-9a72-a3cf8f299c10"

1113
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

/dev/xvdf: UUID="aebf131c-6957-451e-8d34-ec978d9581ae" TYPE="xfs"

For Ubuntu 18.04 use the lsblk command.

[ec2-user ~]$ sudo lsblk -o +UUID

3. Open the /etc/fstab file using any text editor, such as nano or vim.

[ec2-user ~]$ sudo vim /etc/fstab

4. Add the following entry to /etc/fstab to mount the device at the specified mount point. The
fields are the UUID value returned by blkid (or lsblk for Ubuntu 18.04), the mount point, the file
system, and the recommended file system mount options. For more information, see the manual
page for fstab (run man fstab).

UUID=aebf131c-6957-451e-8d34-ec978d9581ae /data xfs defaults,nofail 0 2

Note
If you ever boot your instance without this volume attached (for example, after moving the
volume to another instance), the nofail mount option enables the instance to boot even if
there are errors mounting the volume. Debian derivatives, including Ubuntu versions earlier
than 16.04, must also add the nobootwait mount option.
5. To verify that your entry works, run the following commands to unmount the device and then
mount all file systems in /etc/fstab. If there are no errors, the /etc/fstab file is OK and your
file system will mount automatically after it is rebooted.

[ec2-user ~]$ sudo umount /data

[ec2-user ~]$ sudo mount -a

If you receive an error message, address the errors in the file.

Warning
Errors in the /etc/fstab file can render a system unbootable. Do not shut down a system
that has errors in the /etc/fstab file.

If you are unsure how to correct errors in /etc/fstab and you created a backup file in the first step
of this procedure, you can restore from your backup file using the following command.

[ec2-user ~]$ sudo mv /etc/fstab.orig /etc/fstab

View information about an Amazon EBS volume

You can view descriptive information about your EBS volumes. For example, you can view information
about all volumes in a specific Region or view detailed information about a single volume, including its
size, volume type, whether the volume is encrypted, which master key was used to encrypt the volume,
and the specific instance to which the volume is attached.

You can get additional information about your EBS volumes, such as how much disk space is available,
from the operating system on the instance.

View volume information

To view information about an EBS volume using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

1114
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

2. In the navigation pane, choose Volumes.

3. (Optional) Use the filter options in the search field to display only the volumes that interest you.
For example, if you know the instance ID, choose Instance ID from the search field menu, and then
choose the instance ID from the list provided. To remove a filter, choose it again.
4. Select the volume.
5. In the details pane, you can inspect the information provided about the volume. Attachment
information shows the instance ID this volume is attached to and the device name under which it is
attached.
6. (Optional) Choose the Attachment information link to view additional details about the instance.

To view the EBS volumes that are attached to an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instance.
4. In the Storage tab, view the information provided about root and block devices.
5. (Optional) Choose a link in the Volume ID column to view additional details for the volume.

To view information about an EBS volume using the command line

You can use one of the following commands to view volume attributes. For more information, see Access
Amazon EC2 (p. 3).

• describe-volumes (AWS CLI)

• Get-EC2Volume (AWS Tools for Windows PowerShell)

Volume state

Volume state describes the availability of an Amazon EBS volume. You can view the volume state in the
State column on the Volumes page in the console, or by using the describe-volumes AWS CLI command.

The possible volume states are:

creating

The volume is being created.

available

The volume is not attached to an instance.

in-use

The volume is attached to an instance.

deleting

The volume is being deleted.

deleted

The volume is deleted.

error

The underlying hardware related to your EBS volume has failed, and the data associated with the
volume is unrecoverable. For information about how to restore the volume or recover the data on
the volume, see My EBS volume has a status of "error".

1115
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

View volume metrics

You can get additional information about your EBS volumes from Amazon CloudWatch. For more
information, see Amazon CloudWatch metrics for Amazon EBS (p. 1252).

View free disk space

You can get additional information about your EBS volumes, such as how much disk space is available,
from the Linux operating system on the instance. For example, use the following command:

[ec2-user ~]$ df -hT /dev/xvda1

Filesystem Type Size Used Avail Use% Mounted on
/dev/xvda1 xfs 8.0G 1.2G 6.9G 15% /

Replace an Amazon EBS volume using a previous snapshot

Amazon EBS snapshots are the preferred backup tool on Amazon EC2 due to their speed, convenience,
and cost. When creating a volume from a snapshot, you recreate its state at a specific point in the past
with all data intact. By attaching a volume created from a snapshot to an instance, you can duplicate
data across Regions, create test environments, replace a damaged or corrupted production volume in its
entirety, or retrieve specific files and directories and transfer them to another attached volume. For more
information, see Amazon EBS snapshots (p. 1126).

You can use the following procedure to replace an EBS volume with another volume created from a
previous snapshot of that volume. You must detach the current volume and then attach the new volume.

Note that EBS volumes can only be attached to EC2 instances in the same Availability Zone.

To replace a volume

1. Create a volume from the snapshot and write down the ID of the new volume. For more information,
see Create a volume from a snapshot (p. 1105).
2. On the volumes page, select the check box for the volume to replace. On the Description tab, find
Attachment information and write down the device name of the volume (for example, /dev/sda1
or /dev/xvda for a root volume, or /dev/sdb or xvdb) and the ID of the instance.
3. (Optional) Before you can detach the root volume of an instance, you must stop the instance. If you
are not replacing the root volume, you can continue to the next step without stopping the instance.
Otherwise, to stop the instance, from Attachment information, hover over the instance ID, right-
click, and open the instance in a new browser tab. Choose Instance state, Stop instance. Leave the
tab with the instances page open and return to the browser tab with the volumes page.
4. With the volume still selected, choose Actions, Detach Volume. When prompted for confirmation,
choose Yes, Detach. Clear the check box for this volume.
5. Select the check box for the new volume that you created in step 1. Choose Actions, Attach Volume.
Enter the instance ID and device name that you wrote down in step 2, and then choose Attach.
6. (Optional) If you stopped the instance, you must restart it. Return to the browser tab with the
instances page and choose Instance state, Start instance.
7. Connect to your instance and mount the volume. For more information, see Make an Amazon EBS
volume available for use on Linux (p. 1111).

Monitor the status of your volumes

Amazon Web Services (AWS) automatically provides data that you can use to monitor your Amazon
Elastic Block Store (Amazon EBS) volumes.

1116
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Contents
• EBS volume status checks (p. 1117)
• EBS volume events (p. 1119)
• Work with an impaired volume (p. 1120)
• Work with the Auto-Enabled IO volume attribute (p. 1122)

For additional monitoring information, see Amazon CloudWatch metrics for Amazon EBS (p. 1252) and
Amazon CloudWatch Events for Amazon EBS (p. 1258).

EBS volume status checks

Volume status checks enable you to better understand, track, and manage potential inconsistencies in
the data on an Amazon EBS volume. They are designed to provide you with the information that you
need to determine whether your Amazon EBS volumes are impaired, and to help you control how a
potentially inconsistent volume is handled.

Volume status checks are automated tests that run every 5 minutes and return a pass or fail status. If
all checks pass, the status of the volume is ok. If a check fails, the status of the volume is impaired. If
the status is insufficient-data, the checks may still be in progress on the volume. You can view the
results of volume status checks to identify any impaired volumes and take any necessary actions.

When Amazon EBS determines that a volume's data is potentially inconsistent, the default is that it
disables I/O to the volume from any attached EC2 instances, which helps to prevent data corruption.
After I/O is disabled, the next volume status check fails, and the volume status is impaired. In addition,
you'll see an event that lets you know that I/O is disabled, and that you can resolve the impaired status
of the volume by enabling I/O to the volume. We wait until you enable I/O to give you the opportunity
to decide whether to continue to let your instances use the volume, or to run a consistency check using a
command, such as fsck, before doing so.
Note
Volume status is based on the volume status checks, and does not reflect the volume state.
Therefore, volume status does not indicate volumes in the error state (for example, when
a volume is incapable of accepting I/O.) For information about volume states, see Volume
state (p. 1115).

If the consistency of a particular volume is not a concern, and you'd prefer that the volume be made
available immediately if it's impaired, you can override the default behavior by configuring the volume
to automatically enable I/O. If you enable the Auto-Enable IO volume attribute (autoEnableIO in the
API), the volume status check continues to pass. In addition, you'll see an event that lets you know that
the volume was determined to be potentially inconsistent, but that its I/O was automatically enabled.
This enables you to check the volume's consistency or replace it at a later time.

The I/O performance status check compares actual volume performance to the expected performance
of a volume and alerts you if the volume is performing below expectations. This status check is only
available for Provisioned IOPS SSD (io1 and io2) volumes that are attached to an instance. It is not
valid for General Purpose SSD (gp2 and gp3), Throughput Optimized HDD (st1), Cold HDD (sc1), or
Magnetic (standard) volumes. The I/O performance status check is performed once every minute and
CloudWatch collects this data every 5 minutes, so it might take up to 5 minutes from the moment you
attach an io1 or io2 volume to an instance for this check to report the I/O performance status.
Important
While initializing Provisioned IOPS SSD volumes that were restored from snapshots, the
performance of the volume may drop below 50 percent of its expected level, which causes the
volume to display a warning state in the I/O Performance status check. This is expected, and
you can ignore the warning state on Provisioned IOPS SSD volumes while you are initializing
them. For more information, see Initialize Amazon EBS volumes (p. 1241).

1117
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

The following table lists statuses for Amazon EBS volumes.

Volume status I/O enabled status I/O performance status (only

available for Provisioned IOPS
volumes)

ok Enabled (I/O Enabled or I/O Normal (Volume performance is

Auto-Enabled) as expected)

warning Enabled (I/O Enabled or I/O Degraded (Volume performance

Auto-Enabled) is below expectations)

Severely Degraded (Volume

performance is well below
expectations)

impaired Enabled (I/O Enabled or I/O Stalled (Volume performance is

Auto-Enabled) severely impacted)

Disabled (Volume is offline and Not Available (Unable to

pending recovery, or is waiting determine I/O performance
for the user to enable I/O) because I/O is disabled)

insufficient-data Enabled (I/O Enabled or I/O Insufficient Data

Auto-Enabled)

Insufficient Data

To view and work with status checks, you can use the Amazon EC2 console, the API, or the command line
interface.

To view status checks in the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Volumes. The Volume Status column displays the operational status
of each volume.
3. To view the status details of a volume, select the volume and choose Status Checks.

4. If you have a volume with a failed status check (status is impaired), see Work with an impaired
volume (p. 1120).

Alternatively, you can choose Events in the navigator to view all the events for your instances and
volumes. For more information, see EBS volume events (p. 1119).

To view volume status information with the command line

You can use one of the following commands to view the status of your Amazon EBS volumes. For more
information about these command line interfaces, see Access Amazon EC2 (p. 3).

1118
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

• describe-volume-status (AWS CLI)

• Get-EC2VolumeStatus (AWS Tools for Windows PowerShell)

EBS volume events

When Amazon EBS determines that a volume's data is potentially inconsistent, it disables I/O to the
volume from any attached EC2 instances by default. This causes the volume status check to fail, and
creates a volume status event that indicates the cause of the failure.

To automatically enable I/O on a volume with potential data inconsistencies, change the setting of the
Auto-Enabled IO volume attribute (autoEnableIO in the API). For more information about changing
this attribute, see Work with an impaired volume (p. 1120).

Each event includes a start time that indicates the time at which the event occurred, and a duration that
indicates how long I/O for the volume was disabled. The end time is added to the event when I/O for the
volume is enabled.

Volume status events include one of the following descriptions:

Awaiting Action: Enable IO

Volume data is potentially inconsistent. I/O is disabled for the volume until you explicitly enable it.
The event description changes to IO Enabled after you explicitly enable I/O.
IO Enabled

I/O operations were explicitly enabled for this volume.

IO Auto-Enabled

I/O operations were automatically enabled on this volume after an event occurred. We recommend
that you check for data inconsistencies before continuing to use the data.
Normal

For io1 and io2 volumes only. Volume performance is as expected.

Degraded

For io1 and io2 volumes only. Volume performance is below expectations.
Severely Degraded

For io1 and io2 volumes only. Volume performance is well below expectations.
Stalled

For io1 and io2 volumes only. Volume performance is severely impacted.

You can view events for your volumes using the Amazon EC2 console, the API, or the command line
interface.

To view events for your volumes in the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Events. All instances and volumes that have events are listed.
3. You can filter by volume to view only volume status. You can also filter on specific status types.
4. Select a volume to view its specific event.

1119
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

If you have a volume where I/O is disabled, see Work with an impaired volume (p. 1120). If you have a
volume where I/O performance is below normal, this might be a temporary condition due to an action
you have taken (for example, creating a snapshot of a volume during peak usage, running the volume on
an instance that cannot support the I/O bandwidth required, accessing data on the volume for the first
time, etc.).

To view events for your volumes with the command line

You can use one of the following commands to view event information for your Amazon EBS volumes.
For more information about these command line interfaces, see Access Amazon EC2 (p. 3).

• describe-volume-status (AWS CLI)

• Get-EC2VolumeStatus (AWS Tools for Windows PowerShell)

Work with an impaired volume

Use the following options if a volume is impaired because the volume's data is potentially inconsistent.

Options
• Option 1: Perform a consistency check on the volume attached to its instance (p. 1120)
• Option 2: Perform a consistency check on the volume using another instance (p. 1121)
• Option 3: Delete the volume if you no longer need it (p. 1122)

Option 1: Perform a consistency check on the volume attached to its instance

The simplest option is to enable I/O and then perform a data consistency check on the volume while the
volume is still attached to its Amazon EC2 instance.

To perform a consistency check on an attached volume

1. Stop any applications from using the volume.

2. Enable I/O on the volume.

a. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

b. In the navigation pane, choose Volumes.
c. Select the volume on which to enable I/O operations.

1120
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

d. In the details pane, choose Enable Volume IO, and then choose Yes, Enable.

3. Check the data on the volume.

a. Run the fsck command.

b. (Optional) Review any available application or system logs for relevant error messages.
c. If the volume has been impaired for more than 20 minutes, you can contact the AWS Support
Center. Choose Troubleshoot, and then in the Troubleshoot Status Checks dialog box, choose
Contact Support to submit a support case.

To enable I/O for a volume with the command line

You can use one of the following commands to view event information for your Amazon EBS volumes.
For more information about these command line interfaces, see Access Amazon EC2 (p. 3).

• enable-volume-io (AWS CLI)

• Enable-EC2VolumeIO (AWS Tools for Windows PowerShell)

Option 2: Perform a consistency check on the volume using another instance

Use the following procedure to check the volume outside your production environment.
Important
This procedure may cause the loss of write I/Os that were suspended when volume I/O was
disabled.

To perform a consistency check on a volume in isolation

1. Stop any applications from using the volume.

2. Detach the volume from the instance.

a. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

b. In the navigation pane, choose Volumes.
c. Select the volume to detach.
d. Choose Actions, Force Detach Volume. You'll be prompted for confirmation.
3. Enable I/O on the volume.

a. In the navigation pane, choose Volumes.

b. Select the volume that you detached in the previous step.
c. In the details pane, choose Enable Volume IO, and then choose Yes, Enable.

1121
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

4. Attach the volume to another instance. For more information, see Launch your instance (p. 525) and
Attach an Amazon EBS volume to an instance (p. 1106).
5. Check the data on the volume.

a. Run the fsck command.

b. (Optional) Review any available application or system logs for relevant error messages.
c. If the volume has been impaired for more than 20 minutes, you can contact the AWS Support
Center. Choose Troubleshoot, and then in the troubleshooting dialog box, choose Contact
Support to submit a support case.

To enable I/O for a volume with the command line

You can use one of the following commands to view event information for your Amazon EBS volumes.
For more information about these command line interfaces, see Access Amazon EC2 (p. 3).

• enable-volume-io (AWS CLI)

• Enable-EC2VolumeIO (AWS Tools for Windows PowerShell)

Option 3: Delete the volume if you no longer need it

If you want to remove the volume from your environment, simply delete it. For information about
deleting a volume, see Delete an Amazon EBS volume (p. 1126).

If you have a recent snapshot that backs up the data on the volume, you can create a new volume from
the snapshot. For more information, see Create a volume from a snapshot (p. 1105).

Work with the Auto-Enabled IO volume attribute

When Amazon EBS determines that a volume's data is potentially inconsistent, it disables I/O to the
volume from any attached EC2 instances by default. This causes the volume status check to fail, and
creates a volume status event that indicates the cause of the failure. If the consistency of a particular
volume is not a concern, and you prefer that the volume be made available immediately if it's impaired,
you can override the default behavior by configuring the volume to automatically enable I/O. If you
enable the Auto-Enabled IO volume attribute (autoEnableIO in the API), I/O between the volume and
the instance is automatically re-enabled and the volume's status check will pass. In addition, you'll see
an event that lets you know that the volume was in a potentially inconsistent state, but that its I/O was
automatically enabled. When this event occurs, you should check the volume's consistency and replace it
if necessary. For more information, see EBS volume events (p. 1119).

This procedure explains how to view and modify the Auto-Enabled IO attribute of a volume.

To view the Auto-Enabled IO attribute of a volume in the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Volumes.
3. Select the volume and choose Status Checks. Auto-Enabled IO displays the current setting
(Enabled or Disabled) for your volume.

1122
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

To modify the Auto-Enabled IO attribute of a volume in the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Volumes.
3. Select the volume and choose Actions, Change Auto-Enable IO Setting. Alternatively, choose the
Status Checks tab, and for Auto-Enabled IO, choose Edit.

4. Select the Auto-Enable Volume IO check box to automatically enable I/O for an impaired volume.
To disable the feature, clear the check box.

5. Choose Save.

To view or modify the autoEnableIO attribute of a volume with the command line

You can use one of the following commands to view the autoEnableIO attribute of your Amazon EBS
volumes. For more information about these command line interfaces, see Access Amazon EC2 (p. 3).

• describe-volume-attribute (AWS CLI)

• Get-EC2VolumeAttribute (AWS Tools for Windows PowerShell)

To modify the autoEnableIO attribute of a volume, you can use one of the commands below.

• modify-volume-attribute (AWS CLI)

• Edit-EC2VolumeAttribute (AWS Tools for Windows PowerShell)

Detach an Amazon EBS volume from a Linux instance

You need to detach an Amazon Elastic Block Store (Amazon EBS) volume from an instance before you
can attach it to a different instance or delete it. Detaching a volume does not affect the data on the
volume.

1123
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

For information about detaching volumes from a Windows instance, see Detaching a volume from a
Windows instance in the Amazon EC2 User Guide for Windows Instances.

Topics
• Considerations (p. 319)
• Unmount and detach a volume (p. 1124)
• Troubleshoot (p. 1125)

Considerations
• You can detach an Amazon EBS volume from an instance explicitly or by terminating the instance.
However, if the instance is running, you must first unmount the volume from the instance.
• If an EBS volume is the root device of an instance, you must stop the instance before you can detach
the volume.
• You can reattach a volume that you detached (without unmounting it), but it might not get the same
mount point. If there were writes to the volume in progress when it was detached, the data on the
volume might be out of sync.
• After you detach a volume, you are still charged for volume storage as long as the storage amount
exceeds the limit of the AWS Free Tier. You must delete a volume to avoid incurring further charges.
For more information, see Delete an Amazon EBS volume (p. 1126).

Unmount and detach a volume

Use the following procedures to unmount and detach a volume from an instance. This can be useful
when you need to attach the volume to a different instance or when you need to delete the volume.

Steps
• Step 1: Unmount the volume (p. 1124)
• Step 2: Detach the volume from the instance (p. 1124)

Step 1: Unmount the volume

From your Linux instance, use the following command to unmount the /dev/sdh device.

[ec2-user ~]$ umount -d /dev/sdh

Step 2: Detach the volume from the instance

To detach the volume from the instance, use one of the following methods:

Console

To detach an EBS volume using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Volumes.
3. Select a volume and choose Actions, Detach Volume.
4. When prompted for confirmation, choose Yes, Detach.

1124
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes

Command line

To detach an EBS volume from an instance using the command line

After unmounting the volume, you can use one of the following commands to detach it. For more
information about these command line interfaces, see Access Amazon EC2 (p. 3).

• detach-volume (AWS CLI)

• Dismount-EC2Volume (AWS Tools for Windows PowerShell)

Troubleshoot
The following are common problems encountered when detaching volumes, and how to resolve them.
Note
To guard against the possibility of data loss, take a snapshot of your volume before attempting
to unmount it. Forced detachment of a stuck volume can cause damage to the file system or the
data it contains or an inability to attach a new volume using the same device name, unless you
reboot the instance.

• If you encounter problems while detaching a volume through the Amazon EC2 console, it can be
helpful to use the describe-volumes CLI command to diagnose the issue. For more information, see
describe-volumes.
• If your volume stays in the detaching state, you can force the detachment by choosing Force Detach.
Use this option only as a last resort to detach a volume from a failed instance, or if you are detaching
a volume with the intention of deleting it. The instance doesn't get an opportunity to flush file system
caches or file system metadata. If you use this option, you must perform the file system check and
repair procedures.
• If you've tried to force the volume to detach multiple times over several minutes and it stays in the
detaching state, you can post a request for help to the Amazon EC2 forum. To help expedite a
resolution, include the volume ID and describe the steps that you've already taken.
• When you attempt to detach a volume that is still mounted, the volume can become stuck in the busy
state while it is trying to detach. The following output from describe-volumes shows an example of
this condition:

"Volumes": [
{
"AvailabilityZone": "us-west-2b",
"Attachments": [
{
"AttachTime": "2016-07-21T23:44:52.000Z",
"InstanceId": "i-fedc9876",
"VolumeId": "vol-1234abcd",
"State": "busy",
"DeleteOnTermination": false,
"Device": "/dev/sdf"
}
...
}
]

When you encounter this state, detachment can be delayed indefinitely until you unmount the volume,
force detachment, reboot the instance, or all three.

1125
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

Delete an Amazon EBS volume

After you no longer need an Amazon EBS volume, you can delete it. After deletion, its data is gone and
the volume can't be attached to any instance. However, before deletion, you can store a snapshot of the
volume, which you can use to re-create the volume later.
Note
You can't delete a volume if it's attached to an instance. To delete a volume, you must
first detach it. For more information, see Detach an Amazon EBS volume from a Linux
instance (p. 1123).
You can check if a volume is attached to an instance. In the console, on the Volumes page, you
can view the state of your volumes.

• If a volume is attached to an instance, it’s in the in-use state.

• If a volume is detached from an instance, it’s in the available state. You can delete this
volume.

To delete an EBS volume using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Volumes.
3. Select a volume and choose Actions, Delete Volume. If Delete Volume is greyed out, the volume is
attached to an instance.
4. In the confirmation dialog box, choose Yes, Delete.

To delete an EBS volume using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• delete-volume (AWS CLI)

• Remove-EC2Volume (AWS Tools for Windows PowerShell)

Amazon EBS snapshots

You can back up the data on your Amazon EBS volumes to Amazon S3 by taking point-in-time snapshots.
Snapshots are incremental backups, which means that only the blocks on the device that have changed
after your most recent snapshot are saved. This minimizes the time required to create the snapshot and
saves on storage costs by not duplicating data. Each snapshot contains all of the information that is
needed to restore your data (from the moment when the snapshot was taken) to a new EBS volume.

When you create an EBS volume based on a snapshot, the new volume begins as an exact replica of
the original volume that was used to create the snapshot. The replicated volume loads data in the
background so that you can begin using it immediately. If you access data that hasn't been loaded
yet, the volume immediately downloads the requested data from Amazon S3, and then continues
loading the rest of the volume's data in the background. For more information, see Create Amazon EBS
snapshots (p. 1129).

When you delete a snapshot, only the data unique to that snapshot is removed. For more information,
see Delete an Amazon EBS snapshot (p. 1132).

Snapshot events

You can track the status of your EBS snapshots through CloudWatch Events. For more information, see
EBS snapshot events (p. 1261).

1126
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

Multi-volume snapshots

Snapshots can be used to create a backup of critical workloads, such as a large database or a file system
that spans across multiple EBS volumes. Multi-volume snapshots allow you to take exact point-in-
time, data coordinated, and crash-consistent snapshots across multiple EBS volumes attached to an
EC2 instance. You are no longer required to stop your instance or to coordinate between volumes to
ensure crash consistency, because snapshots are automatically taken across multiple EBS volumes. For
more information, see the steps for creating a multi-volume EBS snapshot under Create Amazon EBS
snapshots (p. 1129) .

Snapshot pricing

Charges for your snapshots are based on the amount of data stored. Because snapshots are incremental,
deleting a snapshot might not reduce your data storage costs. Data referenced exclusively by a snapshot
is removed when that snapshot is deleted, but data referenced by other snapshots is preserved. For
more information, see Amazon Elastic Block Store Volumes and Snapshots in the AWS Billing and Cost
Management User Guide.

Contents
• How incremental snapshots work (p. 1127)
• Copy and share snapshots (p. 1128)
• Encryption support for snapshots (p. 1129)
• Create Amazon EBS snapshots (p. 1129)
• Delete an Amazon EBS snapshot (p. 1132)
• Copy an Amazon EBS snapshot (p. 1134)
• View Amazon EBS snapshot information (p. 1138)
• Share an Amazon EBS snapshot (p. 1139)
• Use EBS direct APIs to access the contents of an EBS snapshot (p. 1142)
• Automate the snapshot lifecycle (p. 1163)

How incremental snapshots work

This section provides illustrations of how an EBS snapshot captures the state of a volume at a point in
time, and also how successive snapshots of a changing volume create a history of those changes.

In the diagram below, Volume 1 is shown at three points in time. A snapshot is taken of each of these
three volume states.

• In State 1, the volume has 10 GiB of data. Because Snap A is the first snapshot taken of the volume,
the entire 10 GiB of data must be copied.
• In State 2, the volume still contains 10 GiB of data, but 4 GiB have changed. Snap B needs to copy and
store only the 4 GiB that changed after Snap A was taken. The other 6 GiB of unchanged data, which
are already copied and stored in Snap A, are referenced by Snap B rather than (again) copied. This is
indicated by the dashed arrow.
• In State 3, 2 GiB of data have been added to the volume, for a total of 12 GiB. Snap C needs to copy
the 2 GiB that were added after Snap B was taken. As shown by the dashed arrows, Snap C also
references 4 GiB of data stored in Snap B, and 6 GiB of data stored in Snap A.
• The total storage required for the three snapshots is 16 GiB.

Relations among multiple snapshots of a volume

1127
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

Note
If you copy a snapshot and encrypt it to a new CMK, a complete (non-incremental) copy is
always created, resulting in additional delay and storage costs.

For more information about how data is managed when you delete a snapshot, see Delete an Amazon
EBS snapshot (p. 1132).

Copy and share snapshots

You can share a snapshot across AWS accounts by modifying its access permissions. You can make copies
of your own snapshots as well as snapshots that have been shared with you. For more information, see
Share an Amazon EBS snapshot (p. 1139).

A snapshot is constrained to the AWS Region where it was created. After you create a snapshot of an EBS
volume, you can use it to create new volumes in the same Region. For more information, see Create a
volume from a snapshot (p. 1105). You can also copy snapshots across Regions, making it possible to use
multiple Regions for geographical expansion, data center migration, and disaster recovery. You can copy

1128
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

any accessible snapshot that has a completed status. For more information, see Copy an Amazon EBS
snapshot (p. 1134).

Encryption support for snapshots

EBS snapshots fully support EBS encryption.

• Snapshots of encrypted volumes are automatically encrypted.

• Volumes that you create from encrypted snapshots are automatically encrypted.
• Volumes that you create from an unencrypted snapshot that you own or have access to can be
encrypted on-the-fly.
• When you copy an unencrypted snapshot that you own, you can encrypt it during the copy process.
• When you copy an encrypted snapshot that you own or have access to, you can reencrypt it with a
different key during the copy process.
• The first snapshot you take of an encrypted volume that has been created from an unencrypted
snapshot is always a full snapshot.
• The first snapshot you take of a reencrypted volume, which has a different CMK compared to the
source snapshot, is always a full snapshot.

Note
If you copy a snapshot and encrypt it to a new CMK, a complete (non-incremental) copy is
always created, resulting in additional delay and storage costs.

Complete documentation of possible snapshot encryption scenarios is provided in Create Amazon EBS
snapshots (p. 1129) and in Copy an Amazon EBS snapshot (p. 1134).

For more information, see Amazon EBS encryption (p. 1176).

Create Amazon EBS snapshots

You can create a point-in-time snapshot of an EBS volume and use it as a baseline for new volumes or
for data backup. If you make periodic snapshots of a volume, the snapshots are incremental—the new
snapshot saves only the blocks that have changed since your last snapshot.

Snapshots occur asynchronously; the point-in-time snapshot is created immediately, but the status of
the snapshot is pending until the snapshot is complete (when all of the modified blocks have been
transferred to Amazon S3), which can take several hours for large initial snapshots or subsequent
snapshots where many blocks have changed. While it is completing, an in-progress snapshot is not
affected by ongoing reads and writes to the volume.

You can take a snapshot of an attached volume that is in use. However, snapshots only capture data
that has been written to your Amazon EBS volume at the time the snapshot command is issued. This
might exclude any data that has been cached by any applications or the operating system. If you can
pause any file writes to the volume long enough to take a snapshot, your snapshot should be complete.
However, if you can't pause all file writes to the volume, you should unmount the volume from within
the instance, issue the snapshot command, and then remount the volume to ensure a consistent and
complete snapshot. You can remount and use your volume while the snapshot status is pending.

To make snapshot management easier, you can tag your snapshots during creation or add tags
afterward. For example, you can apply tags describing the original volume from which the snapshot
was created, or the device name that was used to attach the original volume to an instance. For more
information, see Tag your Amazon EC2 resources (p. 1311).

Snapshot encryption
Snapshots that are taken from encrypted volumes are automatically encrypted. Volumes that are created
from encrypted snapshots are also automatically encrypted. The data in your encrypted volumes and

1129
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

any associated snapshots is protected both at rest and in motion. For more information, see Amazon EBS
encryption (p. 1176).

By default, only you can create volumes from snapshots that you own. However, you can share your
unencrypted snapshots with specific AWS accounts, or you can share them with the entire AWS
community by making them public. For more information, see Share an Amazon EBS snapshot (p. 1139).

You can share an encrypted snapshot only with specific AWS accounts. For others to use your shared,
encrypted snapshot, you must also share the CMK key that was used to encrypt it. Users with access to
your encrypted snapshot must create their own personal copy of it and then use that copy. Your copy of
a shared, encrypted snapshot can also be re-encrypted using a different key. For more information, see
Share an Amazon EBS snapshot (p. 1139).
Note
If you copy a snapshot and encrypt it to a new CMK, a complete (non-incremental) copy is
always created, resulting in additional delay and storage costs.

Multi-volume snapshots
You can create multi-volume snapshots, which are point-in-time snapshots for all EBS volumes attached
to an EC2 instance. You can also create lifecycle policies to automate the creation and retention of multi-
volume snapshots. For more information, see Amazon Data Lifecycle Manager (p. 1186).

After the snapshots are created, each snapshot is treated as an individual snapshot. You can perform all
snapshot operations, such as restore, delete, and copy across Regions or accounts, just as you would with
a single volume snapshot. You can also tag your multi-volume snapshots as you would a single volume
snapshot. We recommend you tag your multiple volume snapshots to manage them collectively during
restore, copy, or retention.

Multi-volume, crash-consistent snapshots are typically restored as a set. It is helpful to identify

the snapshots that are in a crash-consistent set by tagging your set with the instance ID, name, or
other relevant details. You can also choose to automatically copy tags from the source volume to
the corresponding snapshots. This helps you to set the snapshot metadata, such as access policies,
attachment information, and cost allocation, to match the source volume.

After creating your snapshots, they appear in your EC2 console created at the exact point-in-time. The
snapshots are collectively managed and, therefore, if any one snapshot for the volume set fails, all of the
other snapshots display an error status.

Amazon Data Lifecycle Manager

You can create, retain, and delete snapshots manually, or you can use Amazon Data Lifecycle Manager to
manage your snapshots for you. For more information, see Data Lifecycle Manager (p. 1186).

Considerations
The following considerations apply to creating snapshots:

• When you create a snapshot for an EBS volume that serves as a root device, you should stop the
instance before taking the snapshot.
• You cannot create snapshots from instances for which hibernation is enabled.
• You cannot create snapshots from hibernated instances.
• Although you can take a snapshot of a volume while a previous snapshot of that volume is in the
pending status, having multiple pending snapshots of a volume can result in reduced volume
performance until the snapshots complete.
• There is a limit of one pending snapshot for a single st1 or sc1 volume, or five
pending snapshots for a single volume of the other volume types. If you receive a
ConcurrentSnapshotLimitExceeded error while trying to create multiple concurrent snapshots of

1130
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

the same volume, wait for one or more of the pending snapshots to complete before creating another
snapshot of that volume.
• When a snapshot is created from a volume with an AWS Marketplace product code, the product code is
propagated to the snapshot.

Create a snapshot
Use the following procedure to create a snapshot from the specified volume.

To create a snapshot using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Snapshots under Elastic Block Store in the navigation pane.
3. Choose Create Snapshot.
4. For Select resource type, choose Volume.
5. For Volume, select the volume.
6. (Optional) Enter a description for the snapshot.
7. (Optional) Choose Add Tag to add tags to your snapshot. For each tag, provide a tag key and a tag
value.
8. Choose Create Snapshot.

To create a snapshot using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• create-snapshot (AWS CLI)

• New-EC2Snapshot (AWS Tools for Windows PowerShell)

Create a multi-volume snapshot

Use the following procedure to create a snapshot from the volumes of an instance.

To create multi-volume snapshots using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Snapshots under Elastic Block Store in the navigation pane.
3. Choose Create Snapshot.
4. For Select resource type, choose Instance.
5. Select the instance ID for which you want to create simultaneous backups for all of the attached EBS
volumes. Multi-volume snapshots support up to 40 EBS volumes per instance.
6. (Optional) Set Exclude root volume.
7. (Optional) Set Copy tags from volume flag to automatically copy tags from the source volume to
the corresponding snapshots. This sets snapshot metadata—such as access policies, attachment
information, and cost allocation—to match the source volume.
8. (Optional) Choose Add Tag to add tags to your snapshot. For each tag, provide a tag key and a tag
value.
9. Choose Create Snapshot.

During snapshot creation, the snapshots are managed together. If one of the snapshots in the
volume set fails, the other snapshots are moved to error status for the volume set. You can

1131
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

monitor the progress of your snapshots using CloudWatch Events. After the snapshot creation
process completes, CloudWatch generates an event that contains the status and all of the relevant
snapshots details for the affected instance.

To create multi-volume snapshots using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• create-snapshots (AWS CLI)

• New-EC2SnapshotBatch (AWS Tools for Windows PowerShell)

Work with EBS snapshots

You can copy snapshots, share snapshots, and create volumes from snapshots. For more information, see
the following:

• Copy an Amazon EBS snapshot (p. 1134)

• Share an Amazon EBS snapshot (p. 1139)
• Create a volume from a snapshot (p. 1105)

Delete an Amazon EBS snapshot

After you no longer need an Amazon EBS snapshot of a volume, you can delete it. Deleting a snapshot
has no effect on the volume. Deleting a volume has no effect on the snapshots made from it.

Incremental snapshot deletion

If you make periodic snapshots of a volume, the snapshots are incremental. This means that only the
blocks on the device that have changed after your most recent snapshot are saved in the new snapshot.
Even though snapshots are saved incrementally, the snapshot deletion process is designed so that you
need to retain only the most recent snapshot in order to create volumes.

If data was present on a volume held in an earlier snapshot or series of snapshots, and that data is
subsequently deleted from the volume later on, the data is still considered to be unique data of the
earlier snapshots. Unique data is only deleted from the sequence of snapshots if all snapshots that
reference the unique data are deleted.

When you delete a snapshot, only the data that is referenced exclusively by that snapshot is removed.
Unique data is only deleted if all of the snapshots that reference it are deleted. Deleting previous
snapshots of a volume does not affect your ability to create volumes from later snapshots of that
volume.

Deleting a snapshot might not reduce your organization's data storage costs. Other snapshots might
reference that snapshot's data, and referenced data is always preserved. If you delete a snapshot
containing data being used by a later snapshot, costs associated with the referenced data are allocated
to the later snapshot. For more information about how snapshots store data, see How incremental
snapshots work (p. 1127) and the following example.

In the following diagram, Volume 1 is shown at three points in time. A snapshot has captured each of the
first two states, and in the third, a snapshot has been deleted.

• In State 1, the volume has 10 GiB of data. Because Snap A is the first snapshot taken of the volume,
the entire 10 GiB of data must be copied.

1132
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

• In State 2, the volume still contains 10 GiB of data, but 4 GiB have changed. Snap B needs to copy and
store only the 4 GiB that changed after Snap A was taken. The other 6 GiB of unchanged data, which
are already copied and stored in Snap A, are referenced by Snap B rather than (again) copied. This is
indicated by the dashed arrow.
• In state 3, the volume has not changed since State 2, but Snapshot A has been deleted. The 6 GiB of
data stored in Snapshot A that were referenced by Snapshot B have now been moved to Snapshot
B, as shown by the heavy arrow. As a result, you are still charged for storing 10 GiB of data; 6 GiB of
unchanged data preserved from Snap A and 4 GiB of changed data from Snap B.

Deleting a snapshot with some of its data referenced by another snapshot

Considerations
The following considerations apply to deleting snapshots:

• You can't delete a snapshot of the root device of an EBS volume used by a registered AMI. You must
first deregister the AMI before you can delete the snapshot. For more information, see Deregister your
Linux AMI (p. 175).
• You can't delete a snapshot that is managed by the AWS Backup service using Amazon EC2. Instead,
use AWS Backup to delete the corresponding recovery points in the backup vault.
• You can create, retain, and delete snapshots manually, or you can use Amazon Data Lifecycle Manager
to manage your snapshots for you. For more information, see Data Lifecycle Manager (p. 1186).
• Although you can delete a snapshot that is still in progress, the snapshot must complete before the
deletion takes effect. This might take a long time. If you are also at your concurrent snapshot limit, and

1133
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

you attempt to take an additional snapshot, you might get a ConcurrentSnapshotLimitExceeded

error. For more information, see the Service Quotas for Amazon EBS in the Amazon Web Services
General Reference.

Delete a snapshot
Use the following procedure to delete a snapshot.

To delete a snapshot using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Snapshots in the navigation pane.
3. Select a snapshot and then choose Delete from the Actions list.
4. Choose Yes, Delete.

To delete a snapshot using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• delete-snapshot (AWS CLI)

• Remove-EC2Snapshot (AWS Tools for Windows PowerShell)

Delete a multi-volume snapshot

To delete multi-volume snapshots, retrieve all of the snapshots for your multi-volume group using the
tag you applied to the group when you created the snapshots. Then, delete the snapshots individually.
You will not be prevented from deleting individual snapshots in the multi-volume snapshots group.

Copy an Amazon EBS snapshot

With Amazon EBS, you can create point-in-time snapshots of volumes, which we store for you in Amazon
S3. After you create a snapshot and it has finished copying to Amazon S3 (when the snapshot status is
completed), you can copy it from one AWS Region to another, or within the same Region. Amazon S3
server-side encryption (256-bit AES) protects a snapshot's data in transit during a copy operation. The
snapshot copy receives an ID that is different from the ID of the original snapshot.

To copy multi-volume snapshots to another AWS Region, retrieve the snapshots using the tag you
applied to the multi-volume snapshots group when you created it. Then individually copy the snapshots
to another Region.

For information about copying an Amazon RDS snapshot, see Copying a DB Snapshot in the Amazon RDS
User Guide.

If you would like another account to be able to copy your snapshot, you must either modify the snapshot
permissions to allow access to that account or make the snapshot public so that all AWS accounts can
copy it. For more information, see Share an Amazon EBS snapshot (p. 1139).

For pricing information about copying snapshots across AWS Regions and accounts, see Amazon EBS
Pricing.
Note
Snapshot copy operations within a single account and Region do not copy any actual data and
therefore are cost-free as long as the encryption status of the snapshot copy does not change.

1134
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

Note
If you copy a snapshot to a new Region, a complete (non-incremental) copy is always created,
resulting in additional delay and storage costs.
Note
If you copy a snapshot and encrypt it to a new CMK, a complete (non-incremental) copy is
always created, resulting in additional delay and storage costs.

Use cases

• Geographic expansion: Launch your applications in a new AWS Region.

• Migration: Move an application to a new Region, to enable better availability and to minimize cost.
• Disaster recovery: Back up your data and logs across different geographical locations at regular
intervals. In case of disaster, you can restore your applications using point-in-time backups stored in
the secondary Region. This minimizes data loss and recovery time.
• Encryption: Encrypt a previously unencrypted snapshot, change the key with which the snapshot is
encrypted, or create a copy that you own in order to create a volume from it (for encrypted snapshots
that have been shared with you).
• Data retention and auditing requirements: Copy your encrypted EBS snapshots from one AWS
account to another to preserve data logs or other files for auditing or data retention. Using a different
account helps prevent accidental snapshot deletions, and protects you if your main AWS account is
compromised.

Prerequisites

• You can copy any accessible snapshots that have a completed status, including shared snapshots and
snapshots that you have created.
• You can copy AWS Marketplace, VM Import/Export, and AWS Storage Gateway snapshots, but you
must verify that the snapshot is supported in the destination Region.

Limits

• Each account can have up to twenty concurrent snapshot copy requests to a single destination Region.
• User-defined tags are not copied from the source snapshot to the new snapshot. You can add user-
defined tags during or after the copy operation. For more information, see Tag your Amazon EC2
resources (p. 1311).
• Snapshots created by the CopySnapshot action have an arbitrary volume ID that should not be used
for any purpose.

Incremental snapshot copying

Whether a snapshot copy is incremental is determined by the most recently completed snapshot copy.
When you copy a snapshot across Regions or accounts, the copy is an incremental copy if the following
conditions are met:

• The snapshot was copied to the destination Region or account previously.

• The most recent snapshot copy still exists in the destination Region or account.
• All copies of the snapshot in the destination Region or account are either unencrypted or were
encrypted using the same CMK.

If the most recent snapshot copy was deleted, the next copy is a full copy, not an incremental copy. If
a copy is still pending when you start a another copy, the second copy starts only after the first copy
finishes.

1135
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

We recommend that you tag your snapshots with the volume ID and creation time so that you can keep
track of the most recent snapshot copy of a volume in the destination Region or account.

To see whether your snapshot copies are incremental, check the copySnapshot (p. 1263) CloudWatch
event.

Encryption and snapshot copying

When you copy a snapshot, you can encrypt the copy or you can specify a CMK different from the
original one, and the resulting copied snapshot uses the new CMK. However, changing the encryption
status of a snapshot during a copy operation results in a full (not incremental) copy, which might incur
greater data transfer and storage charges.

To copy an encrypted snapshot shared from another AWS account, you must have permissions to use
the snapshot and the customer master key (CMK) that was used to encrypt the snapshot. When using
an encrypted snapshot that was shared with you, we recommend that you re-encrypt the snapshot by
copying it using a CMK that you own. This protects you if the original CMK is compromised, or if the
owner revokes it, which could cause you to lose access to any encrypted volumes that you created using
the snapshot. For more information, see Share an Amazon EBS snapshot (p. 1139).

You apply encryption to EBS snapshot copies by setting the Encrypted parameter to true. (The
Encrypted parameter is optional if encryption by default (p. 1178) is enabled).

Optionally, you can use KmsKeyId to specify a custom key to use to encrypt the snapshot copy. (The
Encrypted parameter must also be set to true, even if encryption by default is enabled.) If KmsKeyId
is not specified, the key that is used for encryption depends on the encryption state of the source
snapshot and its ownership.

The following table describes the encryption outcome for each possible combination of settings.

Encryption outcomes: Copying a snapshot

Is Encrypted Is encryption Source snapshot Default (no KmsKeyId Custom (KmsKeyId
parameter by default specified) specified)
set? set?

No No Unencrypted Unencrypted N/A

snapshot that you
own

No No Encrypted snapshot Encrypted by same key

that you own

No No Unencrypted Unencrypted
snapshot that is
shared with you

No No Encrypted snapshot Encrypted by default

that is shared with CMK*
you

Yes No Unencrypted Encrypted by default Encrypted by a

snapshot that you CMK specified CMK**
own

Yes No Encrypted snapshot Encrypted by same key

that you own

Yes No Unencrypted Encrypted by default

snapshot that is CMK
shared with you

1136
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

Is Encrypted Is encryption Source snapshot Default (no KmsKeyId Custom (KmsKeyId

parameter by default specified) specified)
set? set?

Yes No Encrypted snapshot Encrypted by default

that is shared with CMK
you

No Yes Unencrypted Encrypted by default N/A

snapshot that you CMK
own

No Yes Encrypted snapshot Encrypted by same key

that you own

No Yes Unencrypted Encrypted by default

snapshot that is CMK
shared with you

No Yes Encrypted snapshot Encrypted by default

that is shared with CMK
you

Yes Yes Unencrypted Encrypted by default Encrypted by a

snapshot that you CMK specified CMK
own

Yes Yes Encrypted snapshot Encrypted by same key

that you own

Yes Yes Unencrypted Encrypted by default

snapshot that is CMK
shared with you

Yes Yes Encrypted snapshot Encrypted by default

that is shared with CMK
you

* This is the default CMK used for EBS encryption for the AWS account and Region. By default this is a
unique AWS managed CMK for EBS, or you can specify a customer managed CMK. For more information,
see Default key for EBS encryption (p. 1178).

** This is a customer managed CMK specified for the copy action. This CMK is used instead of the default
CMK for the AWS account and Region.

Copy a snapshot
Use the following procedure to copy a snapshot using the Amazon EC2 console.

To copy a snapshot using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Snapshots.
3. Select the snapshot to copy, and then choose Copy from the Actions list.
4. In the Copy Snapshot dialog box, update the following as necessary:

• Destination region: Select the Region where you want to write the copy of the snapshot.

1137
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

• Description: By default, the description includes information about the source snapshot so that
you can identify a copy from the original. You can change this description as necessary.
• Encryption: If the source snapshot is not encrypted, you can choose to encrypt the copy. If you
have enabled encryption by default (p. 1178), the Encryption option is set and cannot be unset
from the snapshot console. If the Encryption option is set, you can choose to encrypt it to a
customer managed CMK by selecting one in the field, described below.

You cannot strip encryption from an encrypted snapshot.

Note
If you copy a snapshot and encrypt it to a new CMK, a complete (non-incremental) copy is
always created, resulting in additional delay and storage costs.
• Master Key: The customer master key (CMK) to be used to encrypt this snapshot. The default
key for your account is displayed initially, but you can optionally select from the master keys in
your account or type/paste the ARN of a key from a different account. You can create new master
encryption keys in the IAM console https://console.aws.amazon.com/iam/.
5. Choose Copy.
6. In the Copy Snapshot confirmation dialog box, choose Snapshots to go to the Snapshots page in
the Region specified, or choose Close.

To view the progress of the copy process, switch to the destination Region, and then refresh the
Snapshots page. Copies in progress are listed at the top of the page.

To check for failure

If you attempt to copy an encrypted snapshot without having permissions to use the encryption key, the
operation fails silently. The error state is not displayed in the console until you refresh the page. You can
also check the state of the snapshot from the command line, as in the following example.

aws ec2 describe-snapshots --snapshot-id snap-0123abcd

If the copy failed because of insufficient key permissions, you see the following message:
"StateMessage": "Given key ID is not accessible".

When copying an encrypted snapshot, you must have DescribeKey permissions on the default CMK.
Explicitly denying these permissions results in copy failure. For information about managing CMK keys,
see Controlling Access to Customer Master Keys.

To copy a snapshot using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• copy-snapshot (AWS CLI)

• Copy-EC2Snapshot (AWS Tools for Windows PowerShell)

View Amazon EBS snapshot information

You can view detailed information about your snapshots.

To view snapshot information using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Snapshots in the navigation pane.

1138
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

3. To reduce the list, choose an option from the Filter list. For example, to view only your snapshots,
choose Owned By Me. You can also filter your snapshots using tags and snapshot attributes. Choose
the search bar to view the available tags and attributes.
4. To view more information about a snapshot, select it.

To view snapshot information using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• describe-snapshots (AWS CLI)

• Get-EC2Snapshot (AWS Tools for Windows PowerShell)

Example Example: Filter based on tags

The following command describes the snapshots with the tag Stack=production.

aws ec2 describe-snapshots --filters Name=tag:Stack,Values=production

Example Example: Filter based on volume

The following command describes the snapshots created from the specified volume.

aws ec2 describe-snapshots --filters Name=volume-id,Values=vol-049df61146c4d7901

Example Example: Filter based on snapshot age

With the AWS CLI, you can use JMESPath to filter results using expressions. For example, the
following command displays the IDs of all snapshots created by your AWS account (represented by
123456789012) before the specified date (represented by 2020-03-31). If you do not specify the
owner, the results include all public snapshots.

aws ec2 describe-snapshots --filters Name=owner-id,Values=123456789012 --query "Snapshots[?

(StartTime<=`2020-03-31`)].[SnapshotId]" --output text

The following command displays the IDs of all snapshots created in the specified date range.

aws ec2 describe-snapshots --filters Name=owner-id,Values=123456789012 --query "Snapshots[?

(StartTime>=`2019-01-01`) && (StartTime<=`2019-12-31`)].[SnapshotId]" --output text

Share an Amazon EBS snapshot

By modifying the permissions of a snapshot, you can share it with the AWS accounts that you specify.
Users that you have authorized can use the snapshots you share as the basis for creating their own EBS
volumes, while your original snapshot remains unaffected.

If you choose, you can make your unencrypted snapshots available publicly to all AWS users. You can't
make your encrypted snapshots available publicly.

When you share an encrypted snapshot, you must also share the customer managed CMK used to
encrypt the snapshot. You can apply cross-account permissions to a customer managed CMK either when
it is created or at a later time.
Important
When you share a snapshot, you are giving others access to all of the data on the snapshot.
Share snapshots only with people with whom you want to share all of your snapshot data.

1139
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

Considerations
The following considerations apply to sharing snapshots:

• Snapshots are constrained to the Region in which they were created. To share a snapshot with
another Region, copy the snapshot to that Region. For more information, see Copy an Amazon EBS
snapshot (p. 1134).
• AWS prevents you from sharing snapshots that were encrypted with your default CMK. Snapshots that
you intend to share must instead be encrypted with a customer managed CMK. For more information,
see Creating Keys in the AWS Key Management Service Developer Guide.
• Users of your shared CMK who are accessing encrypted snapshots must be granted permissions
to perform the following actions on the key: kms:DescribeKey, kms:CreateGrant,
GenerateDataKey, and kms:ReEncrypt. For more information, see Controlling Access to Customer
Master Keys in the AWS Key Management Service Developer Guide.

Share an unencrypted snapshot using the console

To share a snapshot using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Snapshots in the navigation pane.
3. Select the snapshot and then choose Actions, Modify Permissions.
4. Make the snapshot public or share it with specific AWS accounts as follows:
• To make the snapshot public, choose Public.

This option is not valid for encrypted snapshots or snapshots with an AWS Marketplace product
code.
• To share the snapshot with one or more AWS accounts, choose Private, enter the AWS account
ID (without hyphens) in AWS Account Number, and choose Add Permission. Repeat for any
additional AWS accounts.
5. Choose Save.

To use an unencrypted snapshot that was privately shared with you

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Snapshots in the navigation pane.
3. Choose the Private Snapshots filter.
4. Locate the snapshot by ID or description. You can use this snapshot as you would any other; for
example, you can create a volume from the snapshot or copy the snapshot to a different Region.

Share an encrypted snapshot using the console

To share an encrypted snapshot using the console

1. Open the AWS KMS console at https://console.aws.amazon.com/kms.

2. To change the AWS Region, use the Region selector in the upper-right corner of the page.
3. Choose Customer managed keys in the navigation pane.
4. In the Alias column, choose the alias (text link) of the customer managed key that you used to
encrypt the snapshot. The key details open in a new page.
5. In the Key policy section, you see either the policy view or the default view. The policy view displays
the key policy document. The default view displays sections for Key administrators, Key deletion,
Key Use, and Other AWS accounts. The default view displays if you created the policy in the console

1140
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

and have not customized it. If the default view is not available, you'll need to manually edit the
policy in the policy view. For more information, see Viewing a Key Policy (Console) in the AWS Key
Management Service Developer Guide.

Use either the policy view or the default view, depending on which view you can access, to add one
or more AWS account IDs to the policy, as follows:
• (Policy view) Choose Edit. Add one or more AWS account IDs to the following statements:
"Allow use of the key" and "Allow attachment of persistent resources".
Choose Save changes. In the following example, the AWS account ID 444455556666 is added
to the policy.

{
"Sid": "Allow use of the key",
"Effect": "Allow",
"Principal": {"AWS": [
"arn:aws:iam::111122223333:user/CMKUser",
"arn:aws:iam::444455556666:root"
]},
"Action": [
"kms:Encrypt",
"kms:Decrypt",
"kms:ReEncrypt*",
"kms:GenerateDataKey*",
"kms:DescribeKey"
],
"Resource": "*"
},
{
"Sid": "Allow attachment of persistent resources",
"Effect": "Allow",
"Principal": {"AWS": [
"arn:aws:iam::111122223333:user/CMKUser",
"arn:aws:iam::444455556666:root"
]},
"Action": [
"kms:CreateGrant",
"kms:ListGrants",
"kms:RevokeGrant"
],
"Resource": "*",
"Condition": {"Bool": {"kms:GrantIsForAWSResource": true}}
}

• (Default view) Scroll down to Other AWS accounts. Choose Add other AWS accounts and enter
the AWS account ID as prompted. To add another account, choose Add another AWS account
and enter the AWS account ID. When you have added all AWS accounts, choose Save changes.
6. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.
7. Choose Snapshots in the navigation pane.
8. Select the snapshot and then choose Actions, Modify Permissions.
9. For each AWS account, enter the AWS account ID in AWS Account Number and choose Add
Permission. When you have added all AWS accounts, choose Save.

To use an encrypted snapshot that was shared with you

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Snapshots in the navigation pane.
3. Choose the Private Snapshots filter. Optionally add the Encrypted filter.
4. Locate the snapshot by ID or description.

1141
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

5. Select the snapshot and choose Actions, Copy.

6. (Optional) Select a destination Region.
7. The copy of the snapshot is encrypted by the key displayed in Master Key. By default, the selected
key is your account's default CMK. To select a customer managed CMK, click inside the input box to
see a list of available keys.
8. Choose Copy.

Share a snapshot using the command line

The permissions for a snapshot are specified using the createVolumePermission attribute of the
snapshot. To make a snapshot public, set the group to all. To share a snapshot with a specific AWS
account, set the user to the ID of the AWS account.

To modify snapshot permissions using the command line

Use one of the following commands:

• modify-snapshot-attribute (AWS CLI)

• Edit-EC2SnapshotAttribute (AWS Tools for Windows PowerShell)

To view snapshot permissions using the command line

Use one of the following commands:

• describe-snapshot-attribute (AWS CLI)

• Get-EC2SnapshotAttribute (AWS Tools for Windows PowerShell)

For more information about these command line interfaces, see Access Amazon EC2 (p. 3).

Determine the use of shared snapshots

You can use AWS CloudTrail to monitor whether a snapshot that you have shared with others is copied or
used to create a volume. The following events are logged in CloudTrail:

• SharedSnapshotCopyInitiated — A shared snapshot is being copied.

• SharedSnapshotVolumeCreated — A shared snapshot is being used to create a volume.

For more information about using CloudTrail, see Log Amazon EC2 and Amazon EBS API calls with AWS
CloudTrail (p. 805).

Use EBS direct APIs to access the contents of an EBS snapshot

You can use the Amazon Elastic Block Store (Amazon EBS) direct APIs to create EBS snapshots, write
data directly to your snapshots, read data on your snapshots, and identify the differences or changes
between two snapshots. If you’re an independent software vendor (ISV) who offers backup services for
Amazon EBS, the EBS direct APIs make it more efficient and cost-effective to track incremental changes
on your EBS volumes through snapshots. This can be done without having to create new volumes
from snapshots, and then use Amazon Elastic Compute Cloud (Amazon EC2) instances to compare the
differences.

You can create incremental snapshots directly from data on-premises into EBS volumes and the cloud
to use for quick disaster recovery. With the ability to write and read snapshots, you can write your on-
premises data to an EBS snapshot during a disaster. Then after recovery, you can restore it back to AWS
or on-premises from the snapshot. You no longer need to build and maintain complex mechanisms to
copy data to and from Amazon EBS.

1142
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

This user guide provides an overview of the elements that make up the EBS direct APIs, and examples of
how to use them effectively. For more information about the actions, data types, parameters, and errors
of the APIs, see the EBS direct APIs reference. For more information about the supported AWS Regions,
endpoints, and service quotas for the EBS direct APIs, see Amazon EBS Endpoints and Quotas in the AWS
General Reference.

Contents
• Understand the EBS direct APIs (p. 1143)
• Permissions for IAM users (p. 1146)
• Use encryption (p. 1150)
• Use Signature Version 4 signing (p. 1150)
• Use checksums (p. 1150)
• Work with the EBS direct APIs using the API or AWS SDKs (p. 1151)
• Work with the EBS direct APIs using the command line (p. 1155)
• Optimize performance (p. 1158)
• Frequently asked questions (p. 1159)
• Log API Calls for the EBS direct APIs with AWS CloudTrail (p. 1160)
• EBS direct APIs and interface VPC endpoints (p. 1162)
• Idempotency for StartSnapshot API (p. 1163)

Understand the EBS direct APIs

The following are the key elements that you should understand before getting started with the EBS
direct APIs.

Pricing

The price that you pay to use the EBS direct APIs depends on the requests you make. For more
information, see Amazon EBS pricing.

Snapshots

Snapshots are the primary means to back up data from your EBS volumes. With the EBS direct APIs,
you can also back up data from your on-premises disks to snapshots. To save storage costs, successive
snapshots are incremental, containing only the volume data that changed since the previous snapshot.
For more information, see Amazon EBS snapshots (p. 1126).
Note
Public snapshots are not supported by the EBS direct APIs.

Blocks

A block is a fragment of data within a snapshot. Each snapshot can contain thousands of blocks. All
blocks in a snapshot are of a fixed size.

Block indexes

A block index is the offset position of a block within a snapshot, and it is used to identify the block.
Multiply the BlockIndex value with the BlockSize value (BlockIndex * BlockSize) to identify the logical
offset of the data in the logical volume.

Block tokens

A block token is the identifying hash of a block within a snapshot, and it is used to locate the block data.
Block tokens returned by EBS direct APIs are temporary. They change on the expiry timestamp specified
for them, or if you run another ListSnapshotBlocks or ListChangedBlocks request for the same snapshot.

1143
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

Checksum

A checksum is a small-sized datum derived from a block of data for the purpose of detecting errors that
were introduced during its transmission or storage. The EBS direct APIs use checksums to validate data
integrity. When you read data from an EBS snapshot, the service provides Base64-encoded SHA256
checksums for each block of data transmitted, which you can use for validation. When you write data
to an EBS snapshot, you must provide a Base64 encoded SHA256 checksum for each block of data
transmitted. The service validates the data received using the checksum provided. For more information,
see Use checksums (p. 1150) later in this guide.

Encryption

Encryption protects your data by converting it into unreadable code that can be deciphered only by
people who have access to the key used to encrypt it. You can use the EBS direct APIs to read and write
encrypted snapshots, but there are some limitations. For more information, see Use encryption (p. 1150)
later in this guide.

API actions

The EBS direct APIs consists of six actions. There are three read actions and three write actions. The
read actions are ListSnapshotBlocks, ListChangedBlocks, and GetSnapshotBlock. The write actions are
StartSnapshot, PutSnapshotBlock, and CompleteSnapshot. These actions are described in the following
sections.

List snapshot blocks

The ListSnapshotBlocks action returns the block indexes and block tokens of blocks in the specified
snapshot.

List changed blocks

The ListChangedBlocks action returns the block indexes and block tokens of blocks that are different
between two specified snapshots of the same volume and snapshot lineage.

Get snapshot block

The GetSnapshotBlock action returns the data in a block for the specified snapshot ID, block index, and
block token.

Start snapshot

The StartSnapshot action starts a snapshot, either as an incremental snapshot of an existing one or
as a new snapshot. The started snapshot remains in a pending state until it is completed using the
CompleteSnapshot action.

Put snapshot block

The PutSnapshotBlock action adds data to a started snapshot in the form of individual blocks. You must
specify a Base64-encoded SHA256 checksum for the block of data transmitted. The service validates the
checksum after the transmission is completed. The request fails if the checksum computed by the service
doesn’t match what you specified.

Complete snapshot

The CompleteSnapshot action completes a started snapshot that is in a pending state. The snapshot is
then changed to a completed state.

Use the EBS direct APIs to read snapshots

The following steps describe how to use the EBS direct APIs to read snapshots:

1144
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

1. Use the ListSnapshotBlocks action to view all block indexes and block tokens of blocks in a snapshot.
Or use the ListChangedBlocks action to view only the block indexes and block tokens of blocks that
are different between two snapshots of the same volume and snapshot lineage. These actions help
you identify the block tokens and block indexes of blocks for which you might want to get data.
2. Use the GetSnapshotBlock action, and specify the block index and block token of the block for which
you want to get data.

For examples of how to run these actions, see the Work with the EBS direct APIs using the API or AWS
SDKs (p. 1151) and Work with the EBS direct APIs using the command line (p. 1155) sections later in this
guide.

Use the EBS direct APIs to write incremental snapshots

The following steps describe how to use the EBS direct APIs to write incremental snapshots:

1. Use the StartSnapshot action and specify a parent snapshot ID to start a snapshot as an incremental
snapshot of an existing one, or omit the parent snapshot ID to start a new snapshot. This action
returns the new snapshot ID, which is in a pending state.
2. Use the PutSnapshotBlock action and specify the ID of the pending snapshot to add data to it in the
form of individual blocks. You must specify a Base64-encoded SHA256 checksum for the block of
data transmitted. The service computes the checksum of the data received and validates it with the
checksum that you specified. The action fails if the checksums don't match.
3. When you're done adding data to the pending snapshot, use the CompleteSnapshot action to start an
asynchronous workflow that seals the snapshot and moves it to a completed state.

Repeat these steps to create a new, incremental snapshot using the previously created snapshot as the
parent.

For example, in the following diagram, snapshot A is the first new snapshot started. Snapshot A is used
as the parent snapshot to start snapshot B. Snapshot B is used as the parent snapshot to start and create
snapshot C. Snapshots A, B, and C are incremental snapshots. Snapshot A is used to create EBS volume
1. Snapshot D is created from EBS volume 1. Snapshot D is an incremental snapshot of A; it is not an
incremental snapshot of B or C.

For examples of how to run these actions, see the Work with the EBS direct APIs using the API or AWS
SDKs (p. 1151) and Work with the EBS direct APIs using the command line (p. 1155) sections later in this
guide.

1145
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

Permissions for IAM users

An AWS Identity and Access Management (IAM) user must have the following policies to use the EBS
direct APIs. For more information, see Changing Permissions for an IAM User.

Be cautious when assigning the following policies to IAM users. By assigning these policies, you might
give access to a user who is denied access to the same resource through the Amazon EC2 APIs, such as
the CopySnapshot or CreateVolume actions.

Permissions to read snapshots

The following policy allows the read EBS direct APIs to be used on all snapshots in a specific AWS Region.
In the policy, replace <Region> with the Region of the snapshot.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ebs:ListSnapshotBlocks",
"ebs:ListChangedBlocks",
"ebs:GetSnapshotBlock"
],
"Resource": "arn:aws:ec2:<Region>::snapshot/*"
}
]
}

The following policy allows the read EBS direct APIs to be used on snapshots with a specific key-value
tag. In the policy, replace <Key> with the key value of the tag, and <Value> with the value of the tag.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ebs:ListSnapshotBlocks",
"ebs:ListChangedBlocks",
"ebs:GetSnapshotBlock"
],
"Resource": "arn:aws:ec2:*::snapshot/*",
"Condition": {
"StringEqualsIgnoreCase": {
"aws:ResourceTag/<Key>": "<Value>"
}
}
}
]
}

The following policy allows all of the read EBS direct APIs to be used on all snapshots in the account
only within a specific time range. This policy authorizes use of the EBS direct APIs based on the
aws:CurrentTime global condition key. In the policy, be sure to replace the date and time range shown
with the date and time range for your policy.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",

1146
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

"Action": [
"ebs:ListSnapshotBlocks",
"ebs:ListChangedBlocks",
"ebs:GetSnapshotBlock"
],
"Resource": "arn:aws:ec2:*::snapshot/*",
"Condition": {
"DateGreaterThan": {
"aws:CurrentTime": "2018-05-29T00:00:00Z"
},
"DateLessThan": {
"aws:CurrentTime": "2020-05-29T23:59:59Z"
}
}
}
]
}

The following policy grants access to decrypt an encrypted snapshot using a specific key ID from the
AWS Key Management Service (AWS KMS). It grants access to encrypt new snapshots using the default
AWS KMS key ID for EBS snapshots. It also provides the ability to determine if encrypt by default
is enabled on the account. In the policy, replace <Region> with the Region of the AWS KMS key,
<AccountId> with the ID of the AWS account of the key, and <KeyId> with the ID of the key used to
encrypt the snapshot that you want to read with the EBS direct APIs.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "VisualEditor0",
"Effect": "Allow",
"Action": [
"kms:Encrypt",
"kms:Decrypt",
"kms:GenerateDataKey",
"kms:GenerateDataKeyWithoutPlaintext",
"kms:ReEncrypt*",
"kms:CreateGrant",
"ec2:CreateTags",
"kms:DescribeKey",
"ec2:GetEbsDefaultKmsKeyId",
"ec2:GetEbsEncryptionByDefault"
],
"Resource": "arn:aws:kms:<Region>:<AccountId>:key/<KeyId>"
}
]
}

For more information, see Changing Permissions for an IAM User in the IAM User Guide.

Permissions to write snapshots

The following policy allows the write EBS direct APIs to be used on all snapshots in a specific AWS
Region. In the policy, replace <Region> with the Region of the snapshot.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ebs:StartSnapshot",
"ebs:PutSnapshotBlock",

1147
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

"ebs:CompleteSnapshot"
],
"Resource": "arn:aws:ec2:<Region>::snapshot/*"
}
]
}

The following policy allows the write EBS direct APIs to be used on snapshots with a specific key-value
tag. In the policy, replace <Key> with the key value of the tag, and <Value> with the value of the tag.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ebs:StartSnapshot",
"ebs:PutSnapshotBlock",
"ebs:CompleteSnapshot"
],
"Resource": "arn:aws:ec2:*::snapshot/*",
"Condition": {
"StringEqualsIgnoreCase": {
"aws:ResourceTag/<Key>": "<Value>"
}
}
}
]
}

The following policy allows all of the EBS direct APIs to be used. It also allows the StartSnapshot
action only if a parent snapshot ID is specified. Therefore, this policy blocks the ability to start new
snapshots without using a parent snapshot.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ebs:*",
"Resource": "*",
"Condition": {
"StringEquals": {
"ebs:ParentSnapshot": "arn:aws:ec2:*::snapshot/*"
}
}
}
]
}

The following policy allows all of the EBS direct APIs to be used. It also allows only the user tag key
to be created for a new snapshot. This policy also ensures that the user has access to create tags. The
StartSnapshot action is the only action that can specify tags.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ebs:*",
"Resource": "*",
"Condition": {

1148
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

"ForAllValues:StringEquals": {
"aws:TagKeys": "user"
}
}
},
{
"Effect": "Allow",
"Action": "ec2:CreateTags",
"Resource": "*"
}
]
}

The following policy allows all of the write EBS direct APIs to be used on all snapshots in the account
only within a specific time range. This policy authorizes use of the EBS direct APIs based on the
aws:CurrentTime global condition key. In the policy, be sure to replace the date and time range shown
with the date and time range for your policy.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ebs:StartSnapshot",
"ebs:PutSnapshotBlock",
"ebs:CompleteSnapshot"
],
"Resource": "arn:aws:ec2:*::snapshot/*",
"Condition": {
"DateGreaterThan": {
"aws:CurrentTime": "2018-05-29T00:00:00Z"
},
"DateLessThan": {
"aws:CurrentTime": "2020-05-29T23:59:59Z"
}
}
}
]
}

The following policy grants access to decrypt an encrypted snapshot using a specific key ID from the
AWS Key Management Service (AWS KMS). It grants access to encrypt new snapshots using the default
AWS KMS key ID for EBS snapshots. It also provides the ability to determine if encrypt by default
is enabled on the account. In the policy, replace <Region> with the Region of the AWS KMS key,
<AccountId> with the ID of the AWS account of the key, and <KeyId> with the ID of the key used to
encrypt the snapshot that you want to read with the EBS direct APIs.

{
"Version": "2012-10-17",
"Statement": [
{
"Sid": "VisualEditor0",
"Effect": "Allow",
"Action": [
"kms:Encrypt",
"kms:Decrypt",
"kms:GenerateDataKey",
"kms:GenerateDataKeyWithoutPlaintext",
"kms:ReEncrypt*",
"kms:CreateGrant",
"ec2:CreateTags",

1149
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

"kms:DescribeKey",
"ec2:GetEbsDefaultKmsKeyId",
"ec2:GetEbsEncryptionByDefault"
],
"Resource": "arn:aws:kms:<Region>:<AccountId>:key/<KeyId>"
}
]
}

For more information, see Changing Permissions for an IAM User in the IAM User Guide.

Use encryption
If Amazon EBS encryption by default is enabled on your AWS account, you cannot start a new snapshot
using an un-encrypted parent snapshot. You must first encrypt the parent snapshot by copying it. For
more information, see Copy an Amazon EBS snapshot (p. 1134) and Encryption by default (p. 1178).

To start an encrypted snapshot, specify the Amazon Resource Name (ARN) of an AWS KMS key, or specify
an encrypted parent snapshot in your StartSnapshot request. If neither are specified, and Amazon EBS
encryption by default is enabled on the account, then the default CMK for the account is used. If no
default CMK has been specified for the account, then the AWS managed CMK is used.
Important
By default, all principals in the account have access to the default AWS managed CMK, and they
can use it for EBS encryption and decryption operations. For more information, see Default key
for EBS encryption (p. 1178).

You might need additional IAM permissions to use the EBS direct APIs with encryption. For more
information, see the Permissions for IAM users (p. 1146) section earlier in this guide.

Use Signature Version 4 signing

Signature Version 4 is the process to add authentication information to AWS requests sent by HTTP. For
security, most requests to AWS must be signed with an access key, which consists of an access key ID and
secret access key. These two keys are commonly referred to as your security credentials. For information
about how to obtain credentials for your account, see Understanding and getting your credentials.

If you intend to manually create HTTP requests, you must learn how to sign them. When you use the
AWS Command Line Interface (AWS CLI) or one of the AWS SDKs to make requests to AWS, these tools
automatically sign the requests for you with the access key that you specify when you configure the
tools. When you use these tools, you don't need to learn how to sign requests yourself.

For more information, see Signing AWS requests with Signature Version 4 in the AWS General Reference.

Use checksums
The GetSnapshotBlock action returns data that is in a block of a snapshot, and the PutSnapshotBlock
action adds data to a block in a snapshot. The block data that is transmitted is not signed as part of the
Signature Version 4 signing process. As a result, checksums are used to validate the integrity of the data
as follows:

• When you use the GetSnapshotBlock action, the response provides a Base64-encoded SHA256
checksum for the block data using the x-amz-Checksum header, and the checksum algorithm using
the x-amz-Checksum-Algorithm header. Use the returned checksum to validate the integrity of
the data. If the checksum that you generate doesn't match what Amazon EBS provided, you should
consider the data not valid and retry your request.
• When you use the PutSnapshotBlock action, your request must provide a Base64-encoded SHA256
checksum for the block data using the x-amz-Checksum header, and the checksum algorithm using
the x-amz-Checksum-Algorithm header. The checksum that you provide is validated against a
checksum generated by Amazon EBS to validate the integrity of the data. If the checksums do not
correspond, the request fails.

1150
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

• When you use the CompleteSnapshot action, your request can optionally provide an aggregate
Base64-encoded SHA256 checksum for the complete set of data added to the snapshot. Provide the
checksum using the x-amz-Checksum header, the checksum algorithm using the x-amz-Checksum-
Algorithm header, and the checksum aggregation method using the x-amz-Checksum-Aggregation-
Method header. To generate the aggregated checksum using the linear aggregation method, arrange
the checksums for each written block in ascending order of their block index, concatenate them to
form a single string, and then generate the checksum on the entire string using the SHA256 algorithm.

The checksums in these actions are part of the Signature Version 4 signing process.

Work with the EBS direct APIs using the API or AWS SDKs
The EBS direct APIs Reference provides descriptions and syntax for each of the service’s actions and
data types. You can also use one of the AWS SDKs to access an API that's tailored to the programming
language or platform that you're using. For more information, see AWS SDKs.

The EBS direct APIs require an AWS Signature Version 4 signature. For more information, see Use
Signature Version 4 signing (p. 1150).

Use the API to read snapshots

List blocks in a snapshot
The following ListChangedBlocks example request returns the block indexes and block tokens of blocks
that are in snapshot snap-0acEXAMPLEcf41648. The startingBlockIndex parameter limits the
results to block indexes greater than 1000, and the maxResults parameter limits the results to the first
100 blocks.

GET /snapshots/snap-0acEXAMPLEcf41648/blocks?maxResults=100&startingBlockIndex=1000
HTTP/1.1
Host: ebs.us-east-2.amazonaws.com
Accept-Encoding: identity
User-Agent: <User agent parameter>
X-Amz-Date: 20200617T231953Z
Authorization: <Authentication parameter>

The following example response for the previous request lists the block indexes and block tokens in the
snapshot. Use the GetSnapshotBlock action and specify the block index and block token of the block for
which you want to get data. The block tokens are valid until the expiry time listed.

HTTP/1.1 200 OK
x-amzn-RequestId: d6e5017c-70a8-4539-8830-57f5557f3f27
Content-Type: application/json
Content-Length: 2472
Date: Wed, 17 Jun 2020 23:19:56 GMT
Connection: keep-alive

{
"BlockSize": 524288,
"Blocks": [
{
"BlockIndex": 0,
"BlockToken": "AAUBAcuWqOCnDNuKle11s7IIX6jp6FYcC/q8oT93913HhvLvA+3JRrSybp/0"
},
{
"BlockIndex": 1536,
"BlockToken": "AAUBAWudwfmofcrQhGVlLwuRKm2b8ZXPiyrgoykTRC6IU1NbxKWDY1pPjvnV"
},
{
"BlockIndex": 3072,
"BlockToken": "AAUBAV7p6pC5fKAC7TokoNCtAnZhqq27u6YEXZ3MwRevBkDjmMx6iuA6tsBt"

1151
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

},
{
"BlockIndex": 3073,
"BlockToken": "AAUBAbqt9zpqBUEvtO2HINAfFaWToOwlPjbIsQOlx6JUN/0+iMQl0NtNbnX4"
},
...
],
"ExpiryTime": 1.59298379649E9,
"VolumeSize": 3
}

List blocks that are different between two snapshots

The following ListChangedBlocks example request returns the block indexes and block tokens of blocks
that are different between snapshots snap-0acEXAMPLEcf41648 and snap-0c9EXAMPLE1b30e2f.
The startingBlockIndex parameter limits the results to block indexes greater than 0, and the
maxResults parameter limits the results to the first 500 blocks.

GET /snapshots/snap-0c9EXAMPLE1b30e2f/changedblocks?
firstSnapshotId=snap-0acEXAMPLEcf41648&maxResults=500&startingBlockIndex=0 HTTP/1.1
Host: ebs.us-east-2.amazonaws.com
Accept-Encoding: identity
User-Agent: <User agent parameter>
X-Amz-Date: 20200617T232546Z
Authorization: <Authentication parameter>

The following example response for the previous request shows that block indexes 0, 3072, 6002, and
6003 are different between the two snapshots. Additionally, block indexes 6002, and 6003 exist only in
the first snapshot ID specified, and not in the second snapshot ID because there is no second block token
listed in the response.

Use the GetSnapshotBlock action and specify the block index and block token of the block for which
you want to get data. The block tokens are valid until the expiry time listed.

HTTP/1.1 200 OK
x-amzn-RequestId: fb0f6743-6d81-4be8-afbe-db11a5bb8a1f
Content-Type: application/json
Content-Length: 1456
Date: Wed, 17 Jun 2020 23:25:47 GMT
Connection: keep-alive

{
"BlockSize": 524288,
"ChangedBlocks": [
{
"BlockIndex": 0,
"FirstBlockToken": "AAUBAVaWqOCnDNuKle11s7IIX6jp6FYcC/tJuVT1GgP23AuLntwiMdJ
+OJkL",
"SecondBlockToken": "AAUBASxzy0Y0b33JVRLoYm3NOresCxn5RO+HVFzXW3Y/
RwfFaPX2Edx8QHCh"
},
{
"BlockIndex": 3072,
"FirstBlockToken": "AAUBAcHp6pC5fKAC7TokoNCtAnZhqq27u6fxRfZOLEmeXLmHBf2R/
Yb24MaS",
"SecondBlockToken":
"AAUBARGCaufCqBRZC8tEkPYGGkSv3vqvOjJ2xKDi3ljDFiytUxBLXYgTmkid"
},
{
"BlockIndex": 6002,
"FirstBlockToken": "AAABASqX4/
NWjvNceoyMUljcRd0DnwbSwNnes1UkoP62CrQXvn47BY5435aw"

1152
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

},
{
"BlockIndex": 6003,
"FirstBlockToken":
"AAABASmJ0O5JxAOce25rF4P1sdRtyIDsX12tFEDunnePYUKOf4PBROuICb2A"
},
...
],
"ExpiryTime": 1.592976647009E9,
"VolumeSize": 3
}

Get block data from a snapshot

The following GetSnapshotBlock example request returns the data in the block index 3072 with block
token AAUBARGCaufCqBRZC8tEkPYGGkSv3vqvOjJ2xKDi3ljDFiytUxBLXYgTmkid, in snapshot
snap-0c9EXAMPLE1b30e2f.

GET /snapshots/snap-0c9EXAMPLE1b30e2f/blocks/3072?
blockToken=AAUBARGCaufCqBRZC8tEkPYGGkSv3vqvOjJ2xKDi3ljDFiytUxBLXYgTmkid HTTP/1.1
Host: ebs.us-east-2.amazonaws.com
Accept-Encoding: identity
User-Agent: <User agent parameter>
X-Amz-Date: 20200617T232838Z
Authorization: <Authentication parameter>

The following example response for the previous request shows the size of the data returned, the
checksum to validate the data, and the algorithm used to generate the checksum. The binary data is
transmitted in the body of the response and is represented as BlockData in the following example.

HTTP/1.1 200 OK
x-amzn-RequestId: 2d0db2fb-bd88-474d-a137-81c4e57d7b9f
x-amz-Data-Length: 524288
x-amz-Checksum: Vc0yY2j3qg8bUL9I6GQuI2orTudrQRBDMIhcy7bdEsw=
x-amz-Checksum-Algorithm: SHA256
Content-Type: application/octet-stream
Content-Length: 524288
Date: Wed, 17 Jun 2020 23:28:38 GMT
Connection: keep-alive

BlockData

Use the API to write incremental snapshots

Start a snapshot
The following StartSnapshot example request starts an 8 GiB snapshot, using snapshot
snap-123EXAMPLE1234567 as the parent snapshot. The new snapshot will be an incremental snapshot
of the parent snapshot. The snapshot moves to an error state if there are no put or complete requests
made for the snapshot within the specified 60 minute timeout period. The 550e8400-e29b-41d4-
a716-446655440000 client token ensures idempotency for the request. If the client token is omitted,
the AWS SDK automatically generates one for you. For more information about idempotency, see
Idempotency for StartSnapshot API (p. 1163).

POST /snapshots HTTP/1.1

Host: ebs.us-east-2.amazonaws.com
Accept-Encoding: identity
User-Agent: <User agent parameter>
X-Amz-Date: 20200618T040724Z
Authorization: <Authentication parameter>

1153
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

{
"VolumeSize": 8,
"ParentSnapshot": snap-123EXAMPLE1234567,
"ClientToken": "550e8400-e29b-41d4-a716-446655440000",
"Timeout": 60
}

The following example response for the previous request shows the snapshot ID, AWS account ID, status,
volume size in GiB, and size of the blocks in the snapshot. The snapshot is started in a pending state.
Specify the snapshot ID in a subsequent PutSnapshotBlocks request to write data to the snapshot.

HTTP/1.1 201 Created

x-amzn-RequestId: 929e6eb9-7183-405a-9502-5b7da37c1b18
Content-Type: application/json
Content-Length: 181
Date: Thu, 18 Jun 2020 04:07:29 GMT
Connection: keep-alive

{
"BlockSize": 524288,
"Description": null,
"OwnerId": "138695307491",
"Progress": null,
"SnapshotId": "snap-052EXAMPLEc85d8dd",
"StartTime": null,
"Status": "pending",
"Tags": null,
"VolumeSize": 8
}

Put data into a snapshot

The following PutSnapshot example request writes 524288 Bytes of data to

block index 1000 on snapshot snap-052EXAMPLEc85d8dd. The Base64 encoded
QOD3gmEQOXATfJx2Aa34W4FU2nZGyXfqtsUuktOw8DM= checksum was generated using the SHA256
algorithm. The data is transmitted in the body of the request and is represented as BlockData in the
following example.

PUT /snapshots/snap-052EXAMPLEc85d8dd/blocks/1000 HTTP/1.1

Host: ebs.us-east-2.amazonaws.com
Accept-Encoding: identity
x-amz-Data-Length: 524288
x-amz-Checksum: QOD3gmEQOXATfJx2Aa34W4FU2nZGyXfqtsUuktOw8DM=
x-amz-Checksum-Algorithm: SHA256
User-Agent: <User agent parameter>
X-Amz-Date: 20200618T042215Z
X-Amz-Content-SHA256: UNSIGNED-PAYLOAD
Authorization: <Authentication parameter>

BlockData

The following is example response for the previous request confirms the data length, checksum, and
checksum algorithm for the data received by the service.

HTTP/1.1 201 Created

x-amzn-RequestId: 643ac797-7e0c-4ad0-8417-97b77b43c57b
x-amz-Checksum: QOD3gmEQOXATfJx2Aa34W4FU2nZGyXfqtsUuktOw8DM=
x-amz-Checksum-Algorithm: SHA256
Content-Type: application/json
Content-Length: 2
Date: Thu, 18 Jun 2020 04:22:12 GMT

1154
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

Connection: keep-alive

{}

Complete a snapshot
The following CompleteSnapshot example request completes snapshot snap-052EXAMPLEc85d8dd.
The command specifies that 5 blocks were written to the snapshot. The
6D3nmwi5f2F0wlh7xX8QprrJBFzDX8aacdOcA3KCM3c= checksum represents the checksum for the
complete set of data written to a snapshot.

POST /snapshots/completion/snap-052EXAMPLEc85d8dd HTTP/1.1

Host: ebs.us-east-2.amazonaws.com
Accept-Encoding: identity
x-amz-ChangedBlocksCount: 5
x-amz-Checksum: 6D3nmwi5f2F0wlh7xX8QprrJBFzDX8aacdOcA3KCM3c=
x-amz-Checksum-Algorithm: SHA256
x-amz-Checksum-Aggregation-Method: LINEAR
User-Agent: <User agent parameter>
X-Amz-Date: 20200618T043158Z
Authorization: <Authentication parameter>

The following is an example response for the previous request.

HTTP/1.1 202 Accepted

x-amzn-RequestId: 06cba5b5-b731-49de-af40-80333ac3a117
Content-Type: application/json
Content-Length: 20
Date: Thu, 18 Jun 2020 04:31:50 GMT
Connection: keep-alive

{"Status":"pending"}

Work with the EBS direct APIs using the command line
The following examples show how to use the EBS direct APIs using the AWS Command Line Interface
(AWS CLI). For more information about installing and configuring the AWS CLI, see Installing the AWS CLI
and Quickly Configuring the AWS CLI.

Use the AWS CLI to read snapshots

List blocks in a snapshot

The following list-snapshot-blocks example command returns the block indexes and block tokens of
blocks that are in snapshot snap-0987654321. The --starting-block-index parameter limits the
results to block indexes greater than 1000, and the --max-results parameter limits the results to the
first 100 blocks.

aws ebs list-snapshot-blocks --snapshot-id snap-0987654321 --starting-block-index 1000 --

max-results 100

The following example response for the previous command lists the block indexes and block tokens in
the snapshot. Use the get-snapshot-block command and specify the block index and block token of
the block for which you want to get data. The block tokens are valid until the expiry time listed.

{
"Blocks": [
{
"BlockIndex": 1001,

1155
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

"BlockToken": "AAABAV3/PNhXOynVdMYHUpPsetaSvjLB1dtIGfbJv5OJ0sX855EzGTWos4a4"
},
{
"BlockIndex": 1002,
"BlockToken": "AAABATGQIgwr0WwIuqIMjCA/Sy7e/YoQFZsHejzGNvjKauzNgzeI13YHBfQB"
},
{
"BlockIndex": 1007,
"BlockToken": "AAABAZ9CTuQtUvp/dXqRWw4d07eOgTZ3jvn6hiW30W9duM8MiMw6yQayzF2c"
},
{
"BlockIndex": 1012,
"BlockToken": "AAABAQdzxhw0rVV6PNmsfo/YRIxo9JPR85XxPf1BLjg0Hec6pygYr6laE1p0"
},
{
"BlockIndex": 1030,
"BlockToken": "AAABAaYvPax6mv+iGWLdTUjQtFWouQ7Dqz6nSD9L+CbXnvpkswA6iDID523d"
},
{
"BlockIndex": 1031,
"BlockToken": "AAABATgWZC0XcFwUKvTJbUXMiSPg59KVxJGL+BWBClkw6spzCxJVqDVaTskJ"
},
...
],
"ExpiryTime": 1576287332.806,
"VolumeSize": 32212254720,
"BlockSize": 524288
}

List blocks that are different between two snapshots

The following list-changed-blocks example command returns the block indexes and block tokens of
blocks that are different between snapshots snap-1234567890 and snap-0987654321. The --
starting-block-index parameter limits the results to block indexes greater than 0, and the --max-
results parameter limits the results to the first 500 blocks..

aws ebs list-changed-blocks --first-snapshot-id snap-1234567890 --second-snapshot-

id snap-0987654321 --starting-block-index 0 --max-results 500

The following example response for the previous command shows that block indexes 0, 6000, 6001,
6002, and 6003 are different between the two snapshots. Additionally, block indexes 6001, 6002, and
6003 exist only in the first snapshot ID specified, and not in the second snapshot ID because there is no
second block token listed in the response.

Use the get-snapshot-block command and specify the block index and block token of the block for
which you want to get data. The block tokens are valid until the expiry time listed.

{
"ChangedBlocks": [
{
"BlockIndex": 0,
"FirstBlockToken": "AAABAVahm9SO60Dyi0ORySzn2ZjGjW/
KN3uygGlS0QOYWesbzBbDnX2dGpmC",
"SecondBlockToken":
"AAABAf8o0o6UFi1rDbSZGIRaCEdDyBu9TlvtCQxxoKV8qrUPQP7vcM6iWGSr"
},
{
"BlockIndex": 6000,
"FirstBlockToken": "AAABAbYSiZvJ0/
R9tz8suI8dSzecLjN4kkazK8inFXVintPkdaVFLfCMQsKe",
"SecondBlockToken":
"AAABAZnqTdzFmKRpsaMAsDxviVqEI/3jJzI2crq2eFDCgHmyNf777elD9oVR"

1156
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

},
{
"BlockIndex": 6001,
"FirstBlockToken": "AAABASBpSJ2UAD3PLxJnCt6zun4/
T4sU25Bnb8jB5Q6FRXHFqAIAqE04hJoR"
},
{
"BlockIndex": 6002,
"FirstBlockToken": "AAABASqX4/
NWjvNceoyMUljcRd0DnwbSwNnes1UkoP62CrQXvn47BY5435aw"
},
{
"BlockIndex": 6003,
"FirstBlockToken":
"AAABASmJ0O5JxAOce25rF4P1sdRtyIDsX12tFEDunnePYUKOf4PBROuICb2A"
},
...
],
"ExpiryTime": 1576308931.973,
"VolumeSize": 32212254720,
"BlockSize": 524288,
"NextToken": "AAADARqElNng/sV98CYk/bJDCXeLJmLJHnNSkHvLzVaO0zsPH/QM3Bi3zF//O6Mdi/
BbJarBnp8h"
}

Get block data from a snapshot

The following get-snapshot-block example command returns the data in the block index 6001 with
block token AAABASBpSJ2UAD3PLxJnCt6zun4/T4sU25Bnb8jB5Q6FRXHFqAIAqE04hJoR, in snapshot
snap-1234567890. The binary data is output to the data file in the C:\Temp directory on a Windows
computer. If you run the command on a Linux or Unix computer, replace the output path with /tmp/
data to output the data to the data file in the /tmp directory.

aws ebs get-snapshot-block --snapshot-id snap-1234567890 --block-index 6001 --block-

token AAABASBpSJ2UAD3PLxJnCt6zun4/T4sU25Bnb8jB5Q6FRXHFqAIAqE04hJoR C:/Temp/data

The following example response for the previous command shows the size of the data returned, the
checksum to validate the data, and the algorithm of the checksum. The binary data is automatically
saved to the directory and file you specified in the request command.

{
"DataLength": "524288",
"Checksum": "cf0Y6/Fn0oFa4VyjQPOa/iD0zhTflPTKzxGv2OKowXc=",
"ChecksumAlgorithm": "SHA256"
}

Use the AWS CLI to write incremental snapshots

Start a snapshot
The following start-snapshot example command starts an 8 GiB snapshot, using snapshot
snap-123EXAMPLE1234567 as the parent snapshot. The new snapshot will be an incremental snapshot
of the parent snapshot. The snapshot moves to an error state if there are no put or complete requests
made for the snapshot within the specified 60 minute timeout period. The 550e8400-e29b-41d4-
a716-446655440000 client token ensures idempotency for the request. If the client token is omitted,
the AWS SDK automatically generates one for you. For more information about idempotency, see
Idempotency for StartSnapshot API (p. 1163).

aws ebs start-snapshot --volume-size 8 --parent-snapshot snap-123EXAMPLE1234567 --

timeout 60 --client-token 550e8400-e29b-41d4-a716-446655440000

1157
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

The following example response for the previous command shows the snapshot ID, AWS account ID,
status, volume size in GiB, and size of the blocks in the snapshot. The snapshot is started in a pending
state. Specify the snapshot ID in subsequent put-snapshot-block commands to write data to the
snapshot, then use the complete-snapshot command to complete the snapshot and change its status
to completed.

{
"SnapshotId": "snap-0aaEXAMPLEe306d62",
"OwnerId": "111122223333",
"Status": "pending",
"VolumeSize": 8,
"BlockSize": 524288
}

Put data into a snapshot

The following put-snapshot example command writes 524288 Bytes of data to
block index 1000 on snapshot snap-0aaEXAMPLEe306d62. The Base64 encoded
QOD3gmEQOXATfJx2Aa34W4FU2nZGyXfqtsUuktOw8DM= checksum was generated using the SHA256
algorithm. The data that is transmitted is in the /tmp/data file.

aws ebs put-snapshot-block --snapshot-id snap-0aaEXAMPLEe306d62

--block-index 1000 --data-length 524288 --block-data /tmp/data --
checksum QOD3gmEQOXATfJx2Aa34W4FU2nZGyXfqtsUuktOw8DM= --checksum-algorithm SHA256

The following example response for the previous command confirms the data length, checksum, and
checksum algorithm for the data received by the service.

{
"DataLength": "524288",
"Checksum": "QOD3gmEQOXATfJx2Aa34W4FU2nZGyXfqtsUuktOw8DM=",
"ChecksumAlgorithm": "SHA256"
}

Complete a snapshot
The following complete-snapshot example command completes snapshot
snap-0aaEXAMPLEe306d62. The command specifies that 5 blocks were written to the snapshot.
The 6D3nmwi5f2F0wlh7xX8QprrJBFzDX8aacdOcA3KCM3c= checksum represents the checksum
for the complete set of data written to a snapshot. For more information about checksums, see Use
checksums (p. 1150) earlier in this guide.

aws ebs complete-snapshot --snapshot-id snap-0aaEXAMPLEe306d62 --changed-blocks-count 5

--checksum 6D3nmwi5f2F0wlh7xX8QprrJBFzDX8aacdOcA3KCM3c= --checksum-algorithm SHA256 --
checksum-aggregation-method LINEAR

The following is an example response for the previous command.

{
"Status": "pending"
}

Optimize performance
You can run API requests concurrently. Assuming PutSnapshotBlock latency is 100ms, then a thread
can process 10 requests in one second. Furthermore, assuming your client application creates multiple
threads and connections (for example, 100 connections), it can make 1000 (10 * 100) requests per
second in total. This will correspond to a throughput of around 500 MB per second.

1158
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

The following list contains few things to look for in your application:

• Is each thread using a separate connection? If the connections are limited on the application then
multiple threads will wait for the connection to be available and you will notice lower throughput.
• Is there any wait time in the application between two put requests? This will reduce the effective
throughput of a thread.
• The bandwidth limit on the instance – If bandwidth on the instance is shared by other applications, it
could limit the available throughput for PutSnapshotBlock requests.

Be sure to take note of other workloads that might be running in the account to avoid bottlenecks. You
should also build retry mechanisms into your EBS direct APIs workflows to handle throttling, timeouts,
and service unavailability.

Review the EBS direct APIs service quotas to determine the maximum API requests that you can run per
second. For more information, see Amazon Elastic Block Store Endpoints and Quotas in the AWS General
Reference.

Frequently asked questions

Can a snapshot be accessed using the EBS direct APIs if it has a pending status?

No. The snapshot can be accessed only if it has a completed status.

Are the block indexes returned by the EBS direct APIs in numerical order?

Yes. The block indexes returned are unique, and in numerical order.
Can I submit a request with a MaxResults parameter value of under 100?

No. The minimum MaxResult parameter value you can use is 100. If you submit a request with a
MaxResult parameter value of under 100, and there are more than 100 blocks in the snapshot, then
the API will return at least 100 results.
Can I run API requests concurrently?

You can run API requests concurrently. Be sure to take note of other workloads that might be
running in the account to avoid bottlenecks. You should also build retry mechanisms into your
EBS direct APIs workflows to handle throttling, timeouts, and service unavailability. For more
information, see Optimize performance (p. 1158).

Review the EBS direct APIs service quotas to determine the API requests that you can run per
second. For more information, see Amazon Elastic Block Store Endpoints and Quotas in the AWS
General Reference.
When running the ListChangedBlocks action, is it possible to get an empty response even though
there are blocks in the snapshot?

Yes. If the changed blocks are scarce in the snapshot, the response may be empty but the API will
return a next page token value. Use the next page token value to continue to the next page of
results. You can confirm that you have reached the last page of results when the API returns a next
page token value of null.
If the NextToken parameter is specified together with a StartingBlockIndex parameter, which of the
two is used?

The NextToken is used, and the StartingBlockIndex is ignored.

How long are the block tokens and next tokens valid?

Block tokens are valid for seven days, and next tokens are valid for 60 minutes.
Are encrypted snapshots supported?

Yes. Encrypted snapshots can be accessed using the EBS direct APIs.

1159
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

To access an encrypted snapshot, the user must have access to the key used to encrypt the snapshot,
and the AWS KMS decrypt action. See the Permissions for IAM users (p. 1146) section earlier in this
guide for the AWS KMS policy to assign to a user.
Are public snapshots supported?

Public snapshots are not supported.

Does list snapshot block return all block indexes and block tokens in a snapshot, or only those that
have data written to them?

It returns only block indexes and tokens that have data written to them.
Can I get a history of the API calls made by the EBS direct APIs on my account for security analysis
and operational troubleshooting purposes?

Yes. To receive a history of EBS direct APIs API calls made on your account, turn on AWS CloudTrail in
the AWS Management Console. For more information, see Log API Calls for the EBS direct APIs with
AWS CloudTrail (p. 1160).

Log API Calls for the EBS direct APIs with AWS CloudTrail
The EBS direct APIs service is integrated with AWS CloudTrail. CloudTrail is a service that provides a
record of actions taken by a user, role, or an AWS service in the EBS direct APIs. CloudTrail captures
StartSnapshot and CompleteSnapshot API calls for the EBS direct APIs as events. If you create a trail,
you can enable continuous delivery of CloudTrail events to an Amazon Simple Storage Service (Amazon
S3) bucket, including events for the EBS direct APIs. If you don't configure a trail, you can still view the
most recent events in the CloudTrail console in Event history. You can use the information collected by
CloudTrail to determine the request that was made to the EBS direct APIs, the IP address from which the
request was made, who made the request, when it was made, and additional details.

For more information about CloudTrail, see the AWS CloudTrail User Guide.

EBS direct APIs Information in CloudTrail

CloudTrail is enabled on your AWS account when you create the account. When supported event activity
occurs in the EBS direct APIs, that activity is recorded in a CloudTrail event along with other AWS service
events in Event history. You can view, search, and download recent events in your AWS account. For
more information, see Viewing Events with CloudTrail Event History.

For an ongoing record of events in your AWS account, including events for the EBS direct APIs, create
a trail. A trail enables CloudTrail to deliver log files to an S3 bucket. By default, when you create a trail
in the console, the trail applies to all AWS Regions. The trail logs events from all Regions in the AWS
partition and delivers the log files to the S3 bucket that you specify. Additionally, you can configure
other AWS services to further analyze and act upon the event data collected in CloudTrail logs. For more
information, see the following:

• Overview for Creating a Trail

• CloudTrail Supported Services and Integrations
• Configuring Amazon SNS Notifications for CloudTrail
• Receiving CloudTrail Log Files from Multiple Regions and Receiving CloudTrail Log Files from Multiple
Accounts

Supported API actions

The following API actions support logging as events in CloudTrail log files:

• StartSnapshot
• CompleteSnapshot

1160
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

Identity information

Every event or log entry contains information about who generated the request. The identity
information helps you determine the following:

• Whether the request was made with root or AWS Identity and Access Management (IAM) user
credentials.
• Whether the request was made with temporary security credentials for a role or federated user.
• Whether the request was made by another AWS service.

For more information, see the CloudTrail userIdentityElement.

Understand EBS direct APIs Log File Entries

A trail is a configuration that enables delivery of events as log files to an S3 bucket that you specify.
CloudTrail log files contain one or more log entries. An event represents a single request from any
source and includes information about the requested action, the date and time of the action, request
parameters, and so on. CloudTrail log files aren't an ordered stack trace of the public API calls, so they
don't appear in any specific order.

The following examples show CloudTrail log entries that demonstrates the StartSnapshot and
CompleteSnapshot actions.

StartSnapshot example:

{
"eventVersion": "1.05",
"userIdentity": {
"type": "IAMUser",
"principalId": "123456789012",
"arn": "arn:aws:iam::123456789012:root",
"accountId": "123456789012",
"accessKeyId": "AKIAIOSFODNN7EXAMPLE",
"userName": "user"
},
"eventTime": "2020-07-03T23:27:26Z",
"eventSource": "ebs.amazonaws.com",
"eventName": "StartSnapshot",
"awsRegion": "eu-west-1",
"sourceIPAddress": "192.0.2.0",
"userAgent": "PostmanRuntime/7.25.0",
"requestParameters": {
"volumeSize": 8,
"clientToken": "token",
"encrypted": true
},
"responseElements": {
"snapshotId": "snap-123456789012",
"ownerId": "123456789012",
"status": "pending",
"startTime": "Jul 3, 2020 11:27:26 PM",
"volumeSize": 8,
"blockSize": 524288,
"kmsKeyArn": "HIDDEN_DUE_TO_SECURITY_REASONS"
},
"requestID": "be112233-1ba5-4ae0-8e2b-1c302EXAMPLE",
"eventID": "6e12345-2a4e-417c-aa78-7594fEXAMPLE",
"eventType": "AwsApiCall",
"recipientAccountId": "123456789012"
}

1161
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS snapshots

CompleteSnapshot example:

{
"eventVersion": "1.05",
"userIdentity": {
"type": "IAMUser",
"principalId": "123456789012",
"arn": "arn:aws:iam::123456789012:root",
"accountId": "123456789012",
"accessKeyId": "AKIAIOSFODNN7EXAMPLE",
"userName": "user"
},
"eventTime": "2020-07-03T23:28:24Z",
"eventSource": "ebs.amazonaws.com",
"eventName": "CompleteSnapshot",
"awsRegion": "eu-west-1",
"sourceIPAddress": "192.0.2.0",
"userAgent": "PostmanRuntime/7.25.0",
"requestParameters": {
"snapshotId": "snap-123456789012",
"changedBlocksCount": 5
},
"responseElements": {
"status": "completed"
},
"requestID": "be112233-1ba5-4ae0-8e2b-1c302EXAMPLE",
"eventID": "6e12345-2a4e-417c-aa78-7594fEXAMPLE",
"eventType": "AwsApiCall",
"recipientAccountId": "123456789012"
}

EBS direct APIs and interface VPC endpoints

You can establish a private connection between your VPC and EBS direct APIs by creating an interface
VPC endpoint. Interface endpoints are powered by AWS PrivateLink, a technology that enables you to
privately access EBS direct APIs without an internet gateway, NAT device, VPN connection, or AWS Direct
Connect connection. Instances in your VPC don't need public IP addresses to communicate with EBS
direct APIs. Traffic between your VPC and EBS direct APIs does not leave the Amazon network.

Each interface endpoint is represented by one or more Elastic Network Interfaces in your subnets.

For more information, see Interface VPC endpoints (AWS PrivateLink) in the Amazon VPC User Guide.

Considerations for EBS direct APIs VPC endpoints

Before you set up an interface VPC endpoint for EBS direct APIs, ensure that you review Interface
endpoint properties and limitations in the Amazon VPC User Guide.

VPC endpoint policies are not supported for EBS direct APIs. By default, full access to EBS direct APIs is
allowed through the endpoint. However, you can control access to the interface endpoint using security
groups. For more information, see Controlling access to services with VPC endpoints in the Amazon VPC
User Guide.

Create an interface VPC endpoint for EBS direct APIs

You can create a VPC endpoint for the EBS direct APIs service using either the Amazon VPC console or
the AWS Command Line Interface (AWS CLI). For more information, see Creating an interface endpoint in
the Amazon VPC User Guide.

Create a VPC endpoint for EBS direct APIs using the following service name:

• com.amazonaws.region.ebs

1162
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

If you enable private DNS for the endpoint, you can make API requests to EBS direct APIs using
its default DNS name for the Region, for example, ebs.us-east-1.amazonaws.com. For more
information, see Accessing a service through an interface endpoint in the Amazon VPC User Guide.

Idempotency for StartSnapshot API

Idempotency ensures that an API request completes only once. With an idempotent request, if the
original request completes successful, the subsequent retries return the result from the original
successful request and they have no additional effect.

The StartSnapshot API supports idempotency using a client token. A client token is a unique string
that you specify when you make an API request. If you retry an API request with the same client token
and the same request parameters after it has completed successfully, the result of the original request
is returned. If you retry a request with the same client token, but change one or more of the request
parameters, the ConflictException error is returned.

If you do not specify your own client token, the AWS SDKs automatically generates a client token for the
request to ensure that it is idempotent.

A client token can be any string that includes up to up to 64 ASCII characters. You should not reuse the
same client tokens for different requests.

To make an idempotent StartSnapshot request with your own client token using the API

Specify the ClientToken request parameter.

POST /snapshots HTTP/1.1

Host: ebs.us-east-2.amazonaws.com
Accept-Encoding: identity
User-Agent: <User agent parameter>
X-Amz-Date: 20200618T040724Z
Authorization: <Authentication parameter>

{
"VolumeSize": 8,
"ParentSnapshot": snap-123EXAMPLE1234567,
"ClientToken": "550e8400-e29b-41d4-a716-446655440000",
"Timeout": 60
}

To make an idempotent StartSnapshot request with your own client token using the AWS CLI

Specify the client-token request parameter.

$ aws ebs start-snapshot --region us-east-2 --volume-size 8 --parent-snapshot

snap-123EXAMPLE1234567 --timeout 60 --client-token 550e8400-e29b-41d4-a716-446655440000

Automate the snapshot lifecycle

You can use Amazon Data Lifecycle Manager to automate the creation, retention, and deletion of
snapshots that you use to back up your Amazon EBS volumes.

For more information, see Amazon Data Lifecycle Manager (p. 1186).

Amazon EBS data services

Amazon EBS provides the following data services.

1163
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Data services
• Amazon EBS Elastic Volumes (p. 1164)
• Amazon EBS encryption (p. 1176)
• Amazon Data Lifecycle Manager (p. 1186)
• Amazon EBS fast snapshot restore (p. 1210)

Amazon EBS Elastic Volumes

With Amazon EBS Elastic Volumes, you can increase the volume size, change the volume type, or adjust
the performance of your EBS volumes. If your instance supports Elastic Volumes, you can do so without
detaching the volume or restarting the instance. This enables you to continue using your application
while the changes take effect.

There is no charge to modify the configuration of a volume. You are charged for the new volume
configuration after volume modification starts. For more information, see the Amazon EBS Pricing page.

Contents
• Requirements when modifying volumes (p. 1164)
• Request modifications to your EBS volumes (p. 1166)
• Monitor the progress of volume modifications (p. 1169)
• Extend a Linux file system after resizing a volume (p. 1172)

Requirements when modifying volumes

The following requirements and limitations apply when you modify an Amazon EBS volume. To learn
more about the general requirements for EBS volumes, see Constraints on the size and configuration of
an EBS volume (p. 1101).

Supported instance types

Elastic Volumes are supported on the following instances:

• All current-generation instances (p. 204)

• The following previous-generation instances: C1, C3, CC2, CR1, G2, I2, M1, M3, and R3

If your instance type does not support Elastic Volumes, see Modify an EBS volume if Elastic Volumes is
not supported (p. 1168).

Requirements for Linux volumes

Linux AMIs require a GUID partition table (GPT) and GRUB 2 for boot volumes that are 2 TiB (2,048
GiB) or larger. Many Linux AMIs today still use the MBR partitioning scheme, which only supports boot
volume sizes up to 2 TiB. If your instance does not boot with a boot volume larger than 2 TiB, the AMI
you are using may be limited to a boot volume size of less than 2 TiB. Non-boot volumes do not have
this limitation on Linux instances. For requirements affecting Windows volumes, see Requirements for
Windows Volumes in the Amazon EC2 User Guide for Windows Instances.

Before attempting to resize a boot volume beyond 2 TiB, you can determine whether the volume is using
MBR or GPT partitioning by running the following command on your instance:

[ec2-user ~]$ sudo gdisk -l /dev/xvda

An Amazon Linux instance with GPT partitioning returns the following information:

1164
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

GPT fdisk (gdisk) version 0.8.10

Partition table scan:

MBR: protective
BSD: not present
APM: not present
GPT: present

Found valid GPT with protective MBR; using GPT.

A SUSE instance with MBR partitioning returns the following information:

GPT fdisk (gdisk) version 0.8.8

Partition table scan:

MBR: MBR only
BSD: not present
APM: not present
GPT: not present

Limitations

• There are limits to the maximum aggregated storage that can be requested across volume
modifications. For more information, see Amazon EBS service quotas in the Amazon Web Services
General Reference.
• The new volume size cannot exceed the supported volume capacity. For more information, see
Constraints on the size and configuration of an EBS volume (p. 1101).
• You can't use Elastic Volume operations to modify the volume type of Multi-Attach enabled io2
volumes.
• You can't use Elastic Volume operations to change the volume type, size, or Provisioned IOPS of Multi-
Attach enabled io1 volumes.
• If the volume was attached before November 3, 2016 23:40 UTC, you must initialize Elastic Volumes
support. For more information, see Initializing Elastic Volumes Support (p. 1167).
• If you are using an unsupported previous-generation instance type, or if you encounter an error
while attempting a volume modification, see Modify an EBS volume if Elastic Volumes is not
supported (p. 1168).
• A gp2 volume that is attached to an instance as a root volume cannot be modified to an st1 or sc1
volume. If detached and modified to st1 or sc1, it cannot be attached to an instance as the root
volume.
• A gp2 volume cannot be modified to an st1 or sc1 volume if the requested volume size is below the
minimum size for st1 and sc1 volumes.
• In some cases, you must detach the volume or stop the instance for modification to proceed. If you
encounter an error message while attempting to modify an EBS volume, or if you are modifying an
EBS volume attached to a previous-generation instance type, take one of the following steps:
• For a non-root volume, detach the volume from the instance, apply the modifications, and then re-
attach the volume.
• For a root (boot) volume, stop the instance, apply the modifications, and then restart the instance.
• After provisioning over 32,000 IOPS on an existing io1 or io2 volume, you may need to do one of the
following to see the full performance improvements:
• Detach and attach the volume.
• Restart the instance.
• Decreasing the size of an EBS volume is not supported. However, you can create a smaller volume and
then migrate your data to it using an application-level tool such as rsync.

1165
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

• Modification time is increased if you modify a volume that has not been fully initialized. For more
information see Initialize Amazon EBS volumes (p. 1241).
• After modifying a volume, wait at least six hours and ensure that the volume is in the in-use or
available state before making additional modifications to the same volume.
• While m3.medium instances fully support volume modification, m3.large, m3.xlarge, and
m3.2xlarge instances might not support all volume modification features.

Request modifications to your EBS volumes

With Elastic Volumes, you can dynamically modify the size, performance, and volume type of your
Amazon EBS volumes without detaching them.

Use the following process when modifying a volume:

1. (Optional) Before modifying a volume that contains valuable data, it is a best practice to create a
snapshot of the volume in case you need to roll back your changes. For more information, see Create
Amazon EBS snapshots (p. 1129).
2. Request the volume modification.
3. Monitor the progress of the volume modification. For more information, see Monitor the progress of
volume modifications (p. 1169).
4. If the size of the volume was modified, extend the volume's file system to take advantage of the
increased storage capacity. For more information, see Extend a Linux file system after resizing a
volume (p. 1172).

Contents
• Modify an EBS volume using Elastic Volumes (console) (p. 1166)
• Modify an EBS volume using Elastic Volumes (AWS CLI) (p. 1167)
• Initialize Elastic Volumes support (if needed) (p. 1167)
• Modify an EBS volume if Elastic Volumes is not supported (p. 1168)

Modify an EBS volume using Elastic Volumes (console)

Use the following procedure to modify an EBS volume.

To modify an EBS volume using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Volumes, select the volume to modify, and then choose Actions, Modify Volume.
3. The Modify Volume window displays the volume ID and the volume's current configuration,
including type, size, IOPS, and throughput. Set new configuration values as follows:

• To modify the type, choose a value for Volume Type.

• To modify the size, enter a new value for Size.
• To modify the IOPS, if the volume type is gp3, io1, or io2, enter a new value for IOPS.
• To modify the throughput, if the volume type is gp3, enter a new value for Throughput.
4. After you have finished changing the volume settings, choose Modify. When prompted for
confirmation, choose Yes.
5. Modifying volume size has no practical effect until you also extend the volume's file system to make
use of the new storage capacity. For more information, see Extend a Linux file system after resizing a
volume (p. 1172).

1166
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Modify an EBS volume using Elastic Volumes (AWS CLI)

Use the modify-volume command to modify one or more configuration settings for a volume. For
example, if you have a volume of type gp2 with a size of 100 GiB, the following command changes its
configuration to a volume of type io1 with 10,000 IOPS and a size of 200 GiB.

aws ec2 modify-volume --volume-type io1 --iops 10000 --size 200 --volume-
id vol-11111111111111111

The following is example output:

{
"VolumeModification": {
"TargetSize": 200,
"TargetVolumeType": "io1",
"ModificationState": "modifying",
"VolumeId": "vol-11111111111111111",
"TargetIops": 10000,
"StartTime": "2017-01-19T22:21:02.959Z",
"Progress": 0,
"OriginalVolumeType": "gp2",
"OriginalIops": 300,
"OriginalSize": 100
}
}

Modifying volume size has no practical effect until you also extend the volume's file system to make
use of the new storage capacity. For more information, see Extend a Linux file system after resizing a
volume (p. 1172).

Initialize Elastic Volumes support (if needed)

Before you can modify a volume that was attached to an instance before November 3, 2016 23:40 UTC,
you must initialize volume modification support using one of the following actions:

• Detach and attach the volume

• Stop and start the instance

Use one of the following procedures to determine whether your instances are ready for volume
modification.

New console

To determine whether your instances are ready using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. On the navigation pane, choose Instances.
3. Choose the Show/Hide Columns icon (the gear). Select the Launch time attribute column and
then choose Confirm.
4. Sort the list of instances by the Launch Time column. For each instance that was started before
the cutoff date, choose the Storage tab and check the Attachment time column to see when its
volumes were attached.

Old console

To determine whether your instances are ready using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

1167
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

2. On the navigation pane, choose Instances.

3. Choose the Show/Hide Columns icon (the gear). Select the Launch Time and Block Devices
attributes and then choose Close.
4. Sort the list of instances by the Launch Time column. For instances that were started before the
cutoff date, check when the devices were attached. In the following example, you must initialize
volume modification for the first instance because it was started before the cutoff date and its
root volume was attached before the cutoff date. The other instances are ready because they
were started after the cutoff date.

To determine whether your instances are ready using the CLI

Use the following describe-instances command to determine whether the volume was attached before
November 3, 2016 23:40 UTC.

aws ec2 describe-instances --query "Reservations[*].Instances[*].

[InstanceId,LaunchTime<='2016-11-01',BlockDeviceMappings[*][Ebs.AttachTime<='2016-11-01']]"
--output text

The first line of the output for each instance shows its ID and whether it was started before the cutoff
date (True or False). The first line is followed by one or more lines that show whether each EBS volume
was attached before the cutoff date (True or False). In the following example output, you must initialize
volume modification for the first instance because it was started before the cutoff date and its root
volume was attached before the cutoff date. The other instances are ready because they were started
after the cutoff date.

i-e905622e True
True
i-719f99a8 False
True
i-006b02c1b78381e57 False
False
False
i-e3d172ed False
True

Modify an EBS volume if Elastic Volumes is not supported

If you are using a supported instance type, you can use Elastic Volumes to dynamically modify the size,
performance, and volume type of your Amazon EBS volumes without detaching them.

If you cannot use Elastic Volumes but you need to modify the root (boot) volume, you must stop the
instance, modify the volume, and then restart the instance.

After the instance has started, you can check the file system size to see if your instance recognizes the
larger volume space. On Linux, use the df -h command to check the file system size.

[ec2-user ~]$ df -h
Filesystem Size Used Avail Use% Mounted on
/dev/xvda1 7.9G 943M 6.9G 12% /
tmpfs 1.9G 0 1.9G 0% /dev/shm

1168
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

If the size does not reflect your newly expanded volume, you must extend the file system of your device
so that your instance can use the new space. For more information, see Extend a Linux file system after
resizing a volume (p. 1172).

Monitor the progress of volume modifications

When you modify an EBS volume, it goes through a sequence of states. The volume enters the
modifying state, the optimizing state, and finally the completed state. At this point, the volume is
ready to be further modified.
Note
Rarely, a transient AWS fault can result in a failed state. This is not an indication of volume
health; it merely indicates that the modification to the volume failed. If this occurs, retry the
volume modification.

While the volume is in the optimizing state, your volume performance is in between the source and
target configuration specifications. Transitional volume performance will be no less than the source
volume performance. If you are downgrading IOPS, transitional volume performance is no less than the
target volume performance.

Volume modification changes take effect as follows:

• Size changes usually take a few seconds to complete and take effect after a volume is in the
Optimizing state.
• Performance (IOPS) changes can take from a few minutes to a few hours to complete and are
dependent on the configuration change being made.
• It might take up to 24 hours for a new configuration to take effect, and in some cases more, such as
when the volume has not been fully initialized. Typically, a fully used 1-TiB volume takes about 6 hours
to migrate to a new performance configuration.

Use one of the following methods to monitor the progress of a volume modification.

Contents
• Monitor the progress of a volume modification (console) (p. 1169)
• Monitor the progress of a volume modification (AWS CLI) (p. 1170)
• Monitor the progress of a volume modification (CloudWatch Events) (p. 1171)

Monitor the progress of a volume modification (console)

Use the following procedure to view the progress of one or more volume modifications.

To monitor progress of a modification using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Volumes.
3. Select the volume.
4. The State column and the State field in the details pane contain information in the following
format: volume-state - modification-state (progress%). The possible volume states are creating,
available, in-use, deleting, deleted, and error. The possible modification states are modifying,
optimizing, and completed. Shortly after the volume modification is completed, we remove the
modification state and progress, leaving only the volume state.

In this example, the modification state of the selected volume is optimizing. The modification state
of the next volume is modifying.

1169
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

5. Choose the text in the State field in the details pane to display information about the most recent
modification action, as shown in the previous step.

Monitor the progress of a volume modification (AWS CLI)

Use the describe-volumes-modifications command to view the progress of one or more volume
modifications. The following example describes the volume modifications for two volumes.

aws ec2 describe-volumes-modifications --volume-

id vol-11111111111111111 vol-22222222222222222

In the following example output, the volume modifications are still in the modifying state. Progress is
reported as a percentage.

{
"VolumesModifications": [
{
"TargetSize": 200,
"TargetVolumeType": "io1",
"ModificationState": "modifying",
"VolumeId": "vol-11111111111111111",
"TargetIops": 10000,
"StartTime": "2017-01-19T22:21:02.959Z",
"Progress": 0,
"OriginalVolumeType": "gp2",
"OriginalIops": 300,
"OriginalSize": 100
},
{
"TargetSize": 2000,
"TargetVolumeType": "sc1",
"ModificationState": "modifying",
"VolumeId": "vol-22222222222222222",
"StartTime": "2017-01-19T22:23:22.158Z",
"Progress": 0,
"OriginalVolumeType": "gp2",
"OriginalIops": 300,
"OriginalSize": 1000
}
]

1170
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

The next example describes all volumes with a modification state of either optimizing or completed,
and then filters and formats the results to show only modifications that were initiated on or after
February 1, 2017:

aws ec2 describe-volumes-modifications --filters Name=modification-

state,Values="optimizing","completed" --query "VolumesModifications[?
StartTime>='2017-02-01'].{ID:VolumeId,STATE:ModificationState}"

The following is example output with information about two volumes:

[
{
"STATE": "optimizing",
"ID": "vol-06397e7a0eEXAMPLE"
},
{
"STATE": "completed",
"ID": "vol-ba74e18c2aEXAMPLE"
}
]

Monitor the progress of a volume modification (CloudWatch Events)

With CloudWatch Events, you can create a notification rule for volume modification events. You can
use your rule to generate a notification message using Amazon SNS or to invoke a Lambda function in
response to matching events.

To monitor progress of a modification using CloudWatch Events

1. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.

2. Choose Events, Create rule.
3. For Build event pattern to match events by service, choose Custom event pattern.
4. For Build custom event pattern, replace the contents with the following and choose Save.

{
"source": [
"aws.ec2"
],
"detail-type": [
"EBS Volume Notification"
],
"detail": {
"event": [
"modifyVolume"
]
}
}

The following is example event data:

{
"version": "0",
"id": "01234567-0123-0123-0123-012345678901",
"detail-type": "EBS Volume Notification",
"source": "aws.ec2",
"account": "012345678901",

1171
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

"time": "2017-01-12T21:09:07Z",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:012345678901:volume/vol-03a55cf56513fa1b6"
],
"detail": {
"result": "optimizing",
"cause": "",
"event": "modifyVolume",
"request-id": "01234567-0123-0123-0123-0123456789ab"
}
}

Extend a Linux file system after resizing a volume

After you increase the size of an EBS volume (p. 1166), you must use file system–specific commands to
extend the file system to the larger size. You can resize the file system as soon as the volume enters the
optimizing state.
Important
Before extending a file system that contains valuable data, it is best practice to create a
snapshot of the volume, in case you need to roll back your changes. For more information,
see Create Amazon EBS snapshots (p. 1129). If your Linux AMI uses the MBR partitioning
scheme, you are limited to a boot volume size of up to 2 TiB. For more information, see
Requirements for Linux volumes (p. 1164) and Constraints on the size and configuration of an
EBS volume (p. 1101).

The process for extending a file system on Linux is as follows:

1. Your EBS volume might have a partition that contains the file system and data. Increasing the size of
a volume does not increase the size of the partition. Before you extend the file system on a resized
volume, check whether the volume has a partition that must be extended to the new size of the
volume.
2. Use a file system-specific command to resize each file system to the new volume capacity.

For information about extending a Windows file system, see Extending a Windows file system after
resizing a volume in the Amazon EC2 User Guide for Windows Instances.

The following examples walk you through the process of extending a Linux file system. For file systems
and partitioning schemes other than the ones shown here, refer to the documentation for those file
systems and partitioning schemes for instructions.

Examples
• Example: Extend the file system of NVMe EBS volumes (p. 1172)
• Example: Extend the file system of EBS volumes (p. 1174)

Example: Extend the file system of NVMe EBS volumes

For this example, suppose that you have an instance built on the Nitro System (p. 209), such as an M5
instance. You resized the boot volume from 8 GB to 16 GB and an additional volume from 8 GB to 30 GB.
Use the following procedure to extend the file system of the resized volumes.

To extend the file system of NVMe EBS volumes

1. Connect to your instance (p. 598).

2. To verify the file system for each volume, use the df -hT command.

1172
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

[ec2-user ~]$ df -hT

The following is example output for an instance that has a boot volume with an XFS file system
and an additional volume with an XFS file system. The naming convention /dev/nvme[0-26]n1
indicates that the volumes are exposed as NVMe block devices.

[ec2-user ~]$ df -hT

Filesystem Type Size Used Avail Use% Mounted on
/dev/nvme0n1p1 xfs 8.0G 1.6G 6.5G 20% /
/dev/nvme1n1 xfs 8.0G 33M 8.0G 1% /data
...

3. To check whether the volume has a partition that must be extended, use the lsblk command to
display information about the NVMe block devices attached to your instance.

[ec2-user ~]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
nvme1n1 259:0 0 30G 0 disk /data
nvme0n1 259:1 0 16G 0 disk
##nvme0n1p1 259:2 0 8G 0 part /
##nvme0n1p128 259:3 0 1M 0 part

This example output shows the following:

• The root volume, /dev/nvme0n1, has a partition, /dev/nvme0n1p1. While the size of the root
volume reflects the new size, 16 GB, the size of the partition reflects the original size, 8 GB, and
must be extended before you can extend the file system.
• The volume /dev/nvme1n1 has no partitions. The size of the volume reflects the new size, 30 GB.
4. For volumes that have a partition, such as the root volume shown in the previous step, use the
growpart command to extend the partition. Notice that there is a space between the device name
and the partition number.

[ec2-user ~]$ sudo growpart /dev/nvme0n1 1

5. (Optional) To verify that the partition reflects the increased volume size, use the lsblk command
again.

[ec2-user ~]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
nvme1n1 259:0 0 30G 0 disk /data
nvme0n1 259:1 0 16G 0 disk
##nvme0n1p1 259:2 0 16G 0 part /
##nvme0n1p128 259:3 0 1M 0 part

6. To verify the size of the file system for each volume, use the df -h command. In this example output,
both file systems reflect the original volume size, 8 GB.

[ec2-user ~]$ df -h
Filesystem Size Used Avail Use% Mounted on
/dev/nvme0n1p1 8.0G 1.6G 6.5G 20% /
/dev/nvme1n1 8.0G 33M 8.0G 1% /data
...

7. To extend the file system on each volume, use the correct command for your file system, as follows:

• [XFS file system] To extend the file system on each volume, use the xfs_growfs command. In this
example, / and /data are the volume mount points shown in the output for df -h.

1173
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

[ec2-user ~]$ sudo xfs_growfs -d /

[ec2-user ~]$ sudo xfs_growfs -d /data

If the XFS tools are not already installed, you can install them as follows.

[ec2-user ~]$ sudo yum install xfsprogs

• [ext4 file system] To extend the file system on each volume, use the resize2fs command.

[ec2-user ~]$ sudo resize2fs /dev/nvme0n1p1

[ec2-user ~]$ sudo resize2fs /dev/nvme1n1

• [Other file system] To extend the file system on each volume, refer to the documentation for your
file system for instructions.
8. (Optional) To verify that each file system reflects the increased volume size, use the df -h command
again.

[ec2-user ~]$ df -h
Filesystem Size Used Avail Use% Mounted on
/dev/nvme0n1p1 16G 1.6G 15G 10% /
/dev/nvme1n1 30G 33M 30G 1% /data
...

Example: Extend the file system of EBS volumes

For this example, suppose that you have resized the boot volume of an instance, such as a T2 instance,
from 8 GB to 16 GB and an additional volume from 8 GB to 30 GB. Use the following procedure to
extend the file system of the resized volumes.

To extend the file system of EBS volumes

1. Connect to your instance (p. 598).

2. To verify the file system in use for each volume, use the df -hT command.

[ec2-user ~]$ df -hT

The following is example output for an instance that has a boot volume with an ext4 file system and
an additional volume with an XFS file system.

[ec2-user ~]$ df -hT

Filesystem Type Size Used Avail Use% Mounted on
/dev/xvda1 ext4 8.0G 1.9G 6.2G 24% /
/dev/xvdf1 xfs 8.0G 45M 8.0G 1% /data
...

3. To check whether the volume has a partition that must be extended, use the lsblk command to
display information about the block devices attached to your instance.

[ec2-user ~]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
xvda 202:0 0 16G 0 disk
##xvda1 202:1 0 8G 0 part /
xvdf 202:80 0 30G 0 disk
##xvdf1 202:81 0 8G 0 part /data

1174
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

This example output shows the following:

• The root volume, /dev/xvda, has a partition, /dev/xvda1. While the size of the volume is 16 GB,
the size of the partition is still 8 GB and must be extended.
• The volume /dev/xvdf has a partition, /dev/xvdf1. While the size of the volume is 30G, the
size of the partition is still 8 GB and must be extended.
4. For volumes that have a partition, such as the volumes shown in the previous step, use the growpart
command to extend the partition. Notice that there is a space between the device name and the
partition number.

[ec2-user ~]$ sudo growpart /dev/xvda 1

[ec2-user ~]$ sudo growpart /dev/xvdf 1

5. (Optional) To verify that the partitions reflect the increased volume size, use the lsblk command
again.

[ec2-user ~]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
xvda 202:0 0 16G 0 disk
##xvda1 202:1 0 16G 0 part /
xvdf 202:80 0 30G 0 disk
##xvdf1 202:81 0 30G 0 part /data

6. To verify the size of the file system for each volume, use the df -h command. In this example output,
both file systems reflect the original volume size, 8 GB.

[ec2-user ~]$ df -h
Filesystem Size Used Avail Use% Mounted on
/dev/xvda1 8.0G 1.9G 6.2G 24% /
/dev/xvdf1 8.0G 45M 8.0G 1% /data
...

7. To extend the file system on each volume, use the correct command for your file system, as follows:

• [XFS volumes] To extend the file system on each volume, use the xfs_growfs command. In this
example, / and /data are the volume mount points shown in the output for df -h.

[ec2-user ~]$ sudo xfs_growfs -d /

[ec2-user ~]$ sudo xfs_growfs -d /data

If the XFS tools are not already installed, you can install them as follows.

[ec2-user ~]$ sudo yum install xfsprogs

• [ext4 volumes] To extend the file system on each volume, use the resize2fs command.

[ec2-user ~]$ sudo resize2fs /dev/xvda1

[ec2-user ~]$ sudo resize2fs /dev/xvdf1

• [Other file system] To extend the file system on each volume, refer to the documentation for your
file system for instructions.
8. (Optional) To verify that each file system reflects the increased volume size, use the df -h command
again.

[ec2-user ~]$ df -h
Filesystem Size Used Avail Use% Mounted on
/dev/xvda1 16G 1.9G 14G 12% /

1175
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

/dev/xvdf1 30G 45M 30G 1% /data

...

Amazon EBS encryption

Use Amazon EBS encryption as a straight-forward encryption solution for your EBS resources associated
with your EC2 instances. With Amazon EBS encryption, you aren't required to build, maintain, and secure
your own key management infrastructure. Amazon EBS encryption uses AWS Key Management Service
(AWS KMS) customer master keys (CMK) when creating encrypted volumes and snapshots.

Encryption operations occur on the servers that host EC2 instances, ensuring the security of both data-
at-rest and data-in-transit between an instance and its attached EBS storage.

You can attach both encrypted and unencrypted volumes to an instance simultaneously

Contents
• How EBS encryption works (p. 1176)
• Requirements (p. 1177)
• Default key for EBS encryption (p. 1178)
• Encryption by default (p. 1178)
• Encrypt EBS resources (p. 1179)
• Encryption scenarios (p. 1180)
• Set encryption defaults using the API and CLI (p. 1186)

How EBS encryption works

You can encrypt both the boot and data volumes of an EC2 instance. When you create an encrypted EBS
volume and attach it to a supported instance type, the following types of data are encrypted:

• Data at rest inside the volume

• All data moving between the volume and the instance
• All snapshots created from the volume
• All volumes created from those snapshots

EBS encrypts your volume with a data key using the industry-standard AES-256 algorithm. Your data
key is stored on-disk with your encrypted data, but not before EBS encrypts it with your CMK. Your data
key never appears on disk in plaintext. The same data key is shared by snapshots of the volume and any
subsequent volumes created from those snapshots. For more information, see Data keys in the AWS Key
Management Service Developer Guide.

Amazon EBS works with AWS KMS to encrypt and decrypt your EBS volumes as follows:

1. Amazon EBS sends a GenerateDataKeyWithoutPlaintext request to AWS KMS, specifying the CMK that
you chose for volume encryption.
2. AWS KMS generates a new data key, encrypts it under the CMK that you chose for volume encryption,
and sends the encrypted data key to Amazon EBS to be stored with the volume metadata.
3. When you attach an encrypted volume to an instance, Amazon EC2 sends a Decrypt request to AWS
KMS, specifying the encrypted data key.
4. Amazon EBS sends a CreateGrant request to AWS KMS, so that it can decrypt the data key.
5. AWS KMS decrypts the encrypted data key and sends the decrypted data key to Amazon EC2.

1176
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

6. Amazon EC2 uses the plaintext data key in hypervisor memory to encrypt disk I/O to the volume. The
plaintext data key persists in memory as long as the volume is attached to the instance.

For more information, see How Amazon Elastic Block Store (Amazon EBS) uses AWS KMS and Amazon
EC2 example two in the AWS Key Management Service Developer Guide.

Requirements
Before you begin, verify that the following requirements are met.

Supported volume types

Encryption is supported by all EBS volume types. You can expect the same IOPS performance on
encrypted volumes as on unencrypted volumes, with a minimal effect on latency. You can access
encrypted volumes the same way that you access unencrypted volumes. Encryption and decryption are
handled transparently, and they require no additional action from you or your applications.

Supported instance types

Amazon EBS encryption is available on all current generation (p. 204) instance types and the following
previous generation (p. 207) instance types: A1, C3, cr1.8xlarge, G2, I2, M3, and R3.

Permissions for IAM users

When you configure a CMK as the default key for EBS encryption, the default key policy allows any IAM
user with access to the required KMS actions to use this key to encrypt or decrypt EBS resources. You
must grant IAM users permission to call the following actions in order to use EBS encryption:

• kms:CreateGrant
• kms:Decrypt
• kms:DescribeKey
• kms:GenerateDataKeyWithoutPlainText
• kms:ReEncrypt

To follow the principle of least privilege, do not allow full access to kms:CreateGrant. Instead, allow
the user to create grants on the CMK only when the grant is created on the user's behalf by an AWS
service, as shown in the following example.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "kms:CreateGrant",
"Resource": [
"arn:aws:kms:us-east-2:123456789012:key/abcd1234-a123-456d-a12b-
a123b4cd56ef"
],
"Condition": {
"Bool": {
"kms:GrantIsForAWSResource": true
}
}
}
]
}

For more information, see Allows access to the AWS account and enables IAM policies in the Default key
policy section in the AWS Key Management Service Developer Guide.

1177
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Default key for EBS encryption

Amazon EBS automatically creates a unique AWS managed CMK in each Region where you store AWS
resources. This key has the alias alias/aws/ebs. By default, Amazon EBS uses this key for encryption.
Alternatively, you can specify a symmetric customer managed CMK that you created as the default key
for EBS encryption. Using your own CMK gives you more flexibility, including the ability to create, rotate,
and disable keys.
Important
Amazon EBS does not support asymmetric CMKs. For more information, see Using symmetric
and asymmetric keys in the AWS Key Management Service Developer Guide.

New console

To configure the default key for EBS encryption for a Region

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region.
3. From the navigation pane, select EC2 Dashboard.
4. In the upper-right corner of the page, choose Account Attributes, EBS encryption.
5. Choose Manage.
6. For Default encryption key, choose a symmetric customer managed CMK.
7. Choose Update EBS encryption.

Old console

To configure the default key for EBS encryption for a Region

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region.
3. From the navigation pane, select EC2 Dashboard.
4. In the upper-right corner of the page, choose Account Attributes, Settings.
5. Choose Change the default key and then choose an available key.
6. Choose Save settings.

Encryption by default
You can configure your AWS account to enforce the encryption of the new EBS volumes and snapshot
copies that you create. For example, Amazon EBS encrypts the EBS volumes created when you launch an
instance and the snapshots that you copy from an unencrypted snapshot. For examples of transitioning
from unencrypted to encrypted EBS resources, see Encrypt unencrypted resources (p. 1180).

Encryption by default has no effect on existing EBS volumes or snapshots.

Considerations

• Encryption by default is a Region-specific setting. If you enable it for a Region, you cannot disable it
for individual volumes or snapshots in that Region.
• When you enable encryption by default, you can launch an instance only if the instance type supports
EBS encryption. For more information, see Supported instance types (p. 1177).
• When migrating servers using AWS Server Migration Service (SMS), do not turn on encryption by
default. If encryption by default is already on and you are experiencing delta replication failures, turn
off encryption by default. Instead, enable AMI encryption when you create the replication job.

1178
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

New console

To enable encryption by default for a Region

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region.
3. From the navigation pane, select EC2 Dashboard.
4. In the upper-right corner of the page, choose Account Attributes, EBS encryption.
5. Choose Manage.
6. Select Enable. You keep the AWS managed CMK with the alias alias/aws/ebs created on your
behalf as the default encryption key, or choose a symmetric customer managed CMK.
7. Choose Update EBS encryption.

Old console

To enable encryption by default for a Region

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region.
3. From the navigation pane, select EC2 Dashboard.
4. In the upper-right corner of the page, choose Account Attributes, Settings.
5. Under EBS Storage, select Always encrypt new EBS volumes.
6. Choose Save settings.

You cannot change the CMK that is associated with an existing snapshot or encrypted volume. However,
you can associate a different CMK during a snapshot copy operation so that the resulting copied
snapshot is encrypted by the new CMK.

Encrypt EBS resources

You encrypt EBS volumes by enabling encryption, either using encryption by default (p. 1178) or by
enabling encryption when you create a volume that you want to encrypt.

When you encrypt a volume, you can specify the symmetric CMK to use to encrypt the volume. If you
do not specify a CMK, the key that is used for encryption depends on the encryption state of the source
snapshot and its ownership. For more information, see the encryption outcomes table (p. 1184).
Note
If you are using the API or AWS CLI to specify a CMK, be aware that AWS authenticates the
CMK asynchronously. If you specify a key ID, an alias, or an ARN that is not valid, the action can
appear to complete, but it eventually fails.

You cannot change the CMK that is associated with an existing snapshot or volume. However, you can
associate a different CMK during a snapshot copy operation so that the resulting copied snapshot is
encrypted by the new CMK.

Encrypt an empty volume on creation

When you create a new, empty EBS volume, you can encrypt it by enabling encryption for the specific
volume creation operation. If you enabled EBS encryption by default, the volume is automatically
encrypted. By default, the volume is encrypted to your default key for EBS encryption. Alternatively,
you can specify a different symmetric CMK for the specific volume creation operation. The volume is
encrypted by the time it is first available, so your data is always secured. For detailed procedures, see
Create an Amazon EBS volume (p. 1104).

1179
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

By default, the CMK that you selected when creating a volume encrypts the snapshots that you make
from the volume and the volumes that you restore from those encrypted snapshots. You cannot remove
encryption from an encrypted volume or snapshot, which means that a volume restored from an
encrypted snapshot, or a copy of an encrypted snapshot, is always encrypted.

Public snapshots of encrypted volumes are not supported, but you can share an encrypted snapshot with
specific accounts. For detailed directions, see Share an Amazon EBS snapshot (p. 1139).

Encrypt unencrypted resources

Although there is no direct way to encrypt an existing unencrypted volume or snapshot, you can encrypt
them by creating either a volume or a snapshot. If you enabled encryption by default, Amazon EBS
encrypts the resulting new volume or snapshot using your default key for EBS encryption. Even if you
have not enabled encryption by default, you can enable encryption when you create an individual
volume or snapshot. Whether you enable encryption by default or in individual creation operations,
you can override the default key for EBS encryption and select a symmetric customer managed
CMK. For more information, see Create an Amazon EBS volume (p. 1104) and Copy an Amazon EBS
snapshot (p. 1134).

To encrypt the snapshot copy to a customer managed CMK, you must both enable encryption
and specify the key, as shown in Copy an unencrypted snapshot (encryption by default not
enabled) (p. 1182).
Important
Amazon EBS does not support asymmetric CMKs. For more information, see Using Symmetric
and Asymmetric Keys in the AWS Key Management Service Developer Guide.

You can also apply new encryption states when launching an instance from an EBS-backed AMI. This is
because EBS-backed AMIs include snapshots of EBS volumes that can be encrypted as described. For
more information, see Use encryption with EBS-backed AMIs (p. 161).

Encryption scenarios
When you create an encrypted EBS resource, it is encrypted by your account's default key for EBS
encryption unless you specify a different customer managed CMK in the volume creation parameters
or the block device mapping for the AMI or instance. For more information, see Default key for EBS
encryption (p. 1178).

The following examples illustrate how you can manage the encryption state of your volumes and
snapshots. For a full list of encryption cases, see the encryption outcomes table (p. 1184).

Examples
• Restore an unencrypted volume (encryption by default not enabled) (p. 1180)
• Restore an unencrypted volume (encryption by default enabled) (p. 1181)
• Copy an unencrypted snapshot (encryption by default not enabled) (p. 1182)
• Copy an unencrypted snapshot (encryption by default enabled) (p. 1182)
• Re-encrypt an encrypted volume (p. 1183)
• Re-encrypt an encrypted snapshot (p. 1184)
• Migrate data between encrypted and unencrypted volumes (p. 1184)
• Encryption outcomes (p. 1184)

Restore an unencrypted volume (encryption by default not enabled)

Without encryption by default enabled, a volume restored from an unencrypted snapshot is unencrypted
by default. However, you can encrypt the resulting volume by setting the Encrypted parameter and,
optionally, the KmsKeyId parameter. The following diagram illustrates the process.

1180
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

If you leave out the KmsKeyId parameter, the resulting volume is encrypted using your default key for
EBS encryption. You must specify a key ID to encrypt the volume to a different CMK.

For more information, see Create a volume from a snapshot (p. 1105).

Restore an unencrypted volume (encryption by default enabled)

When you have enabled encryption by default, encryption is mandatory for volumes restored from
unencrypted snapshots, and no encryption parameters are required for your default CMK to be used. The
following diagram shows this simple default case:

If you want to encrypt the restored volume to a symmetric customer managed CMK, you must
supply both the Encrypted and KmsKeyId parameters as shown in Restore an unencrypted volume
(encryption by default not enabled) (p. 1180).

1181
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Copy an unencrypted snapshot (encryption by default not enabled)

Without encryption by default enabled, a copy of an unencrypted snapshot is unencrypted by default.

However, you can encrypt the resulting snapshot by setting the Encrypted parameter and, optionally,
the KmsKeyId parameter. If you omit KmsKeyId, the resulting snapshot is encrypted by your default
CMK. You must specify a key ID to encrypt the volume to a different symmetric CMK.

The following diagram illustrates the process.

Note
If you copy a snapshot and encrypt it to a new CMK, a complete (non-incremental) copy is
always created, resulting in additional delay and storage costs.

You can encrypt an EBS volume by copying an unencrypted snapshot to an encrypted snapshot and
then creating a volume from the encrypted snapshot. For more information, see Copy an Amazon EBS
snapshot (p. 1134).

Copy an unencrypted snapshot (encryption by default enabled)

When you have enabled encryption by default, encryption is mandatory for copies of unencrypted
snapshots, and no encryption parameters are required if your default CMK is used. The following
diagram illustrates this default case:

1182
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Note
If you copy a snapshot and encrypt it to a new CMK, a complete (non-incremental) copy is
always created, resulting in additional delay and storage costs.

Re-encrypt an encrypted volume

When the CreateVolume action operates on an encrypted snapshot, you have the option of re-
encrypting it with a different CMK. The following diagram illustrates the process. In this example, you
own two CMKs, CMK A and CMK B. The source snapshot is encrypted by CMK A. During volume creation,
with the key ID of CMK B specified as a parameter, the source data is automatically decrypted, then re-
encrypted by CMK B.

Note
If you copy a snapshot and encrypt it to a new CMK, a complete (non-incremental) copy is
always created, resulting in additional delay and storage costs.

For more information, see Create a volume from a snapshot (p. 1105).

1183
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Re-encrypt an encrypted snapshot

The ability to encrypt a snapshot during copying allows you to apply a new symmetric CMK to an
already-encrypted snapshot that you own. Volumes restored from the resulting copy are only accessible
using the new CMK. The following diagram illustrates the process. In this example, you own two CMKs,
CMK A and CMK B. The source snapshot is encrypted by CMK A. During copy, with the key ID of CMK B
specified as a parameter, the source data is automatically re-encrypted by CMK B.

Note
If you copy a snapshot and encrypt it to a new CMK, a complete (non-incremental) copy is
always created, resulting in additional delay and storage costs.

In a related scenario, you can choose to apply new encryption parameters to a copy of a snapshot that
has been shared with you. By default, the copy is encrypted with a CMK shared by the snapshot's owner.
However, we recommend that you create a copy of the shared snapshot using a different CMK that you
control. This protects your access to the volume if the original CMK is compromised, or if the owner
revokes the CMK for any reason. For more information, see Encryption and snapshot copying (p. 1136).

Migrate data between encrypted and unencrypted volumes

When you have access to both an encrypted and unencrypted volume, you can freely transfer data
between them. EC2 carries out the encryption and decryption operations transparently.

For example, use the rsync command to copy the data. In the following command, the source data is
located in /mnt/source and the destination volume is mounted at /mnt/destination.

[ec2-user ~]$ sudo rsync -avh --progress /mnt/source/ /mnt/destination/

Encryption outcomes

The following table describes the encryption outcome for each possible combination of settings.

Is Is encryption Source of volume Default (no CMK Custom (CMK

encryption by default specified) specified)
enabled? enabled?

No No New (empty) volume Unencrypted N/A

1184
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Is Is encryption Source of volume Default (no CMK Custom (CMK

encryption by default specified) specified)
enabled? enabled?

No No Unencrypted snapshot Unencrypted

that you own

No No Encrypted snapshot that Encrypted by same

you own key

No No Unencrypted snapshot Unencrypted

that is shared with you

No No Encrypted snapshot that is Encrypted by

shared with you default CMK*

Yes No New volume Encrypted by Encrypted by a

default CMK specified CMK**

Yes No Unencrypted snapshot Encrypted by

that you own default CMK

Yes No Encrypted snapshot that Encrypted by same

you own key

Yes No Unencrypted snapshot Encrypted by

that is shared with you default CMK

Yes No Encrypted snapshot that is Encrypted by

shared with you default CMK

No Yes New (empty) volume Encrypted by N/A

default CMK

No Yes Unencrypted snapshot Encrypted by

that you own default CMK

No Yes Encrypted snapshot that Encrypted by same

you own key

No Yes Unencrypted snapshot Encrypted by

that is shared with you default CMK

No Yes Encrypted snapshot that is Encrypted by

shared with you default CMK

Yes Yes New volume Encrypted by Encrypted by a

default CMK specified CMK

Yes Yes Unencrypted snapshot Encrypted by

that you own default CMK

Yes Yes Encrypted snapshot that Encrypted by same

you own key

Yes Yes Unencrypted snapshot Encrypted by

that is shared with you default CMK

Yes Yes Encrypted snapshot that is Encrypted by

shared with you default CMK

1185
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

* This is the default CMK used for EBS encryption for the AWS account and Region. By default this is a
unique AWS managed CMK for EBS, or you can specify a customer managed CMK. For more information,
see Default key for EBS encryption (p. 1178).

** This is a customer managed CMK specified for the volume at launch time. This CMK is used instead of
the default CMK for the AWS account and Region.

Set encryption defaults using the API and CLI

You can manage encryption by default and the default customer master key (CMK) using the following
API actions and CLI commands.

API action CLI command Description

DisableEbsEncryptionByDefault disable-ebs-encryption-by-default Disables encryption by

default.

EnableEbsEncryptionByDefault enable-ebs-encryption-by-default Enables encryption by

default.

GetEbsDefaultKmsKeyId get-ebs-default-kms-key-id Describes the default

CMK.

GetEbsEncryptionByDefault get-ebs-encryption-by-default Indicates whether

encryption by default is
enabled.

ModifyEbsDefaultKmsKeyId modify-ebs-default-kms-key-id Changes the default

CMK used to encrypt
EBS volumes.

ResetEbsDefaultKmsKeyId reset-ebs-default-kms-key-id Resets the AWS

managed default CMK
as the default CMK
used to encrypt EBS
volumes.

Amazon Data Lifecycle Manager

You can use Amazon Data Lifecycle Manager to automate the creation, retention, and deletion of EBS
snapshots and EBS-backed AMIs. When you automate snapshot and AMI management, it helps you to:

• Protect valuable data by enforcing a regular backup schedule.

• Create standardized AMIs that can be refreshed at regular intervals.
• Retain backups as required by auditors or internal compliance.
• Reduce storage costs by deleting outdated backups.
• Create disaster recovery backup policies that back up data to isolated accounts.

When combined with the monitoring features of Amazon CloudWatch Events and AWS CloudTrail,
Amazon Data Lifecycle Manager provides a complete backup solution for Amazon EC2 instances and
individual EBS volumes at no additional cost.
Important
Amazon Data Lifecycle Manager cannot be used to manage snapshots or AMIs that are created
by any other means.

1186
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Amazon Data Lifecycle Manager cannot be used to automate the creation, retention, and
deletion of instance store-backed AMIs.

Contents
• How Amazon Data Lifecycle Manager works (p. 1187)
• Considerations for Amazon Data Lifecycle Manager (p. 1189)
• Prerequisites (p. 1191)
• Automate snapshot and EBS-backed AMI lifecycles (p. 1194)
• Automate cross-account snapshot copies (p. 1197)
• View, modify, and delete lifecycle policies (p. 1206)
• Monitor the lifecycle of snapshots and AMIs (p. 1209)

How Amazon Data Lifecycle Manager works

The following are the key elements of Amazon Data Lifecycle Manager.

Elements
• Snapshots (p. 1187)
• EBS-backed AMIs (p. 1187)
• Target resource tags (p. 1187)
• Amazon Data Lifecycle Manager tags (p. 1188)
• Lifecycle policies (p. 1188)
• Policy schedules (p. 1189)

Snapshots

Snapshots are the primary means to back up data from your EBS volumes. To save storage costs,
successive snapshots are incremental, containing only the volume data that changed since the previous
snapshot. When you delete one snapshot in a series of snapshots for a volume, only the data that's
unique to that snapshot is removed. The rest of the captured history of the volume is preserved.

For more information, see Amazon EBS snapshots (p. 1126).

EBS-backed AMIs

An Amazon Machine Image (AMI) provides the information that's required to launch an instance. You
can launch multiple instances from a single AMI when you need multiple instances with the same
configuration. Amazon Data Lifecycle Manager supports EBS-backed AMIs only. EBS-backed AMIs include
a snapshot for each EBS volume that's attached to the source instance.

For more information, see Amazon Machine Images (AMI) (p. 99).

Target resource tags

Amazon Data Lifecycle Manager uses resource tags to identify the resources to back up. Tags are
customizable metadata that you can assign to your AWS resources (including Amazon EC2 instances, EBS
volumes and snapshots). An Amazon Data Lifecycle Manager policy (described later) targets an instance
or volume for backup using a single tag. Multiple tags can be assigned to an instance or volume if you
want to run multiple policies on it.

You can't use a '\' or '=' character in a tag key.

For more information, see Tag your Amazon EC2 resources (p. 1311).

1187
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Amazon Data Lifecycle Manager tags

Amazon Data Lifecycle Manager applies the following tags to all snapshots and AMIs created by a policy,
to distinguish them from snapshots and AMIs created by any other means:

• aws:dlm:lifecycle-policy-id
• aws:dlm:lifecycle-schedule-name
• aws:dlm:expirationTime
• dlm:managed

You can also specify custom tags to be applied to snapshots and AMIs on creation. You can't use a '\' or
'=' character in a tag key.

The target tags that Amazon Data Lifecycle Manager uses to associate volumes with a snapshot policy
can optionally be applied to snapshots created by the policy. Similarly, the target tags that are used to
associate instances with an AMI policy can optionally be applied to AMIs created by the policy.

Lifecycle policies
A lifecycle policy consists of these core settings:

• Policy type—Defines the type of resources that the policy can manage. Amazon Data Lifecycle
Manager supports two types of lifecycle policies:
• Snapshot lifecycle policy—Used to automate the lifecycle of EBS snapshots. These policies can
target EBS volumes and instances.
• EBS-backed AMI lifecycle policy—Used to automate the lifecycle of EBS-backed AMIs. These policies
can target instances only.
• Cross-account copy event policy—Used to automate the copying of snapshots accross accounts. This
policy type should be used in conjunction with an EBS snapshot policy that shares snapshots accross
accounts.
• Resource type—Defines the type of resources that are targeted by the policy. Snapshot lifecycle
policies can target instances or volumes. Use VOLUME to create snapshots of individual volumes, or use
INSTANCE to create multi-volume snapshots of all of the volumes that are attached to an instance. For
more information, see Multi-volume snapshots (p. 1130). AMI lifecycle policies can target instances
only. One AMI is created that includes snapshots of all of the volumes that are attached to the target
instance.
• Target tags—Specifies the tags that must be assigned to an EBS volume or an Amazon EC2 instance
for it to be targeted by the policy.
• Schedules—The start times and intervals for creating snapshots or AMIs. The first snapshot or
AMI is created by a policy within one hour after the specified start time. Subsequent snapshots or
AMIs are created within one hour of their scheduled time. A policy can have up to four schedules:
one mandatory schedule, and up to three optional schedules. For more information, see Policy
schedules (p. 1189).
• Retention—Specifies how snapshots or AMIs are to be retained. You can retain snapshots or AMIs
based either on their total count (count-based), or their age (age-based). For snapshot policies, when
the retention threshold is reached, the oldest snapshot is deleted. For AMI policies, when the retention
threshold is reached, the oldest AMI is deregistered and its backing snapshots are deleted.

For example, you could create a policy with settings similar to the following:

• Manages all EBS volumes that have a tag with a key of account and a value of finance.
• Creates snapshots every 24 hours at 0900 UTC.
• Retains only the five most recent snapshots.
• Starts snapshot creation no later than 0959 UTC each day.

1188
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Policy schedules

Policy schedules define when snapshots or AMIs are created by the policy. Policies can have up to four
schedules—one mandatory schedule, and up to three optional schedules.

Adding multiple schedules to a single policy lets you create snapshots or AMIs at different frequencies
using the same policy. For example, you can create a single policy that creates daily, weekly, monthly,
and yearly snapshots. This eliminates the need to manage multiple policies.

For each schedule, you can define the frequency, fast snapshot restore settings (snapshot lifecycle
policies only), cross-Region copy rules, and tags. The tags that are assigned to a schedule are
automatically assigned to the snapshots or AMIs that are created when the schedule is triggered. In
addition, Amazon Data Lifecycle Manager automatically assigns a system-generated tag based on the
schedule's frequency to each snapshot or AMI.

Each schedule is triggered individually based on its frequency. If multiple schedules are triggered at the
same time, Amazon Data Lifecycle Manager creates only one snapshot or AMI and applies the retention
settings of the schedule that has the highest retention period. The tags of all of the triggered schedules
are applied to the snapshot or AMI.

• (Snapshot lifecycle policies only) If more than one of the triggered schedules is enabled for fast
snapshot restore, then the snapshot is enabled for fast snapshot restore in all of the Availability
Zones specified across all of the triggered schedules. The highest retention settings of the triggered
schedules is used for each Availability Zone.
• If more than one of the triggered schedules is enabled for cross-Region copy, the snapshot or AMI is
copied to all Regions specified across all of the triggered schedules. The highest retention period of
the triggered schedules is applied.

Considerations for Amazon Data Lifecycle Manager

Your AWS account has the following quotas related to Amazon Data Lifecycle Manager:

• You can create up to 100 lifecycle policies per Region.

• You can add up to 45 tags per resource.

The following considerations apply to lifecycle policies:

• A policy does not begin creating snapshots or AMIs until you set its activation status to enabled. You
can configure a policy to be enabled upon creation.
• The first snapshot or AMI is created by a policy within one hour after the specified start time.
Subsequent snapshots or AMIs are created within one hour of their scheduled time.
• If you modify a policy by removing or changing its target tags, the EBS volumes or instances with
those tags are no longer managed by the policy.
• If you modify a schedule name for a policy, the snapshots or AMIs created under the old schedule
name are no longer affected by the policy.
• If you modify a time-based retention schedule to use a new time interval, the new interval is used only
for new snapshots or AMIs created after the change. The new schedule does not affect the retention
schedule of snapshots or AMIs created before the change.
• You cannot change the retention schedule of a policy from count-based to time-based after creation.
To make this change, you must create a new policy.
• If you disable a policy with an age-based retention schedule, the snapshots or AMIs that are set
to expire while the policy is disabled are retained indefinitely. You must delete the snapshots or
deregister the AMIs manually. When you enable the policy again, Amazon Data Lifecycle Manager
resumes deleting snapshots or deregistering AMIs as their retention periods expire.

1189
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

• If you delete the resource to which a policy with count-based retention applies, the policy no longer
manages the previously created snapshots or AMIs. You must manually delete the snapshots or
deregister the AMIs if they are no longer needed.
• If you delete the resource to which a policy with age-based retention applies, the policy continues to
delete snapshots or deregister AMIs on the defined schedule, up to the last snapshot or AMI. You must
manually delete the last snapshot or deregister the last AMI if it is no longer needed.
• You can create multiple policies to back up an EBS volume or an Amazon EC2 instance. For example,
if an EBS volume has two tags, where tag A is the target for policy A to create a snapshot every 12
hours, and tag B is the target for policy B to create a snapshot every 24 hours, Amazon Data Lifecycle
Manager creates snapshots according to the schedules for both policies. Alternatively, you can achieve
the same result by creating a single policy that has multiple schedules. For example, you can create a
single policy that targets only tag A, and specify two schedules—one for every 12 hours and one for
every 24 hours.
• If you create a policy that targets instances, and new volumes are attached to the instance after the
policy has been created, the newly-added volumes are included in the backup at the next policy run.
All volumes attached to the instance at the time of the policy run are included.
• For AMI lifecycle policies, when the AMI retention threshold is reached, the oldest AMI is deregistered
and its backing snapshots are deleted.

The following considerations apply to snapshot lifecycle policies and fast snapshot restore (p. 1210):

• A snapshot that is enabled for fast snapshot restore remains enabled even if you delete or disable the
lifecycle policy, disable fast snapshot restore for the lifecycle policy, or disable fast snapshot restore
for the Availability Zone. You can disable fast snapshot restore for these snapshots manually.
• If you enable fast snapshot restore and you exceed the maximum number of snapshots that can be
enabled for fast snapshot restore, Amazon Data Lifecycle Manager creates snapshots as scheduled
but does not enable them for fast snapshot restore. After a snapshot that is enabled for fast snapshot
restore is deleted, the next snapshot that Amazon Data Lifecycle Manager creates is enabled for fast
snapshot restore.
• When you enable fast snapshot restore for a snapshot, it takes 60 minutes per TiB to optimize
the snapshot. We recommend that you create a schedule that ensures that each snapshot is fully
optimized before Amazon Data Lifecycle Manager creates the next snapshot.
• You are billed for each minute that fast snapshot restore is enabled for a snapshot in a particular
Availability Zone. Charges are pro-rated with a minimum of one hour. For more information, see
Pricing and Billing (p. 1214).
Note
Depending on the configuration of your lifecycle policies, you could have multiple snapshots
enabled for fast snapshot restore simultaneously.

The following considerations apply to snapshot lifecycle policies and Multi-Attach (p. 1107) enabled
volumes:

• When creating a lifecycle policy based on instance tags for Multi-Volume snapshots, Amazon Data
Lifecycle Manager initiates a snapshot of the volume for each attached instance. Use the timestamp
tag to identify the set of time-consistent snapshots that are created from the attached instances.

The following considerations apply to sharing snasphots across accounts:

• You can only share snapshots that are unencrypted or that are encrypted using a customer-managed
CMK.
• You can't share snapshots that are encrypted with the default EBS encryption key.
• If you share encrypted snanpshots, then you must also share the CMK that was used to encrypt the
source volume with the target accounts.

1190
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

The following considerations apply to cross-account copy event policies:

• You can only copy snapshots that are unencrypted or that are encrypted using a customer-managed
CMK.
• You can create a cross-account copy event policy that copies snapshots that are shared outside of
Amazon Data Lifecycle Manager.
• If you want to encrypt snapshots in the target account, then the IAM role selected for the cross-
account copy event policy must have permission to use the required CMK.

Prerequisites
The following prerequisites are required by Amazon Data Lifecycle Manager.

Prerequisites
• Permissions for Amazon Data Lifecycle Manager (p. 1191)
• Permissions for IAM users (p. 1193)
• Permissions for encryption (p. 1193)

Permissions for Amazon Data Lifecycle Manager

Amazon Data Lifecycle Manager uses IAM roles to get the permissions that are required to manage
snapshots and AMIs on your behalf. Amazon Data Lifecycle Manager creates the following default roles
the first time you create a lifecycle policy using the AWS Management Console.

• AWSDataLifecycleManagerDefaultRole—default role for managing snapshots. It is created the first

time you create a snapshot lifecycle policy using the console.
• AWSDataLifecycleManagerDefaultRoleForAMIManagement—default role for managing AMIs. It is
created the first time you create an AMI lifecycle policy using the console.

You can also create this role manually using the create-default-role command. For --resource-type,
specify one of the following, depending on the role to create:

• snapshot—to create the default role for managing snapshot lifecycle policies
• image—to create the default role for managing AMI lifecycle policies

aws dlm create-default-role --resource-type snapshot|image

Alternatively, you can create custom IAM roles with the required permissions and select them when you
create a lifecycle policy.

To create a custom IAM role

1. Create roles with the following permissions.

• Permissions for managing snapshot lifecycle policies

"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"ec2:CreateSnapshot",
"ec2:CreateSnapshots",
"ec2:DeleteSnapshot",

1191
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

"ec2:DescribeInstances",
"ec2:DescribeVolumes",
"ec2:DescribeSnapshots",
"ec2:EnableFastSnapshotRestores",
"ec2:DescribeFastSnapshotRestores",
"ec2:DisableFastSnapshotRestores",
"ec2:CopySnapshot",
"ec2:ModifySnapshotAttribute",
"ec2:DescribeSnapshotAttribute"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": [
"ec2:CreateTags"
],
"Resource": "arn:aws:ec2:*::snapshot/*"
},
{
"Effect": "Allow",
"Action": [
"events:PutRule",
"events:DeleteRule",
"events:DescribeRule",
"events:EnableRule",
"events:DisableRule",
"events:ListTargetsByRule",
"events:PutTargets",
"events:RemoveTargets"
],
"Resource": "arn:aws:events:*:*:rule/AwsDataLifecycleRule.managed-cwe.*"
}
]
}

• Permissions for managing AMI lifecycle policies

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": "ec2:CreateTags",
"Resource": [
"arn:aws:ec2:*::snapshot/*",
"arn:aws:ec2:*::image/*"
]
},
{
"Effect": "Allow",
"Action": [
"ec2:DescribeImages",
"ec2:DescribeInstances",
"ec2:DescribeImageAttribute",
"ec2:DescribeVolumes",
"ec2:DescribeSnapshots"
],
"Resource": "*"
},
{
"Effect": "Allow",
"Action": "ec2:DeleteSnapshot",
"Resource": "arn:aws:ec2:*::snapshot/*"
},

1192
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

{
"Effect": "Allow",
"Action": [
"ec2:ResetImageAttribute",
"ec2:DeregisterImage",
"ec2:CreateImage",
"ec2:CopyImage",
"ec2:ModifyImageAttribute"
],
"Resource": "*"
}]
}

For more information, see Creating a Role in the IAM User Guide.
2. Add a trust relationship to the roles.

a. In the IAM console, choose Roles.

b. Select the roles that you created and then choose Trust relationships.
c. Choose Edit Trust Relationship, add the following policy, and then choose Update Trust Policy.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Principal": {
"Service": "dlm.amazonaws.com"
},
"Action": "sts:AssumeRole"
} ]
}

Permissions for IAM users

An IAM user must have the following permissions to use Amazon Data Lifecycle Manager.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": ["iam:PassRole", "iam:ListRoles"],
"Resource": "arn:aws:iam::123456789012:role/AWSDataLifecycleManagerDefaultRole"
},
{
"Effect": "Allow",
"Action": "dlm:*",
"Resource": "*"
}]
}

For more information, see Changing Permissions for an IAM User in the IAM User Guide.

Permissions for encryption

If the source volume is encrypted, ensure that the Amazon Data Lifecycle
Manager default roles (AWSDataLifecycleManagerDefaultRole and
AWSDataLifecycleManagerDefaultRoleForAMIManagement) have permission to use the AWS KMS
customer master keys (CMKs) used to encrypt the volume.

1193
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

If you enable Cross Region copy for unencrypted snapshots or AMIs backed by unencrypted snapshots,
and choose to enable encryption in the destination Region, ensure that the default roles have permission
to use the CMK needed to perform the encryption in the destination Region.

If you enable Cross Region copy for encrypted snapshots or AMIs backed by encrypted snapshots, ensure
that the default roles have permission to use both the source and destination CMKs.

For more information, see Managing access to AWS KMS CMKs in the AWS Key Management Service
Developer Guide.

Automate snapshot and EBS-backed AMI lifecycles

The following procedures show how to use Amazon Data Lifecycle Manager to manage the backups of
your EBS volumes and Amazon EC2 instances.

Use one of the following procedure to create a lifecycle policy.

Console

To create a lifecycle policy

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic Block Store, Lifecycle Manager, and then choose Create
lifecycle policy.
3. Provide the following information for your policy as needed:

• Description—A description of the policy.

• Policy type—The type of resource to be managed by this policy. To create a policy that
automates the lifecycle of snapshots, choose EBS snapshot policy. To create a policy that
automates the lifecycle of EBS-backed AMIs, choose EBS-backed AMI policy.
• Resource type—(Snapshot lifecycle policies only) The type of resource to back up. Choose
Volume to create snapshots of individual volumes or choose Instance to create multi-
volume snapshots from the volumes attached to an instance. AMI lifecycle policies can back
up instances only.
• Target with these tags—The resource tags that identify the volumes or instances to back up.
• Lifecycle policy tags—The tags to apply to the lifecycle policy.
4. For IAM role, choose the IAM role that has permissions to create, delete, and describe snapshots
or AMIs, depending on the selected policy type, and to describe volumes and instances. AWS
provides a default roles, or you can create a custom IAM role.
5. Add the policy schedules. Schedule 1 is mandatory. Schedules 2, 3, and 4 are optional. For each
policy schedule, specify the following information:

• Schedule name—A name for the schedule.

• Frequency—The interval between policy runs. You can configure policy runs on a daily,
weekly, monthly, or yearly schedule. Alternatively, choose Custom cron expression to specify
an interval of up to one year. For more information, see Cron expressions in the Amazon
CloudWatch Events User Guide.
• Starting at hh:mm UTC—The time at which the policy runs are scheduled to start. The first
policy run starts within an hour after the scheduled time.
• Retention type—You can retain snapshots or AMIs based on either their total count or their
age. For count-based retention, the range is 1 to 1000. After the maximum count is reached,
the oldest snapshot or AMI is deleted when a new one is created. For age-based retention, the
range is 1 day to 100 years. After the retention period of each snapshot or AMI expires, it is
deleted. The retention period should be greater than or equal to the creation interval.

1194
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Note
All schedules must have the same retention type. You can specify the retention type
for Schedule 1 only. Schedules 2, 3, and 4 inherit the retention type from Schedule 1.
Each schedule can have its own retention count or period.
• Copy tags from source—Choose whether to copy all of the user-defined tags from the source
resource to the snapshots or the AMIs created by this schedule. For snapshot lifecycle policies,
the tags assigned to the source volume are assigned to the snapshot. For AMI lifecycle
policies, the tags assigned to the source instance are assigned to the AMI.
• Dynamic tags—If the source resource is an instance, you can choose to automatically tag your
snapshots or AMIs with the following variable tags:
• instance-id—The ID of the source instance.
• timestamp—(Snapshot lifecycle policies only) The date and time of the policy run.
• Additional tags—Specify any additional tags to assign to the snapshots or AMIs created by
this schedule.
• Fast snapshot restore—(Snapshot lifecycle policies only) Choose whether to enable fast
snapshot restore for all snapshots that are created by this policy. If you enable fast snapshot
restore, you must choose the Availability Zones in which to enable it. You are billed for each
minute that fast snapshot restore is enabled for a snapshot in a particular Availability Zone.
Charges are pro-rated with a minimum of one hour. You can also specify the maximum
number of snapshots that can be enabled for fast snapshot restore.
• Enable cross Region copy—You can copy each snapshot or AMI to up at three additional
Regions. You must ensure that you do not exceed the number of concurrent snapshot or AMI
copies per Region. For each Region, you can choose different retention policies and you can
choose whether to copy all tags or no tags. If the source snapshot or AMI is encrypted, or if
encryption by default is enabled, the copied snapshots or AMIs are encrypted. If the source
snapshot or AMI is unencrypted, you can enable encryption. If you do not specify a CMK, the
snapshots or AMIs are encrypted using the default key for EBS encryption in each destination
Region. If you specify a CMK for the destination Region, you must have access to the CMK.
6. (AMI lifecycle policies only) Indicate whether instances should be rebooted before AMI creation.
To prevent the targeted instances from being rebooted, for Reboot Instance at policy run,
choose No. Choosing this option could cause data consistency issues. To reboot instances
before AMI creation, for Reboot Instance at policy run, choose Yes. Choosing this ensures data
consistency but could result in multiple targeted instances rebooting simultaneously.
7. For Policy status after creation, choose Enable policy to start the policy runs at the next
scheduled time, or Disable policy to prevent the policy from running.
8. Choose Create Policy.

Command line

Use the create-lifecycle-policy command to create a lifecycle policy. To create a snapshot lifecycle
policy, for PolicyType, specify EBS_SNAPSHOT_MANAGEMENT. To create an AMI lifecycle policy, for
PolicyType, specify IMAGE_MANAGEMENT.

To simplify the syntax, the following examples use a JSON file, policyDetails.json, that
includes the policy details.

Example 1—Snapshot lifecycle policy

This example creates a snapshot lifecycle policy that creates snapshots of all volumes that have a
tag key of costcenter with a value of 115. The policy includes two schedules. The first schedule
creates a snapshot every day at 03:00 UTC. The second schedule creates a weekly snapshot every
Friday at 17:00 UTC.

1195
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

aws dlm create-lifecycle-policy --description "My volume policy" --state ENABLED --

execution-role-arn arn:aws:iam::12345678910:role/AWSDataLifecycleManagerDefaultRole --
policy-details file://policyDetails.json

The following is an example of the policyDetails.json file.

{
"PolicyType": "EBS_SNAPSHOT_MANAGEMENT",
"ResourceTypes": [
"VOLUME"
],
"TargetTags": [{
"Key": "costcenter",
"Value": "115"
}],
"Schedules": [{
"Name": "DailySnapshots",
"TagsToAdd": [{
"Key": "type",
"Value": "myDailySnapshot"
}],
"CreateRule": {
"Interval": 24,
"IntervalUnit": "HOURS",
"Times": [
"03:00"
]
},
"RetainRule": {
"Count": 5
},
"CopyTags": false
},
{
"Name": "WeeklySnapshots",
"TagsToAdd": [{
"Key": "type",
"Value": "myWeeklySnapshot"
}],
"CreateRule": {
"CronExpression": "cron(0 0 17 ? * FRI *)"
},
"RetainRule": {
"Count": 5
},
"CopyTags": false
}
]}

Upon success, the command returns the ID of the newly created policy. The following is example
output.

{
"PolicyId": "policy-0123456789abcdef0"
}

Example 2—AMI lifecycle policy

This example creates an AMI lifecycle policy that creates AMIs of all instances that have a tag key of
purpose with a value of production without rebooting the targeted instances. The policy includes
one schedule that creates an AMI every day at 01:00 UTC. The policy will retain the two most recent

1196
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

AMIs (and their backing snapshots), and it will copy the tags from the source instance to the AMIs
that it creates.

aws dlm create-lifecycle-policy --description "My

AMI policy" --state ENABLED --execution-role-arn
arn:aws:iam::12345678910:role/AWSDataLifecycleManagerDefaultRoleForAMIManagement --
policy-details file://policyDetails.json

The following is an example of the policyDetails.json file.

{
"PolicyType": "IMAGE_MANAGEMENT",
"ResourceTypes": [
"INSTANCE"
],
"TargetTags": [{
"Key": "purpose",
"Value": "production"
}],
"Schedules": [{
"Name": "DailyAMIs",
"TagsToAdd": [{
"Key": "type",
"Value": "myDailyAMI"
}],
"CreateRule": {
"Interval": 24,
"IntervalUnit": "HOURS",
"Times": [
"01:00"
]
},
"RetainRule": {
"Count": 2
},
"CopyTags": true
}
],
"Parameters" : {
"NoReboot":true
}
}

Upon success, the command returns the ID of the newly created policy. The following is example
output.

{
"PolicyId": "policy-9876543210abcdef0"
}

Automate cross-account snapshot copies

Automating cross-account snapshot copies enables you to copy your Amazon EBS snapshots to specific
Regions in an isolated account and encrypt those snapshots with an encryption key. This enables you to
protect yourself against data loss in the event of your account being compromised.

Automating cross-account snapshots copies involves two accounts:

1197
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

• Source account—The source account is the account that creates and shares the snapshots with the
target account. In this account, you must create an EBS snapshot policy that creates snapshots at set
intervals and then shares them with other AWS accounts.
• Target account—The target account is the account with destination account with which the snapshots
are shared, and it is the account that creates copies of the shared snapshots. In this account, you must
create a cross-account copy event policy that automatically copies snapshots that are shared with it by
one or more specified source accounts.

Topics
• Create cross-account snapshot copy policies (p. 1198)
• Specify snapshot description filters (p. 1206)

Create cross-account snapshot copy policies

To prepare the source and target accounts for cross-account snapshot copying, you need to perform the
following steps:

Topics
• Step 1: Create the EBS snapshot policy (Source account) (p. 1198)
• Step 2: Share the customer-managed CMK (Source account) (p. 1200)
• Step 3: Create cross-account copy event policy (Target account) (p. 1202)
• Step 4: Allow IAM role to use the required CMKs (Target account) (p. 1204)

Step 1: Create the EBS snapshot policy (Source account)

In the source account, create an EBS snapshot policy that will create the snapshots and share them with
the required target accounts.

Use one of the following methods to create the EBS snapshot policy.

Console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Lifecycle Manager and then choose Create Lifecycle Policy.
3. For Policy Type, choose EBS snapshot policy, and for Resource Type, choose Volume.
4. For Target with these tags, specify the resource tags that identify the volumes from which to
create the snapshots.
5. For IAM role, choose the IAM role that has permissions to create, share, delete, and describe
snapshots, and to describe volumes. AWS provides a default role, or you can create a custom
IAM role.

If you intend to share encrypted snapshots, then you must give this IAM role permission to use
CMK used to encrypt the source volume. For more information, see Step 2: Share the customer-
managed CMK (Source account) (p. 1200).
6. Add the policy schedules that define how snapshots are to be created and shared. Schedule 1 is
mandatory. Schedules 2, 3, and 4 are optional. For each policy schedule, specify the following
information:

• For Schedule name, enter a name for the schedule.

• For Frequency, specify the interval between policy runs. You can configure policy runs on
a daily, weekly, monthly, or yearly schedule. Alternatively, choose Custom cron expression

1198
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

to specify an interval of up to one year. For more information, see Cron expressions in the
Amazon CloudWatch Events User Guide.
• For Starting at hh:mm UTC, specify the time at which the policy runs are scheduled to start.
The first policy run starts within an hour after the scheduled time.
• For Retention type, specify how the snapshots are to be retained. You can retain snapshots
or AMIs based on either their total count or their age. For count-based retention, the range
is 1 to 1000. After the maximum count is reached, the oldest snapshot or AMI is deleted
when a new one is created. For age-based retention, the range is 1 day to 100 years. After the
retention period of each snapshot or AMI expires, it is deleted. The retention period should be
greater than or equal to the creation interval.
Note
All schedules must have the same retention type. You can specify the retention type
for Schedule 1 only. Schedules 2, 3, and 4 inherit the retention type from Schedule 1.
Each schedule can have its own retention count or period.
• Select Enable cross-account sharing, and then specify the target AWS accounts with which to
share the snapshots. These are the accounts with which the snapshots are to be shared. If you
are sharing encrypted snapshots, then you must give the selected target accounts permission
to use the CMK used to encrypt the source volume. For more information, see Step 2: Share
the customer-managed CMK (Source account) (p. 1200).
• To automatically unshare shared snapshots, select Unshare automatically, and then specify
when the snapshots are to be unshared.

If you chose to automatically unshare shared snapshots, the period after which to
automatically unshare the snapshots cannot be longer than the period for which the policy
retains its snapshots. For example, if the policy's retention configuration retains snapshots
for a period of 5 days, you can only configure the policy to automatically unshare shared
snapshots after periods up to 4 days. This applies to policies with age-based and count-based
snapshot retention configurations.
7. For Policy status after creation, choose Enable policy to start the policy runs at the next
scheduled time.
8. Choose Create Policy.

Command line

Use the create-lifecycle-policy command to create an EBS snapshot policy. To create the snapshot
policy, for PolicyType, specify EBS_SNAPSHOT_MANAGEMENT.

For example, the following command creates a snapshot lifecycle policy in source account
111111111111. The policy creates snapshots of all volumes that have a tag key of inventory_db
with a value of alpha. The policy includes one schedule that creates a snapshot every day at 15:30
UTC and retains the snapshot for two days. The policy also shares the snapshots with target account
222222222222 and then automatically unshares them after one day.

$ aws dlm create-lifecycle-policy --description "Sharing Policy" --state ENABLED --

execution-role-arn arn:aws:iam::111111111111:role/AWSDataLifecycleManagerDefaultRole --
policy-details file://policyDetails.json

The following shows the contents of the policyDetails.json file.

{
"PolicyType": "EBS_SNAPSHOT_MANAGEMENT",
"ResourceTypes": [
"VOLUME"
],
"TargetTags": [{

1199
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

"Key": "inventory_db",
"Value": "alpha"
}],
"Schedules": [{
"Name": "Daily Snapshots",
"CreateRule": {
"Interval": 24,
"IntervalUnit": "HOURS",
"Times": [
"15:30"
]
},
"RetainRule": {
"Interval": 2,
"IntervalUnit": "DAYS"
},
"ShareRules": [{
"TargetAccounts": ["222222222222"],
"UnshareInterval": 1,
"UnshareIntervalUnit": "DAYS"
}],
"CopyTags": false
}
]}

Upon success, the command returns the ID of the newly created policy. The following is example
output.

{
"PolicyId": "policy-0123456789abcdef0"
}

Step 2: Share the customer-managed CMK (Source account)

If you are sharing encrypted snapshots, you must grant the IAM role and the target AWS accounts (that
you selected in the previous step) permissions to use the customer-managed customer master key (CMK)
that was used to encrypt the source volume.
Note
Only perform this step if you are sharing encrypted snapshots. If you are sharing unencrypted
snapshots, skip this step.

Use one of the following methods to update the key policy of the CMK.

Console

1. Open the AWS KMS console at https://console.aws.amazon.com/kms.

2. To change the AWS Region, use the Region selector in the upper-right corner of the page.
3. In the navigation pane, choose Customer managed keys and then select the CMK that you need
to share with the target accounts.

Make note of the CMK's ARN, you'll need this later.

4. On the Key policy tab, scroll down to the Key users section. Choose Add, enter the name of the
IAM role that you selected in the previous step, and then choose Add.
5. On the Key policy tab, scroll down to the Other AWS accounts section. Choose Add other AWS
accounts, and then add all of the target AWS accounts that you chose to share the snapshots
with in the previous step.
6. Choose Save changes.

1200
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Command line

Use the get-key-policy command to rerieve the key policy that is currently attached to the CMK.

For example, the following command retrieves the key policy for a CMK with an ID of 9d5e2b3d-
e410-4a27-a958-19e220d83a1e and writes it to a file named snapshotKey.json.

$ aws kms get-key-policy --policy-name default --key-id 9d5e2b3d-e410-4a27-

a958-19e220d83a1e --query Policy --output text > snapshotKey.json

Open the key policy using your preferred text editor. Add the ARN of the IAM role that you specified
when you created the snapshot policy and the ARNs of the target accounts with which to share the
CMK.

For example, in the following policy, we added the ARN of the default IAM role, and the ARN of the
root account for target account 222222222222.

{
"Sid" : "Allow use of the key",
"Effect" : "Allow",
"Principal" : {
"AWS" : [
"arn:aws:iam::111111111111:role/service-role/
AWSDataLifecycleManagerDefaultRole",
"arn:aws:iam::222222222222:root"
]
},
"Action" : [
"kms:Encrypt",
"kms:Decrypt",
"kms:ReEncrypt*",
"kms:GenerateDataKey*",
"kms:DescribeKey"
],
"Resource" : "*"
},
{
"Sid" : "Allow attachment of persistent resources",
"Effect" : "Allow",
"Principal" : {
"AWS" : [
"arn:aws:iam::111111111111:role/service-role/
AWSDataLifecycleManagerDefaultRole",
"arn:aws:iam::222222222222:root"
]
},
"Action" : [
"kms:CreateGrant",
"kms:ListGrants",
"kms:RevokeGrant"
],
"Resource" : "*",
"Condition" : {
"Bool" : {
"kms:GrantIsForAWSResource" : "true"
}
}
}

Save and close the file. Then use the put-key-policy command to attach the updated key policy to
the CMK.

1201
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

$ aws kms put-key-policy --policy-name default --key-id 9d5e2b3d-e410-4a27-

a958-19e220d83a1e —policy file://snapshotKey.json

Step 3: Create cross-account copy event policy (Target account)

In the target account, you must create a cross-account copy event policy that will automatically copy
snapshots that are shared by the required source accounts.

This policy will only run in the target account when one of the specified source accounts shares snapshot
with the account.

Use one of the following methods to create the cross-account copy event policy.

Console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Lifecycle Manager and then choose Create Lifecycle Policy.
3. For Policy Type, choose Cross-account copy event policy. For Description, enter a brief
description for the policy.
4. In the Cross-account copy event settings section, for Copy snapshots shared by, enter the
source AWS accounts from which you want to copy the shared snapshots.
5. For Snapshot description filter, enter the required snapshot description using a regular
expression. Only snapshots that are shared by the specified sources accounts and that have
descriptions that match the specified filter are copied by the policy. For more information, see
Specify snapshot description filters (p. 1206).
6. For IAM role, choose the IAM role that has permissions to perform the snapshot copy action.
AWS provides a default role, or you can create a custom IAM role.

If you are copying encrypted snapshots, you must grant the selected IAM role permissions to
use the encryption CMK used to encrypt the source volume. Similarly, if you are encrypting
the snapshot in the destination Region using a different CMK, you must grant the IAM role
permission to use the destination CMK. For more information, see Step 4: Allow IAM role to use
the required CMKs (Target account) (p. 1204).
7. In the Copy settings section, you can configure the policy to copy snapshots to up to three
Regions in the target account. Do the following:

a. For Name, enter a descriptive name for the copy action.

b. For Target Region, select the Region to which to copy the snapshots.
c. For Retain copy for, specify how long to retain the snapshot copies in the target Region
after creation.
d. For Encryption, select Enable to encrypt the snapshot copy in the target Region. If the
source snapshot is encrypted, or if encryption by default is enabled for your account, the
snapshot copy is always encrypted, even if you do enable encryption here. If the source
snapshot is unencrypted and encryption by default is not enabled for your account, you can
choose to enable or disable encryption. If you enable encryption, but do not specify a CMK,
the snapshots are encrypted using the default encryption CMK in each destination Region.
If you specify a CMK for the destination Region, you must have access to the CMK.
e. (Optional) To copy the snapshot to additional Regions, choose Add additional region, and
then complete the required fields.
8. For Policy status after creation, choose Enable policy to start the policy runs at the next
scheduled time.

1202
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

9. Choose Create Policy.

Command line

Use the create-lifecycle-policy command to create a policy. To create a cross-account copy event
policy, for PolicyType, specify EVENT_BASED_POLICY.

For example, the following command creates a cross-account copy event policy in target account
222222222222. The policy copies snapshots that are shared by source account 111111111111.
The policy copies snapshots to sa-east-1 and eu-west-2. Snapshots copied to sa-east-1 are
unencrypted and they are retained for 3 days. Snapshots copied to eu-west-2 are encrypted using
CMK 8af79514-350d-4c52-bac8-8985e84171c7 and they are retained for 1 month. The policy
uses the default IAM role.

$ aws dlm create-lifecycle-policy --description "Copy policy" --state

ENABLED --execution-role-arn arn:aws:iam::222222222222:role/service-role/
AWSDataLifecycleManagerDefaultRole --policy-details file://policyDetails.json

The following shows the contents of the policyDetails.json file.

{
"PolicyType" : "EVENT_BASED_POLICY",
"EventSource" : {
"Type" : "MANAGED_CWE",
"Parameters": {
"EventType" : "shareSnapshot",
"SnapshotOwner": ["111111111111"]
}
},
"Actions" : [{
"Name" :"Copy Snapshot to Sao Paulo and London",
"CrossRegionCopy" : [{
"Target" : "sa-east-1",
"EncryptionConfiguration" : {
"Encrypted" : false
},
"RetainRule" : {
"Interval" : 3,
"IntervalUnit" : "DAYS"
}
},
{
"Target" : "eu-west-2",
"EncryptionConfiguration" : {
"Encrypted" : true,
"CmkArn" : "arn:aws:kms:eu-west-2:222222222222:key/8af79514-350d-4c52-
bac8-8985e84171c7"
},
"RetainRule" : {
"Interval" : 1,
"IntervalUnit" : "MONTHS"
}
}]
}]
}

Upon success, the command returns the ID of the newly created policy. The following is example
output.

1203
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

"PolicyId": "policy-9876543210abcdef0"
}

Step 4: Allow IAM role to use the required CMKs (Target account)

If you are copying encrypted snapshots, you must grant the IAM role (that you selected in the previous
step) permissions to use the customer-managed customer master key (CMK) that was used to encrypt
the source volume.
Note
Only perform this step if you are copying encrypted snapshots. If you are copying unencrypted
snapshots, skip this step.

Use one of the following methods to add the required policies to the IAM role.

Console

1. Open the IAM console at https://console.aws.amazon.com/iam/.

2. In the navigation pane, select Roles. Search for and select the IAM role that you selected when
you created the cross-account copy event policy in the previous step. If you chose to use the
default role, the role is named AWSDataLifecycleManagerDefaultRole.
3. Choose Add inline policy and then select the JSON tab.
4. Replace the existing policy with the following, and specify the ARNs of the CMKs.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"kms:RevokeGrant",
"kms:CreateGrant",
"kms:ListGrants"
],
"Resource": [
"arn:aws:kms:region:source_account_id:key/shared_cmk_id",
"arn:aws:kms:region:source_account_id:key/shared_cmk_id"
],
"Condition": {
"Bool": {
"kms:GrantIsForAWSResource": "true"
}
}
},
{
"Effect": "Allow",
"Action": [
"kms:Encrypt",
"kms:Decrypt",
"kms:ReEncrypt*",
"kms:GenerateDataKey*",
"kms:DescribeKey"
],
"Resource": [
"arn:aws:kms:region:source_account_id:key/shared_cmk_id",
"arn:aws:kms:region:source_account_id:key/shared_cmk_id"
]
}
]

1204
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

5. Choose Review policy

6. For Name, enter a descriptive name for the policy, and then choose Create policy.

Command line

Using your preferred text editor, create a new JSON file named policyDetails.json.
Add the following policy and specify the ARNs of the CMKs that the role needs permissions
to use. In the following example, the policy grants the IAM role permission to use CMK
1234abcd-12ab-34cd-56ef-1234567890ab, which was shared by source account
111111111111, and CMK 4567dcba-23ab-34cd-56ef-0987654321yz, which exists in target
account 222222222222.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"kms:RevokeGrant",
"kms:CreateGrant",
"kms:ListGrants"
],
"Resource": [
"arn:aws:kms:sa-
east-1:111111111111:key/1234abcd-12ab-34cd-56ef-1234567890ab",
"arn:aws:kms:eu-
west-2:222222222222:key/4567dcba-23ab-34cd-56ef-0987654321yz"
],
"Condition": {
"Bool": {
"kms:GrantIsForAWSResource": "true"
}
}
},
{
"Effect": "Allow",
"Action": [
"kms:Encrypt",
"kms:Decrypt",
"kms:ReEncrypt*",
"kms:GenerateDataKey*",
"kms:DescribeKey"
],
"Resource": [
"arn:aws:kms:sa-
east-1:111111111111:key/1234abcd-12ab-34cd-56ef-1234567890ab",
"arn:aws:kms:eu-
west-2:222222222222:key/4567dcba-23ab-34cd-56ef-0987654321yz"
]
}
]
}

Save and close the file. Then use the put-role-policy command to add the policy to the IAM role.

For example

$ aws iam put-role-policy –role-name AWSDataLifecycleManagerDefaultRole --policy-

name CopyPolicy --policy-document file://AdminPolicy.json

1205
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Specify snapshot description filters

When you create the snapshot copy policy in the target account, you must specify a snapshot description
filter. The snapshot description filter enables you to specify an additional level of filtering that lets you
control which snapshots are copied by the policy. This means that a snapshot is only copied by the policy
if it is shared by one of the specified source accounts, and it has a snapshot description that matches the
specified filter. In other words, if a snapshot is shared by one of the specified course accounts, but it does
not have a description that matches the specified filter, it is not copied by the policy.

The snapshot filter description must be specified using a regular expression. It is a mandatory field when
creating cross-account copy event policies using the console and the command line. The following are
example regular expressions that can be used:

• .*—This filter matches all snapshot descriptions. If you use this expression the policy will copy all
snapshots that are shared by one of the specified source accounts.
• Created for policy: policy-0123456789abcdef0.*—This filter matches only snapshots that
are created by a policy with an ID of policy-0123456789abcdef0. If you use an expression like this,
only snapshots that are shared with your account by one of the specified source accounts, and that
have been created by a policy with the specified ID are copied by the policy.
• .*production.*—This filter matches any snapshot that has the word production anywhere in its
description. If you use this expression the policy will copy all snapshots that are shared by one of the
specified source accounts and that have the specified text in their description.

View, modify, and delete lifecycle policies

Use the following procedures to view, modify and delete existing lifecycle policies.

Topics
• View lifecycle policies (p. 1206)
• Modify lifecycle policies (p. 1207)
• Delete lifecycle policies (p. 1110)

View lifecycle policies

Use one of the following procedures to view a lifecycle policy.

Console

To view a lifecycle policy

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic Block Store, Lifecycle Manager.
3. Select a lifecycle policy from the list. The Details tab displays information about the policy.

Command line

Use the get-lifecycle-policy command to display information about a lifecycle policy.

aws dlm get-lifecycle-policy --policy-id policy-0123456789abcdef0

The following is example output. It includes the information that you specified, plus metadata
inserted by AWS.

1206
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

"Policy":{
"Description": "My first policy",
"DateCreated": "2018-05-15T00:16:21+0000",
"State": "ENABLED",
"ExecutionRoleArn":
"arn:aws:iam::210774411744:role/AWSDataLifecycleManagerDefaultRole",
"PolicyId": "policy-0123456789abcdef0",
"DateModified": "2018-05-15T00:16:22+0000",
"PolicyDetails": {
"PolicyType":"EBS_SNAPSHOT_MANAGEMENT",
"ResourceTypes": [
"VOLUME"
],
"TargetTags": [
{
"Value": "115",
"Key": "costcenter"
}
],
"Schedules": [
{
"TagsToAdd": [
{
"Value": "myDailySnapshot",
"Key": "type"
}
],
"RetainRule": {
"Count": 5
},
"CopyTags": false,
"CreateRule": {
"Interval": 24,
"IntervalUnit": "HOURS",
"Times": [
"03:00"
]
},
"Name": "DailySnapshots"
}
]
}
}
}

Modify lifecycle policies

Use one of the following procedures to modify a lifecycle policy.

Console

To modify a lifecycle policy

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Elastic Block Store, Lifecycle Manager.
3. Select a lifecycle policy from the list.
4. Choose Actions, Modify Lifecycle Policy.
5. Modify the policy settings as needed. For example, you can modify the schedule, add or remove
tags, or enable or disable the policy.
6. Choose Update policy.

1207
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

Command line

Use the update-lifecycle-policy command to modify the information in a lifecycle policy. To simplify
the syntax, this example references a JSON file, policyDetailsUpdated.json, that includes the
policy details.

aws dlm update-lifecycle-policy --state DISABLED --execution-role-arn

arn:aws:iam::12345678910:role/AWSDataLifecycleManagerDefaultRole" --policy-details
file://policyDetailsUpdated.json

The following is an example of the policyDetailsUpdated.json file.

{
"ResourceTypes":[
"VOLUME"
],
"TargetTags":[
{
"Key": "costcenter",
"Value": "120"
}
],
"Schedules":[
{
"Name": "DailySnapshots",
"TagsToAdd": [
{
"Key": "type",
"Value": "myDailySnapshot"
}
],
"CreateRule": {
"Interval": 12,
"IntervalUnit": "HOURS",
"Times": [
"15:00"
]
},
"RetainRule": {
"Count" :5
},
"CopyTags": false
}
]
}

To view the updated policy, use the get-lifecycle-policy command. You can see that the state,
the value of the tag, the snapshot interval, and the snapshot start time were changed.

Delete lifecycle policies

Use one of the following procedures to delete a lifecycle policy.

Note
You can delete snapshots created only by Amazon Data Lifecycle Manager.

Old console

To delete a lifecycle policy

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

1208
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

2. In the navigation pane, choose Elastic Block Store, Lifecycle Manager.

3. Select a lifecycle policy from the list.
4. Choose Actions, Delete Lifecycle Policy.
5. When prompted for confirmation, choose Delete Lifecycle Policy.

Command line

Use the delete-lifecycle-policy command to delete a lifecycle policy and free up the target tags
specified in the policy for reuse.
Note
You can delete snapshots created only by Amazon Data Lifecycle Manager.

aws dlm delete-lifecycle-policy --policy-id policy-0123456789abcdef0

The Amazon Data Lifecycle Manager API Reference provides descriptions and syntax for each of the
actions and data types for the Amazon Data Lifecycle Manager Query API.

Alternatively, you can use one of the AWS SDKs to access the API in a way that's tailored to the
programming language or platform that you're using. For more information, see AWS SDKs.

Monitor the lifecycle of snapshots and AMIs

You can use the following features to monitor the lifecycle of your snapshots and AMIs.

Features
• Console and AWS CLI (p. 1209)
• CloudWatch Events (p. 1209)
• AWS CloudTrail (p. 1210)

Console and AWS CLI

You can view your lifecycle policies using the Amazon EC2 console or the AWS CLI. Each snapshot and
AMI created by a policy has a timestamp and policy-related tags. You can filter snapshots and AMIs using
these tags to verify that your backups are being created as you intend. For information about viewing
lifecycle policies using the console, see View lifecycle policies (p. 1206).

CloudWatch Events
Amazon EBS and Amazon Data Lifecycle Manager emit events related to lifecycle policy actions. You
can use AWS Lambda and Amazon CloudWatch Events to handle event notifications programmatically.
Events are emitted on a best effort basis. For more information, see the Amazon CloudWatch Events User
Guide.

The following events are available:

Note
No events are emitted for AMI lifecycle policy actions.

• createSnapshot—An Amazon EBS event emitted when a CreateSnapshot action succeeds or fails.
For more information, see Amazon CloudWatch Events for Amazon EBS (p. 1258).
• DLM Policy State Change—An Amazon Data Lifecycle Manager event emitted when a lifecycle
policy enters an error state. The event contains a description of what caused the error. The following is
an example of an event when the permissions granted by the IAM role are insufficient.

1209
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

"version": "0",
"id": "01234567-0123-0123-0123-0123456789ab",
"detail-type": "DLM Policy State Change",
source": "aws.dlm",
"account": "123456789012",
"time": "2018-05-25T13:12:22Z",
"region": "us-east-1",
"resources": [
"arn:aws:dlm:us-east-1:123456789012:policy/policy-0123456789abcdef"
],
"detail": {
"state": "ERROR",
"cause": "Role provided does not have sufficient permissions",
"policy_id": "arn:aws:dlm:us-east-1:123456789012:policy/policy-0123456789abcdef"
}
}

The following is an example of an event when a limit is exceeded.

{
"version": "0",
"id": "01234567-0123-0123-0123-0123456789ab",
"detail-type": "DLM Policy State Change",
"source": "aws.dlm",
"account": "123456789012",
"time": "2018-05-25T13:12:22Z",
"region": "us-east-1",
"resources": [
"arn:aws:dlm:us-east-1:123456789012:policy/policy-0123456789abcdef"
],
"detail":{
"state": "ERROR",
"cause": "Maximum allowed active snapshot limit exceeded",
"policy_id": "arn:aws:dlm:us-east-1:123456789012:policy/policy-0123456789abcdef"
}
}

AWS CloudTrail
With AWS CloudTrail, you can track user activity and API usage to demonstrate compliance with internal
policies and regulatory standards. For more information, see the AWS CloudTrail User Guide.

Amazon EBS fast snapshot restore

Amazon EBS fast snapshot restore enables you to create a volume from a snapshot that is fully initialized
at creation. This eliminates the latency of I/O operations on a block when it is accessed for the first
time. Volumes that are created using fast snapshot restore instantly deliver all of their provisioned
performance.

To get started, enable fast snapshot restore for specific snapshots in specific Availability Zones. Each
snapshot and Availability Zone pair refers to one fast snapshot restore. When you create a volume
from one of these snapshots in one of its enabled Availability Zones, the volume is restored using fast
snapshot restore.

You can enable fast snapshot restore for snapshots that you own and for public and private snapshots
that are shared with you.

Contents
• Fast snapshot restore quotas (p. 1211)
• Fast snapshot restore states (p. 1211)

1210
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

• Volume creation credits (p. 1211)

• Manage fast snapshot restore (p. 1212)
• View snapshots with fast snapshot restore enabled (p. 1212)
• View volumes restored using fast snapshot restore (p. 1213)
• Monitor fast snapshot restore (p. 1214)
• Pricing and Billing (p. 1214)

Fast snapshot restore quotas

You can enable up to 50 snapshots for fast snapshot restore per Region. The quota applies to snapshots
that you own and snapshots that are shared with you. If you enable fast snapshot restore for a snapshot
that is shared with you, it counts towards your fast snapshot restore quota. It does not count towards the
snapshot owner's fast snapshot restore quota.

Fast snapshot restore states

After you enable fast snapshot restore for a snapshot, it can be in one of the following states.

• enabling — A request was made to enable fast snapshot restore.

• optimizing — Fast snapshot restore is being enabled. It takes 60 minutes per TiB to optimize a
snapshot.
• enabled — Fast snapshot restore is enabled.
• disabling — A request was made to disable fast snapshot restore, or a request to enable fast
snapshot restore failed.
• disabled — Fast snapshot restore is disabled. You can enable fast snapshot restore again as needed.

Volume creation credits

The number of volumes that receive the full performance benefit of fast snapshot restore is determined
by the volume creation credits for the snapshot. There is one credit bucket per snapshot per Availability
Zone. Each volume that you create from a snapshot with fast snapshot restore enabled consumes one
credit from the credit bucket.

When you enable fast snapshot restore for a snapshot that is shared with you, you get a separate credit
bucket for the shared snapshot in your account. If you create volumes from the shared snapshot, the
credits are consumed from your credit bucket; they are not consumed from the snapshot owner's credit
bucket.

The size of a credit bucket depends on the size of the snapshot, not the size of the volumes created from
the snapshot. The size of the credit bucket for each snapshot is calculated as follows:

MAX (1, MIN (10, FLOOR(1024/snapshot_size_gib)))

As you consume credits, the credit bucket is refilled over time. The refill rate for each credit bucket is
calculated as follows:

MIN (10, 1024/snapshot_size_gib)

For example, if you enable fast snapshot restore for a snapshot with a size of 100 GiB, the maximum size
of its credit bucket is 10 credits and the refill rate is 10 credits per hour. When the credit bucket is full,
you can create 10 initialized volumes from this snapshot simultaneously.

You can use Cloudwatch metrics to monitor the size of your credit buckets and the number of credits
available in each bucket. For more information, see Fast snapshot restore metrics (p. 1256).

1211
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

After you create a volume from a snapshot with fast snapshot restore enabled, you can describe the
volume using describe-volumes and check the fastRestored field in the output to determine whether
the volume was created as an initialized volume using fast snapshot restore.

Manage fast snapshot restore

Fast snapshot restore is disabled for a snapshot by default. You can enable or disable fast snapshot
restore for snapshots that you own and for snapshots that are shared with you. When you enable or
disable fast snapshot restore for a snapshot, the changes apply to your account only.
Note
When you enable fast snapshot restore for a snapshot, your account is billed for each minute
that fast snapshot restore is enabled in a particular Availability Zone. Charges are pro-rated and
have a minimum of one hour.

When you delete a snapshot that you own, fast snapshot restore is automatically disabled for that
snapshot in your account. If you enabled fast snapshot restore for a snapshot that is shared with you,
and the snapshot owner deletes or unshares it, fast snapshot restore is automatically disabled for the
shared snapshot in your account.

If you enabled fast snapshot restore for a snapshot that is shared with you, and it's encrypted using a
custom CMK, fast snapshot restore is not automatically disabled for the snapshot when the snapshot
owner revokes your access to the custom CMK. You must manually disable fast snapshot restore for that
snapshot.

Use the following procedure to enable or disable fast snapshot restore for a snapshot that you own or
for a snapshot that is shared with you.

To enable or disable fast snapshot restore

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Snapshots.
3. Select the snapshot.
4. Choose Actions, Manage Fast Snapshot Restore.
5. Select or deselect Availability Zones, and then choose Save.
6. To track the state of fast snapshot restore as it is enabled, see Fast Snapshot Restore on the
Description tab.

To manage fast snapshot restore using the AWS CLI

• enable-fast-snapshot-restores
• disable-fast-snapshot-restores
• describe-fast-snapshot-restores

View snapshots with fast snapshot restore enabled

Use the following procedure to view the state of fast snapshot restore for a snapshot that you own or for
a snapshot that is shared with you.

To view the state of fast snapshot restore using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Snapshots.
3. Select the snapshot.

1212
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS data services

4. On the Description tab, see Fast Snapshot Restore, which indicates the state of fast snapshot
restore. For example, it might show a state of "2 Availability Zones optimizing" or "2 Availability
Zones enabled".

To view snapshots with fast snapshot restore enabled using the AWS CLI

Use the describe-fast-snapshot-restores command to describe the snapshots that are enabled for fast
snapshot restore.

aws ec2 describe-fast-snapshot-restores --filters Name=state,Values=enabled

The following is example output.

{
"FastSnapshotRestores": [
{
"SnapshotId": "snap-0e946653493cb0447",
"AvailabilityZone": "us-east-2a",
"State": "enabled",
"StateTransitionReason": "Client.UserInitiated - Lifecycle state transition",
"OwnerId": "123456789012",
"EnablingTime": "2020-01-25T23:57:49.596Z",
"OptimizingTime": "2020-01-25T23:58:25.573Z",
"EnabledTime": "2020-01-25T23:59:29.852Z"
},
{
"SnapshotId": "snap-0e946653493cb0447",
"AvailabilityZone": "us-east-2b",
"State": "enabled",
"StateTransitionReason": "Client.UserInitiated - Lifecycle state transition",
"OwnerId": "123456789012",
"EnablingTime": "2020-01-25T23:57:49.596Z",
"OptimizingTime": "2020-01-25T23:58:25.573Z",
"EnabledTime": "2020-01-25T23:59:29.852Z"
}
]
}

View volumes restored using fast snapshot restore

When you create a volume from a snapshot that is enabled for fast snapshot restore in the Availability
Zone for the volume, it is restored using fast snapshot restore.

Use the describe-volumes command to view volumes that were created from a snapshot that is enabled
for fast snapshot restore.

aws ec2 describe-volumes --filters Name=fast-restored,Values=true

The following is example output.

{
"Volumes": [
{
"Attachments": [],
"AvailabilityZone": "us-east-2a",
"CreateTime": "2020-01-26T00:34:11.093Z",
"Encrypted": true,
"KmsKeyId": "arn:aws:kms:us-west-2:123456789012:key/8c5b2c63-b9bc-45a3-
a87a-5513e232e843",
"Size": 20,

1213
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes and NVMe

"SnapshotId": "snap-0e946653493cb0447",
"State": "available",
"VolumeId": "vol-0d371921d4ca797b0",
"Iops": 100,
"VolumeType": "gp2",
"FastRestored": true
}
]
}

Monitor fast snapshot restore

Amazon EBS emits Amazon CloudWatch events when the fast snapshot restore state for a snapshot
changes. For more information, see EBS fast snapshot restore events (p. 1265).

Pricing and Billing

You are billed for each minute that fast snapshot restore is enabled for a snapshot in a particular
Availability Zone. Charges are pro-rated with a minimum of one hour.

For example, if you enable fast snapshot restore for one snapshot in US-East-1a for one month (30
days), you are billed $540 (1 snapshot x 1 AZ x 720 hours x $0.75 per hour). If you enable fast snapshot
restore for two snapshots in us-east-1a, us-east-1b, and us-east-1c for the same period, you are
billed $3240 (2 snapshots x 3 AZs x 720 hours x $0.75 per hour).

If you enable fast snapshot restore for a public or private snapshot that is shared with you, your account
is billed; the snapshot owner is not billed. When a snapshot that is shared with you is deleted or
unshared by the snapshot owner, fast snapshot restore is disabled for the snapshot in your account and
billing is stopped.

For more information, see Amazon EBS pricing.

Amazon EBS and NVMe on Linux instances

EBS volumes are exposed as NVMe block devices on instances built on the Nitro System (p. 209). The
device names are /dev/nvme0n1, /dev/nvme1n1, and so on. The device names that you specify in a
block device mapping are renamed using NVMe device names (/dev/nvme[0-26]n1). The block device
driver can assign NVMe device names in a different order than you specified for the volumes in the block
device mapping.

The EBS performance guarantees stated in Amazon EBS Product Details are valid regardless of the block-
device interface.

Contents
• Install or upgrade the NVMe driver (p. 1214)
• Identify the EBS device (p. 1215)
• Work with NVMe EBS volumes (p. 1217)
• I/O operation timeout (p. 1217)

Install or upgrade the NVMe driver

To access NVMe volumes, the NVMe drivers must be installed. Instances can support NVMe EBS volumes,
NVMe instance store volumes, both types of NVMe volumes, or no NVMe volumes. For more information,
see Summary of networking and storage features (p. 210).

The following AMIs include the required NVMe drivers:

• Amazon Linux 2

1214
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes and NVMe

• Amazon Linux AMI 2018.03

• Ubuntu 14.04 (with linux-aws kernel) or later
• Red Hat Enterprise Linux 7.4 or later
• SUSE Linux Enterprise Server 12 SP2 or later
• CentOS 7.4.1708 or later
• FreeBSD 11.1 or later
• Debian GNU/Linux 9 or later

For more information about NVMe drivers on Windows instances, see Amazon EBS and NVMe on
Windows Instances in the Amazon EC2 User Guide for Windows Instances.

To confirm that your instance has the NVMe driver

You can confirm that your instance has the NVMe driver and check the driver version using the following
command. If the instance has the NVMe driver, the command returns information about the driver.

$ modinfo nvme

To update the NVMe driver

If your instance has the NVMe driver, you can update the driver to the latest version using the following
procedure.

1. Connect to your instance.

2. Update your package cache to get necessary package updates as follows.

• For Amazon Linux 2, Amazon Linux, CentOS, and Red Hat Enterprise Linux:

[ec2-user ~]$ sudo yum update -y

• For Ubuntu and Debian:

[ec2-user ~]$ sudo apt-get update -y

3. Ubuntu 16.04 and later include the linux-aws package, which contains the NVMe and ENA drivers
required by Nitro-based instances. Upgrade the linux-aws package to receive the latest version as
follows:

[ec2-user ~]$ sudo apt-get install --only-upgrade -y linux-aws

For Ubuntu 14.04, you can install the latest linux-aws package as follows:

[ec2-user ~]$ sudo apt-get install linux-aws

4. Reboot your instance to load the latest kernel version.

sudo reboot

5. Reconnect to your instance after it has rebooted.

Identify the EBS device

EBS uses single-root I/O virtualization (SR-IOV) to provide volume attachments on Nitro-based
instances using the NVMe specification. These devices rely on standard NVMe drivers on the operating

1215
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS volumes and NVMe

system. These drivers typically discover attached devices by scanning the PCI bus during instance boot,
and create device nodes based on the order in which the devices respond, not on how the devices
are specified in the block device mapping. In Linux, NVMe device names follow the pattern /dev/
nvme<x>n<y>, where <x> is the enumeration order, and, for EBS, <y> is 1. Occasionally, devices can
respond to discovery in a different order in subsequent instance starts, which causes the device name to
change.

We recommend that you use stable identifiers for your EBS volumes within your instance, such as one of
the following:

• For Nitro-based instances, the block device mappings that are specified in the Amazon EC2 console
when you are attaching an EBS volume or during AttachVolume or RunInstances API calls are
captured in the vendor-specific data field of the NVMe controller identification. With Amazon Linux
AMIs later than version 2017.09.01, we provide a udev rule that reads this data and creates a symbolic
link to the block-device mapping.
• NVMe EBS volumes have the EBS volume ID set as the serial number in the device identification. Use
the lsblk -o +SERIAL command to list the serial number.
• When a device is formatted, a UUID is generated that persists for the life of the filesystem. A device
label can be specified at the same time. For more information, see Make an Amazon EBS volume
available for use on Linux (p. 1111) and Boot from the wrong volume (p. 1366).

Amazon Linux AMIs

With Amazon Linux AMI 2017.09.01 or later (including Amazon Linux 2), you can run the ebsnvme-id
command as follows to map the NVMe device name to a volume ID and device name:

[ec2-user ~]$ sudo /sbin/ebsnvme-id /dev/nvme1n1

Volume ID: vol-01324f611e2463981
/dev/sdf

Amazon Linux also creates a symbolic link from the device name in the block device mapping (for
example, /dev/sdf), to the NVMe device name.

FreeBSD AMIs

Starting with FreeBSD 12.2-RELEASE, you can run the ebsnvme-id command as shown above. Pass
either the name of the NVMe device (for example, nvme0) or the disk device (for example, nvd0
or nda0). FreeBSD also creates symbolic links to the disk devices (for example, /dev/aws/disk/
ebs/volume_id).

Other Linux AMIs

With a kernel version of 4.2 or later, you can run the nvme id-ctrl command as follows to map an NVMe
device to a volume ID. First, install the NVMe command line package, nvme-cli, using the package
management tools for your Linux distribution. For download and installation instructions for other
distributions, refer to the documentation specific to your distribution.

The following example gets the volume ID and device name. The device name is available through the
NVMe controller vendor-specific extension (bytes 384:4095 of the controller identification):

[ec2-user ~]$ sudo nvme id-ctrl -v /dev/nvme1n1

NVME Identify Controller:
vid : 0x1d0f
ssvid : 0x1d0f
sn : vol01234567890abcdef
mn : Amazon Elastic Block Store
...
0000: 2f 64 65 76 2f 73 64 6a 20 20 20 20 20 20 20 20 "/dev/sdf..."

1216
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

The lsblk command lists available devices and their mount points (if applicable). This helps you
determine the correct device name to use. In this example, /dev/nvme0n1p1 is mounted as the root
device and /dev/nvme1n1 is attached but not mounted.

[ec2-user ~]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
nvme1n1 259:3 0 100G 0 disk
nvme0n1 259:0 0 8G 0 disk
nvme0n1p1 259:1 0 8G 0 part /
nvme0n1p128 259:2 0 1M 0 part

Work with NVMe EBS volumes

To format and mount an NVMe EBS volume, see Make an Amazon EBS volume available for use on
Linux (p. 1111).

If you are using Linux kernel 4.2 or later, any change you make to the volume size of an NVMe EBS
volume is automatically reflected in the instance. For older Linux kernels, you might need to detach and
attach the EBS volume or reboot the instance for the size change to be reflected. With Linux kernel 3.19
or later, you can use the hdparm command as follows to force a rescan of the NVMe device:

[ec2-user ~]$ sudo hdparm -z /dev/nvme1n1

When you detach an NVMe EBS volume, the instance does not have an opportunity to flush the file
system caches or metadata before detaching the volume. Therefore, before you detach an NVMe EBS
volume, you should first sync and unmount it. If the volume fails to detach, you can attempt a force-
detach command as described in Detach an Amazon EBS volume from a Linux instance (p. 1123).

I/O operation timeout

EBS volumes attached to Nitro-based instances use the default NVMe driver provided by the operating
system. Most operating systems specify a timeout for I/O operations submitted to NVMe devices. The
default timeout is 30 seconds and can be changed using the nvme_core.io_timeout boot parameter.
For most Linux kernels earlier than version 4.6, this parameter is nvme.io_timeout.

If I/O latency exceeds the value of this timeout parameter, the Linux NVMe driver fails the I/O and
returns an error to the filesystem or application. Depending on the I/O operation, your filesystem or
application can retry the error. In some cases, your filesystem might be remounted as read-only.

For an experience similar to EBS volumes attached to Xen instances, we recommend setting
nvme_core.io_timeout to the highest value possible. For current kernels, the maximum is
4294967295, while for earlier kernels the maximum is 255. Depending on the version of Linux, the
timeout might already be set to the supported maximum value. For example, the timeout is set to
4294967295 by default for Amazon Linux AMI 2017.09.01 and later.

You can verify the maximum value for your Linux distribution by writing a value higher than the
suggested maximum to /sys/module/nvme_core/parameters/io_timeout and checking for the
Numerical result out of range error when attempting to save the file.

Amazon EBS–optimized instances

An Amazon EBS–optimized instance uses an optimized configuration stack and provides additional,
dedicated capacity for Amazon EBS I/O. This optimization provides the best performance for your EBS
volumes by minimizing contention between Amazon EBS I/O and other traffic from your instance.

EBS–optimized instances deliver dedicated bandwidth to Amazon EBS. When attached to an EBS–
optimized instance, General Purpose SSD (gp2 and gp3) volumes are designed to deliver their baseline
and burst performance 99% of the time, and Provisioned IOPS SSD (io1 and io2) volumes are designed

1217
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

to deliver their provisioned performance 99.9% of the time. Both Throughput Optimized HDD (st1) and
Cold HDD (sc1) guarantee performance consistency of 90% of burst throughput 99% of the time. Non-
compliant periods are approximately uniformly distributed, targeting 99% of expected total throughput
each hour. For more information, see Amazon EBS volume types (p. 1084).

Contents
• Supported instance types (p. 1218)
• Get maximum performance (p. 1235)
• Enable EBS optimization at launch (p. 1235)
• Enable EBS optimization for an existing instance (p. 1236)

Supported instance types

The following tables show which instance types support EBS optimization. They include the dedicated
bandwidth to Amazon EBS, the typical maximum aggregate throughput that can be achieved on that
connection with a streaming read workload and 128 KiB I/O size, and the maximum IOPS the instance
can support if you are using a 16 KiB I/O size. Choose an EBS–optimized instance that provides more
dedicated Amazon EBS throughput than your application needs; otherwise, the connection between
Amazon EBS and Amazon EC2 can become a performance bottleneck.

EBS optimized by default

The following table lists the instance types that support EBS optimization and EBS optimization is
enabled by default. There is no need to enable EBS optimization and no effect if you disable EBS
optimization.

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

a1.medium * 3,500 437.5 20,000

a1.large * 3,500 437.5 20,000

a1.xlarge * 3,500 437.5 20,000

a1.2xlarge * 3,500 437.5 20,000

a1.4xlarge 3,500 437.5 20,000

a1.metal 3,500 437.5 20,000

c4.large 500 62.5 4,000

c4.xlarge 750 93.75 6,000

c4.2xlarge 1,000 125 8,000

c4.4xlarge 2,000 250 16,000

c4.8xlarge 4,000 500 32,000

c5.large * 4,750 593.75 20,000

c5.xlarge * 4,750 593.75 20,000

c5.2xlarge * 4,750 593.75 20,000

c5.4xlarge 4,750 593.75 20,000

1218
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

c5.9xlarge 9,500 1,187.5 40,000

c5.12xlarge 9,500 1,187.5 40,000

c5.18xlarge 19,000 2,375 80,000

c5.24xlarge 19,000 2,375 80,000

c5.metal 19,000 2,375 80,000

c5a.large * 3,170 396 13,300

c5a.xlarge * 3,170 396 13,300

c5a.2xlarge * 3,170 396 13,300

c5a.4xlarge * 3,170 396 13,300

c5a.8xlarge 3,170 396 13,300

c5a.12xlarge 4,750 594 20,000

c5a.16xlarge 6,300 788 26,700

c5a.24xlarge 9,500 1,188 40,000

c5ad.large * 3,170 396 13,300

c5ad.xlarge * 3,170 396 13,300

c5ad.2xlarge * 3,170 396 13,300

c5ad.4xlarge * 3,170 396 13,300

c5ad.8xlarge 3,170 396 13,300

c5ad.12xlarge 4,750 594 20,000

c5ad.16xlarge 6,300 788 26,700

c5ad.24xlarge 9,500 1,188 40,000

c5d.large * 4,750 593.75 20,000

c5d.xlarge * 4,750 593.75 20,000

c5d.2xlarge * 4,750 593.75 20,000

c5d.4xlarge 4,750 593.75 20,000

c5d.9xlarge 9,500 1,187.5 40,000

c5d.12xlarge 9,500 1,187.5 40,000

c5d.18xlarge 19,000 2,375 80,000

c5d.24xlarge 19,000 2,375 80,000

c5d.metal 19,000 2,375 80,000

1219
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

c5n.large * 4,750 593.75 20,000

c5n.xlarge * 4,750 593.75 20,000

c5n.2xlarge * 4,750 593.75 20,000

c5n.4xlarge 4,750 593.75 20,000

c5n.9xlarge 9,500 1,187.5 40,000

c5n.18xlarge 19,000 2,375 80,000

c5n.metal 19,000 2,375 80,000

c6g.medium * 4,750 593.75 20,000

c6g.large * 4,750 593.75 20,000

c6g.xlarge * 4,750 593.75 20,000

c6g.2xlarge * 4,750 593.75 20,000

c6g.4xlarge 4,750 593.75 20,000

c6g.8xlarge 9,500 1,187.5 40,000

c6g.12xlarge 14,250 1,781.25 50,000

c6g.16xlarge 19,000 2,375 80,000

c6g.metal 19,000 2,375 80,000

c6gd.medium * 4,750 593.75 20,000

c6gd.large * 4,750 593.75 20,000

c6gd.xlarge * 4,750 593.75 20,000

c6gd.2xlarge * 4,750 593.75 20,000

c6gd.4xlarge 4,750 593.75 20,000

c6gd.8xlarge 9,500 1,187.5 40,000

c6gd.12xlarge 14,250 1,781.25 50,000

c6gd.16xlarge 19,000 2,375 80,000

c6gd.metal 19,000 2,375 80,000

c6gn.medium * 9,500 1,187.5 40,000

c6gn.large * 9,500 1,187.5 40,000

c6gn.xlarge * 9,500 1,187.5 40,000

c6gn.2xlarge * 9,500 1,187.5 40,000

c6gn.4xlarge 9,500 1,187.5 40,000

1220
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

c6gn.8xlarge 19,000 2,375 80,000

c6gn.12xlarge 28,500 3,562.5 120,000

c6gn.16xlarge 38,000 4,750 160,000

d2.xlarge 750 93.75 6,000

d2.2xlarge 1,000 125 8,000

d2.4xlarge 2,000 250 16,000

d2.8xlarge 4,000 500 32,000

d3.xlarge * 2,800 350 15,000

d3.2xlarge * 2,800 350 15,000

d3.4xlarge 2,800 350 15,000

d3.8xlarge 5,000 625 30,000

d3en.xlarge * 2,800 350 15,000

d3en.2xlarge * 2,800 350 15,000

d3en.4xlarge 2,800 350 15,000

d3en.8xlarge 5,000 625 30,000

d3en.12xlarge 7,000 875 40,000

f1.2xlarge 1,700 212.5 12,000

f1.4xlarge 3,500 437.5 44,000

f1.16xlarge 14,000 1,750 75,000

g3s.xlarge 850 106.25 5,000

g3.4xlarge 3,500 437.5 20,000

g3.8xlarge 7,000 875 40,000

g3.16xlarge 14,000 1,750 80,000

g4ad.4xlarge * 3,170 396.25 13,333

g4ad.8xlarge 3,170 396.25 13,333

g4ad.16xlarge 6,300 787.5 26,667

g4dn.xlarge * 3,500 437.5 20,000

g4dn.2xlarge * 3,500 437.5 20,000

g4dn.4xlarge 4,750 593.75 20,000

g4dn.8xlarge 9,500 1,187.5 40,000

1221
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

g4dn.12xlarge 9,500 1,187.5 40,000

g4dn.16xlarge 9,500 1,187.5 40,000

g4dn.metal 19,000 2,375 80,000

h1.2xlarge 1,750 218.75 12,000

h1.4xlarge 3,500 437.5 20,000

h1.8xlarge 7,000 875 40,000

h1.16xlarge 14,000 1,750 80,000

i3.large 425 53.13 3000

i3.xlarge 850 106.25 6000

i3.2xlarge 1,700 212.5 12,000

i3.4xlarge 3,500 437.5 16,000

i3.8xlarge 7,000 875 32,500

i3.16xlarge 14,000 1,750 65,000

i3.metal 19,000 2,375 80,000

i3en.large * 4,750 593.75 20,000

i3en.xlarge * 4,750 593.75 20,000

i3en.2xlarge * 4,750 593.75 20,000

i3en.3xlarge * 4,750 593.75 20,000

i3en.6xlarge 4,750 593.75 20,000

i3en.12xlarge 9,500 1,187.5 40,000

i3en.24xlarge 19,000 2,375 80,000

i3en.metal 19,000 2,375 80,000

inf1.xlarge * 4,750 593.75 20,000

inf1.2xlarge * 4,750 593.75 20,000

inf1.6xlarge 4,750 593.75 20,000

inf1.24xlarge 19,000 2,375 80,000

m4.large 450 56.25 3,600

m4.xlarge 750 93.75 6,000

m4.2xlarge 1,000 125 8,000

m4.4xlarge 2,000 250 16,000

1222
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

m4.10xlarge 4,000 500 32,000

m4.16xlarge 10,000 1,250 65,000

m5.large * 4,750 593.75 18,750

m5.xlarge * 4,750 593.75 18,750

m5.2xlarge * 4,750 593.75 18,750

m5.4xlarge 4,750 593.75 18,750

m5.8xlarge 6,800 850 30,000

m5.12xlarge 9,500 1,187.5 40,000

m5.16xlarge 13,600 1,700 60,000

m5.24xlarge 19,000 2,375 80,000

m5.metal 19,000 2,375 80,000

m5a.large * 2,880 360 16,000

m5a.xlarge * 2,880 360 16,000

m5a.2xlarge * 2,880 360 16,000

m5a.4xlarge 2,880 360 16,000

m5a.8xlarge 4,750 593.75 20,000

m5a.12xlarge 6,780 847.5 30,000

m5a.16xlarge 9,500 1,187.50 40,000

m5a.24xlarge 13,570 1,696.25 60,000

m5ad.large * 2,880 360 16,000

m5ad.xlarge * 2,880 360 16,000

m5ad.2xlarge * 2,880 360 16,000

m5ad.4xlarge 2,880 360 16,000

m5ad.8xlarge 4,750 593.75 20,000

m5ad.12xlarge 6,780 847.5 30,000

m5ad.16xlarge 9,500 1,187.5 40,000

m5ad.24xlarge 13,570 1,696.25 60,000

m5d.large * 4,750 593.75 18,750

m5d.xlarge * 4,750 593.75 18,750

m5d.2xlarge * 4,750 593.75 18,750

1223
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

m5d.4xlarge 4,750 593.75 18,750

m5d.8xlarge 6,800 850 30,000

m5d.12xlarge 9,500 1,187.5 40,000

m5d.16xlarge 13,600 1,700 60,000

m5d.24xlarge 19,000 2,375 80,000

m5d.metal 19,000 2,375 80,000

m5dn.large * 4,750 593.75 18,750

m5dn.xlarge * 4,750 593.75 18,750

m5dn.2xlarge * 4,750 593.75 18,750

m5dn.4xlarge 4,750 593.75 18,750

m5dn.8xlarge 6,800 850 30,000

m5dn.12xlarge 9,500 1,187.5 40,000

m5dn.16xlarge 13,600 1,700 60,000

m5dn.24xlarge 19,000 2,375 80,000

m5n.large * 4,750 593.75 18,750

m5n.xlarge * 4,750 593.75 18,750

m5n.2xlarge * 4,750 593.75 18,750

m5n.4xlarge 4,750 593.75 18,750

m5n.8xlarge 6,800 850 30,000

m5n.12xlarge 9,500 1,187.5 40,000

m5n.16xlarge 13,600 1,700 60,000

m5n.24xlarge 19,000 2,375 80,000

m5zn.large * 3,170 396.25 13,333

m5zn.xlarge * 3,170 396.25 13,333

m5zn.2xlarge 3,170 396.25 13,333

m5zn.3xlarge 4,750 593.75 20,000

m5zn.6xlarge 9,500 1187.5 40,000

m5zn.12xlarge 19,000 2,375 80,000

m5zn.metal 19,000 2,375 80,000

m6g.medium * 4,750 593.75 20,000

1224
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

m6g.large * 4,750 593.75 20,000

m6g.xlarge * 4,750 593.75 20,000

m6g.2xlarge * 4,750 593.75 20,000

m6g.4xlarge 4,750 593.75 20,000

m6g.8xlarge 9,500 1,187.5 40,000

m6g.12xlarge 14,250 1,781.25 50,000

m6g.16xlarge 19,000 2,375 80,000

m6g.metal 19,000 2,375 80,000

m6gd.medium * 4,750 593.75 20,000

m6gd.large * 4,750 593.75 20,000

m6gd.xlarge * 4,750 593.75 20,000

m6gd.2xlarge * 4,750 593.75 20,000

m6gd.4xlarge 4,750 593.75 20,000

m6gd.8xlarge 9,500 1,187.5 40,000

m6gd.12xlarge 14,250 1,781.25 50,000

m6gd.16xlarge 19,000 2,375 80,000

m6gd.metal 19,000 2,375 80,000

mac1.metal 8,000 1,000 55,000

p2.xlarge 750 93.75 6,000

p2.8xlarge 5,000 625 32,500

p2.16xlarge 10,000 1,250 65,000

p3.2xlarge 1,750 218.75 10,000

p3.8xlarge 7,000 875 40,000

p3.16xlarge 14,000 1,750 80,000

p3dn.24xlarge 19,000 2,375 80,000

p4d.2xlarge 19,000 2,375 80,000

r4.large 425 53.13 3,000

r4.xlarge 850 106.25 6,000

r4.2xlarge 1,700 212.5 12,000

r4.4xlarge 3,500 437.5 18,750

1225
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

r4.8xlarge 7,000 875 37,500

r4.16xlarge 14,000 1,750 75,000

r5.large * 4,750 593.75 18,750

r5.xlarge * 4,750 593.75 18,750

r5.2xlarge * 4,750 593.75 18,750

r5.4xlarge 4,750 593.75 18,750

r5.8xlarge 6,800 850 30,000

r5.12xlarge 9,500 1,187.5 40,000

r5.16xlarge 13,600 1,700 60,000

r5.24xlarge 19,000 2,375 80,000

r5.metal 19,000 2,375 80,000

r5a.large * 2,880 360 16,000

r5a.xlarge * 2,880 360 16,000

r5a.2xlarge * 2,880 360 16,000

r5a.4xlarge 2,880 360 16,000

r5a.8xlarge 4,750 593.75 20,000

r5a.12xlarge 6,780 847.5 30,000

r5a.16xlarge 9,500 1,187.5 40,000

r5a.24xlarge 13,570 1,696.25 60,000

r5ad.large * 2,880 360 16,000

r5ad.xlarge * 2,880 360 16,000

r5ad.2xlarge * 2,880 360 16,000

r5ad.4xlarge 2,880 360 16,000

r5ad.8xlarge 4,750 593.75 20,000

r5ad.12xlarge 6,780 847.5 30,000

r5ad.16xlarge 9,500 1,187.5 40,000

r5ad.24xlarge 13,570 1,696.25 60,000

r5b.large * 10,000 1,250 43,333

r5b.xlarge * 10,000 1,250 43,333

r5b.2xlarge * 10,000 1,250 43,333

1226
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

r5b.4xlarge 10,000 1,250 43,333

r5b.8xlarge 20,000 2,500 86,667

r5b.12xlarge 30,000 3,750 130,000

r5b.16xlarge 40,000 5,000 173,333

r5b.24xlarge 60,000 7,500 260,000

r5b.metal 60,000 7,500 260,000

r5d.large * 4,750 593.75 18,750

r5d.xlarge * 4,750 593.75 18,750

r5d.2xlarge * 4,750 593.75 18,750

r5d.4xlarge 4,750 593.75 18,750

r5d.8xlarge 6,800 850 30,000

r5d.12xlarge 9,500 1,187.5 40,000

r5d.16xlarge 13,600 1,700 60,000

r5d.24xlarge 19,000 2,375 80,000

r5d.metal 19,000 2,375 80,000

r5dn.large * 4,750 593.75 18,750

r5dn.xlarge * 4,750 593.75 18,750

r5dn.2xlarge * 4,750 593.75 18,750

r5dn.4xlarge 4,750 593.75 18,750

r5dn.8xlarge 6,800 850 30,000

r5dn.12xlarge 9,500 1,187.5 40,000

r5dn.16xlarge 13,600 1,700 60,000

r5dn.24xlarge 19,000 2,375 80,000

r5n.large * 4,750 593.75 18,750

r5n.xlarge * 4,750 593.75 18,750

r5n.2xlarge * 4,750 593.75 18,750

r5n.4xlarge 4,750 593.75 18,750

r5n.8xlarge 6,800 850 30,000

r5n.12xlarge 9,500 1,187.5 40,000

r5n.16xlarge 13,600 1,700 60,000

1227
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

r5n.24xlarge 19,000 2,375 80,000

r6g.medium * 4,750 593.75 20,000

r6g.large * 4,750 593.75 20,000

r6g.xlarge * 4,750 593.75 20,000

r6g.2xlarge * 4,750 593.75 20,000

r6g.4xlarge 4,750 593.75 20,000

r6g.8xlarge 9,500 1,187.5 40,000

r6g.12xlarge 14,250 1,781.25 50,000

r6g.16xlarge 19,000 2,375 80,000

r6g.metal 19,000 2,375 80,000

r6gd.medium * 4,750 593.75 20,000

r6gd.large * 4,750 593.75 20,000

r6gd.xlarge * 4,750 593.75 20,000

r6gd.2xlarge * 4,750 593.75 20,000

r6gd.4xlarge 4,750 593.75 20,000

r6gd.8xlarge 9,500 1,187.5 40,000

r6gd.12xlarge 14,250 1,781.25 50,000

r6gd.16xlarge 19,000 2,375 80,000

r6gd.metal 19,000 2,375 80,000

t3.nano * 2,085 260.57 11,800

t3.micro * 2,085 260.57 11,800

t3.small * 2,085 260.57 11,800

t3.medium * 2,085 260.57 11,800

t3.large * 2,780 347.5 15,700

t3.xlarge * 2,780 347.5 15,700

t3.2xlarge * 2,780 347.5 15,700

t3a.nano * 2,085 260.57 11,800

t3a.micro * 2,085 260.57 11,800

t3a.small * 2,085 260.57 11,800

t3a.medium * 2,085 260.57 11,800

1228
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

t3a.large * 2,780 347.5 15,700

t3a.xlarge * 2,780 347.5 15,700

t3a.2xlarge * 2,780 347.5 15,700

t4g.nano * 2,606 325.75 11,800

t4g.micro * 2,606 325.75 11,800

t4g.small * 2,606 325.75 11,800

t4g.medium * 2,606 325.75 11,800

t4g.large * 3,475 434.37 15,700

t4g.xlarge * 3,475 434.37 15,700

t4g.2xlarge * 3,475 434.37 15,700

u-6tb1.metal 38,000 4,750 160,000

u-9tb1.metal 38,000 4,750 160,000

u-12tb1.metal 38,000 4,750 160,000

u-18tb1.metal 38,000 4,750 160,000

u-24tb1.metal 38,000 4,750 160,000

x1.16xlarge 7,000 875 40,000

x1.32xlarge 14,000 1,750 80,000

x1e.xlarge 500 62.5 3,700

x1e.2xlarge 1,000 125 7,400

x1e.4xlarge 1,750 218.75 10,000

x1e.8xlarge 3,500 437.5 20,000

x1e.16xlarge 7,000 875 40,000

x1e.32xlarge 14,000 1,750 80,000

z1d.large * 3,170 396.25 13,333

z1d.xlarge * 3,170 396.25 13,333

z1d.2xlarge 3,170 396.25 13,333

z1d.3xlarge 4,750 593.75 20,000

z1d.6xlarge 9,500 1,187.5 40,000

z1d.12xlarge 19,000 2,375 80,000

z1d.metal 19,000 2,375 80,000

1229
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

* These instance types can support maximum performance for 30 minutes at least once every 24 hours.
If you have a workload that requires sustained maximum performance for longer than 30 minutes, select
an instance type according to baseline performance as shown in the following table.

Instance size Baseline bandwidth Baseline throughput Baseline IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

a1.medium 300 37.5 2,500

a1.large 525 65.625 4,000

a1.xlarge 800 100 6,000

a1.2xlarge 1,750 218.75 10,000

c5.large 650 81.25 4,000

c5.xlarge 1,150 143.75 6,000

c5.2xlarge 2,300 287.5 10,000

c5a.large 200 25 800

c5a.xlarge 400 50 1,600

c5a.2xlarge 800 100 3,200

c5a.4xlarge 1,580 198 6,600

c5ad.large 200 25 800

c5ad.xlarge 400 50 1,600

c5ad.2xlarge 800 100 3,200

c5ad.4xlarge 1,580 198 6,600

c5d.large 650 81.25 4,000

c5d.xlarge 1,150 143.75 6,000

c5d.2xlarge 2,300 287.5 10,000

c5n.large 650 81.25 4,000

c5n.xlarge 1,150 143.75 6,000

c5n.2xlarge 2,300 287.5 10,000

c6g.medium 315 39.375 2,500

c6g.large 630 78.75 3,600

c6g.xlarge 1,188 148.5 6,000

c6g.2xlarge 2,375 296.875 12,000

c6gd.medium 315 39.375 2,500

c6gd.large 630 78.75 3,600

c6gd.xlarge 1,188 148.5 6,000

1230
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Baseline bandwidth Baseline throughput Baseline IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

c6gd.2xlarge 2,375 296.875 12,000

c6gn.medium 760 95 2,500

c6gn.large 1,235 154.375 5,000

c6gn.xlarge 1,900 237.5 10,000

c6gn.2xlarge 4,750 593.75 20,000

d3.xlarge 850 106.25 5,000

d3.2xlarge 1,700 212.5 10,000

d3en.large 425 53.125 2,500

d3en.xlarge 850 106.25 5,000

d3en.2xlarge 1,700 212.5 10,000

g4ad.4xlarge 1,580 197.5 6,700

g4dn.xlarge 950 118.75 3,000

g4dn.2xlarge 1,150 143.75 6,000

i3en.large 577 72.1 3,000

i3en.xlarge 1,154 144.2 6,000

i3en.2xlarge 2,307 288.39 12,000

i3en.3xlarge 3,800 475 15,000

inf1.xlarge 1,190 148.75 4,000

inf1.2xlarge 1,190 148.75 6,000

m5.large 650 81.25 3,600

m5.xlarge 1,150 143.75 6,000

m5.2xlarge 2,300 287.5 12,000

m5a.large 650 81.25 3,600

m5a.xlarge 1,085 135.63 6,000

m5a.2xlarge 1,580 197.5 8,333

m5ad.large 650 81.25 3,600

m5ad.xlarge 1,085 135.63 6,000

m5ad.2xlarge 1,580 197.5 8,333

m5d.large 650 81.25 3,600

m5d.xlarge 1,150 143.75 6,000

1231
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Baseline bandwidth Baseline throughput Baseline IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

m5d.2xlarge 2,300 287.5 12,000

m5dn.large 650 81.25 3,600

m5dn.xlarge 1,150 143.75 6,000

m5dn.2xlarge 2,300 287.5 12,000

m5n.large 650 81.25 3,600

m5n.xlarge 1,150 143.75 6,000

m5n.2xlarge 2,300 287.5 12,000

m5zn.large 800 100 3,333

m5zn.xlarge 1,580 195.5 6,667

m6g.medium 315 39.375 2,500

m6g.large 630 78.75 3,600

m6g.xlarge 1,188 148.5 6,000

m6g.2xlarge 2,375 296.875 12,000

m6gd.medium 315 39.375 2,500

m6gd.large 630 78.75 3,600

m6gd.xlarge 1,188 148.5 6,000

m6gd.2xlarge 2,375 296.875 12,000

r5.large 650 81.25 3,600

r5.xlarge 1,150 143.75 6,000

r5.2xlarge 2,300 287.5 12,000

r5a.large 650 81.25 3,600

r5a.xlarge 1,085 135.63 6,000

r5a.2xlarge 1,580 197.5 8,333

r5ad.large 650 81.25 3,600

r5ad.xlarge 1,085 135.63 6,000

r5ad.2xlarge 1,580 197.5 8,333

r5b.large 1,250 156.25 5,417

r5b.xlarge 2,500 312.5 10,833

r5b.2xlarge 5,000 625 21,667

r5d.large 650 81.25 3,600

1232
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Baseline bandwidth Baseline throughput Baseline IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

r5d.xlarge 1,150 143.75 6,000

r5d.2xlarge 2,300 287.5 12,000

r5dn.large 650 81.25 3,600

r5dn.xlarge 1,150 143.75 6,000

r5dn.2xlarge 2,300 287.5 12,000

r5n.large 650 81.25 3,600

r5n.xlarge 1,150 143.75 6,000

r5n.2xlarge 2,300 287.5 12,000

r6g.medium 315 39.375 2,500

r6g.large 630 78.75 3,600

r6g.xlarge 1,188 148.5 6,000

r6g.2xlarge 2,375 296.875 12,000

r6gd.medium * 315 39.375 2,500

r6gd.large * 630 78.75 3,600

r6gd.xlarge * 1,188 148.5 6,000

r6gd.2xlarge * 2,375 296.875 12,000

t3.nano 43 5.43 250

t3.micro 87 10.86 500

t3.small 174 21.71 1,000

t3.medium 347 43.43 2,000

t3.large 695 86.86 4,000

t3.xlarge 695 86.86 4,000

t3.2xlarge 695 86.86 4,000

t3a.nano 45 5.63 250

t3a.micro 90 11.25 500

t3a.small 175 21.88 1,000

t3a.medium 350 43.75 2,000

t3a.large 695 86.86 4,000

t3a.xlarge 695 86.86 4,000

t3a.2xlarge 695 86.86 4,000

1233
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Baseline bandwidth Baseline throughput Baseline IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

t4g.nano 32 4 250

t4g.micro 64 8 500

t4g.small 128 16 1,000

t4g.medium 256 32 2,000

t4g.large 512 64 4,000

t4g.xlarge 1,024 128 4,000

t4g.2xlarge 2,048 256 4,000

z1d.large 800 100 3,333

z1d.xlarge 1,580 197.5 6,667

EBS optimization supported

The following table lists the instance types that support EBS optimization but EBS optimization is
not enabled by default. You can enable EBS optimization when you launch these instances or after
they are running. Instances must have EBS optimization enabled to achieve the level of performance
described. When you enable EBS optimization for an instance that is not EBS-optimized by default, you
pay an additional low, hourly fee for the dedicated capacity. For pricing information, see EBS-Optimized
Instances on the Amazon EC2 Pricing, On-Demand Pricing page.

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

c1.xlarge 1,000 125 8,000

c3.xlarge 500 62.5 4,000

c3.2xlarge 1,000 125 8,000

c3.4xlarge 2,000 250 16,000

g2.2xlarge 1,000 125 8,000

i2.xlarge 500 62.5 4,000

i2.2xlarge 1,000 125 8,000

i2.4xlarge 2,000 250 16,000

m1.large 500 62.5 4,000

m1.xlarge 1,000 125 8,000

m2.2xlarge 500 62.5 4,000

m2.4xlarge 1,000 125 8,000

m3.xlarge 500 62.5 4,000

m3.2xlarge 1,000 125 8,000

1234
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

Instance size Maximum bandwidth Maximum throughput Maximum IOPS (16 KiB
(Mbps) (MB/s, 128 KiB I/O) I/O)

r3.xlarge 500 62.5 4,000

r3.2xlarge 1,000 125 8,000

r3.4xlarge 2,000 250 16,000

The i2.8xlarge, c3.8xlarge, and r3.8xlarge instances do not have dedicated EBS bandwidth and
therefore do not offer EBS optimization. On these instances, network traffic and Amazon EBS traffic
share the same 10-gigabit network interface.

Get maximum performance

You can use the EBSIOBalance% and EBSByteBalance% metrics to help you determine whether your
instances are sized correctly. You can view these metrics in the CloudWatch console and set an alarm that
is triggered based on a threshold you specify. These metrics are expressed as a percentage. Instances
with a consistently low balance percentage are candidates to size up. Instances where the balance
percentage never drops below 100% are candidates for downsizing. For more information, see Monitor
your instances using CloudWatch (p. 761).

The high memory instances are designed to run large in-memory databases, including production
deployments of the SAP HANA in-memory database, in the cloud. To maximize EBS performance,
use high memory instances with an even number of io1 or io2 volumes with identical provisioned
performance. For example, for IOPS heavy workloads, use four io1 or io2 volumes with 40,000
provisioned IOPS to get the maximum 160,000 instance IOPS. Similarly, for throughput heavy workloads,
use six io1 or io2 volumes with 48,000 provisioned IOPS to get the maximum 4,750 MB/s throughput.
For additional recommendations, see Storage Configuration for SAP HANA.

Considerations

• G4dn, I3en, Inf1, M5a, M5ad, R5a, R5ad, T3, T3a, and Z1d instances launched after February 26, 2020
provide the maximum performance listed in the table above. To get the maximum performance from
an instance launched before February 26, 2020, stop and start it.
• C5, C5d, C5n, M5, M5d, M5n, M5dn, R5, R5d, R5n, R5dn, and P3dn instances launched after
December 3, 2019 provide the maximum performance listed in the table above. To get the maximum
performance from an instance launched before December 3, 2019, stop and start it.
• u-6tb1.metal, u-9tb1.metal, and u-12tb1.metal instances launched after March 12, 2020
provide the performance in the table above. Instances of these types launched before March 12, 2020
might provide lower performance. To get the maximum performance from an instance launched
before March 12, 2020, contact your account team to upgrade the instance at no additional cost.

Enable EBS optimization at launch

You can enable optimization for an instance by setting its attribute for EBS optimization.

To enable Amazon EBS optimization when launching an instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Launch Instance.
3. In Step 1: Choose an Amazon Machine Image (AMI), select an AMI.
4. In Step 2: Choose an Instance Type, select an instance type that is listed as supporting Amazon EBS
optimization.

1235
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS optimization

5. In Step 3: Configure Instance Details, complete the fields that you need and choose Launch as
EBS-optimized instance. If the instance type that you selected in the previous step doesn't support
Amazon EBS optimization, this option is not present. If the instance type that you selected is
Amazon EBS–optimized by default, this option is selected and you can't deselect it.
6. Follow the directions to complete the wizard and launch your instance.

To enable EBS optimization when launching an instance using the command line

You can use one of the following commands with the corresponding option. For more information about
these command line interfaces, see Access Amazon EC2 (p. 3).

• run-instances with --ebs-optimized (AWS CLI)

• New-EC2Instance with -EbsOptimized (AWS Tools for Windows PowerShell)

Enable EBS optimization for an existing instance

You can enable or disable optimization for an existing instance by modifying its Amazon EBS–optimized
instance attribute. If the instance is running, you must stop it first.
Warning
When you stop an instance, the data on any instance store volumes is erased. To keep data from
instance store volumes, be sure to back it up to persistent storage.

To enable EBS optimization for an existing instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and select the instance.
3. To stop the instance, choose Actions, Instance state, Stop instance. It can take a few minutes for
the instance to stop.
4. With the instance still selected, choose Actions, Instance settings, Change instance type.
5. For Change Instance Type, do one of the following:

• If the instance type of your instance is Amazon EBS–optimized by default, EBS-optimized is

selected and you can't change it. You can choose Cancel, because Amazon EBS optimization is
already enabled for the instance.
• If the instance type of your instance supports Amazon EBS optimization, choose EBS-optimized
and then choose Apply.
• If the instance type of your instance does not support Amazon EBS optimization, you can't choose
EBS-optimized. You can select an instance type from Instance type that supports Amazon EBS
optimization, choose EBS-optimized, and then choose Apply.
6. Choose Instance state, Start instance.

To enable EBS optimization for an existing instance using the command line

1. If the instance is running, use one of the following commands to stop it:

• stop-instances (AWS CLI)

• Stop-EC2Instance (AWS Tools for Windows PowerShell)
2. To enable EBS optimization, use one of the following commands with the corresponding option:

• modify-instance-attribute with --ebs-optimized (AWS CLI)

• Edit-EC2InstanceAttribute with -EbsOptimized (AWS Tools for Windows PowerShell)

1236
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

Amazon EBS volume performance on Linux instances

Several factors, including I/O characteristics and the configuration of your instances and volumes,
can affect the performance of Amazon EBS. Customers who follow the guidance on our Amazon EBS
and Amazon EC2 product detail pages typically achieve good performance out of the box. However,
there are some cases where you may need to do some tuning in order to achieve peak performance on
the platform. This topic discusses general best practices as well as performance tuning that is specific
to certain use cases. We recommend that you tune performance with information from your actual
workload, in addition to benchmarking, to determine your optimal configuration. After you learn the
basics of working with EBS volumes, it's a good idea to look at the I/O performance you require and at
your options for increasing Amazon EBS performance to meet those requirements.

AWS updates to the performance of EBS volume types might not immediately take effect on your
existing volumes. To see full performance on an older volume, you might first need to perform a
ModifyVolume action on it. For more information, see Modifying the Size, IOPS, or Type of an EBS
Volume on Linux.

Contents
• Amazon EBS performance tips (p. 1237)
• I/O characteristics and monitoring (p. 1239)
• Initialize Amazon EBS volumes (p. 1241)
• RAID configuration on Linux (p. 1243)
• Benchmark EBS volumes (p. 1247)

Amazon EBS performance tips

These tips represent best practices for getting optimal performance from your EBS volumes in a variety
of user scenarios.

Use EBS-optimized instances

On instances without support for EBS-optimized throughput, network traffic can contend with traffic
between your instance and your EBS volumes; on EBS-optimized instances, the two types of traffic are
kept separate. Some EBS-optimized instance configurations incur an extra cost (such as C3, R3, and
M3), while others are always EBS-optimized at no extra cost (such as M4, C4, C5, and D2). For more
information, see Amazon EBS–optimized instances (p. 1217).

Understand how performance is calculated

When you measure the performance of your EBS volumes, it is important to understand the units of
measure involved and how performance is calculated. For more information, see I/O characteristics and
monitoring (p. 1239).

Understand your workload

There is a relationship between the maximum performance of your EBS volumes, the size and number of
I/O operations, and the time it takes for each action to complete. Each of these factors (performance, I/
O, and latency) affects the others, and different applications are more sensitive to one factor or another.
For more information, see Benchmark EBS volumes (p. 1247).

Be aware of the performance penalty When initializing volumes from snapshots

There is a significant increase in latency when you first access each block of data on a new EBS volume
that was created from a snapshot. You can avoid this performance hit using one of the following options:

1237
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

• Access each block prior to putting the volume into production. This process is called initialization
(formerly known as pre-warming). For more information, see Initialize Amazon EBS volumes (p. 1241).
• Enable fast snapshot restore on a snapshot to ensure that the EBS volumes created from it are fully-
initialized at creation and instantly deliver all of their provisioned performance. For more information,
see Amazon EBS fast snapshot restore (p. 1210).

Factors that can degrade HDD performance

When you create a snapshot of a Throughput Optimized HDD (st1) or Cold HDD (sc1) volume,
performance may drop as far as the volume's baseline value while the snapshot is in progress. This
behavior is specific to these volume types. Other factors that can limit performance include driving
more throughput than the instance can support, the performance penalty encountered while initializing
volumes created from a snapshot, and excessive amounts of small, random I/O on the volume. For more
information about calculating throughput for HDD volumes, see Amazon EBS volume types (p. 1084).

Your performance can also be impacted if your application isn’t sending enough I/O requests. This can
be monitored by looking at your volume’s queue length and I/O size. The queue length is the number
of pending I/O requests from your application to your volume. For maximum consistency, HDD-backed
volumes must maintain a queue length (rounded to the nearest whole number) of 4 or more when
performing 1 MiB sequential I/O. For more information about ensuring consistent performance of your
volumes, see I/O characteristics and monitoring (p. 1239)

Increase read-ahead for high-throughput, read-heavy workloads on st1 and

sc1
Some workloads are read-heavy and access the block device through the operating system page cache
(for example, from a file system). In this case, to achieve the maximum throughput, we recommend that
you configure the read-ahead setting to 1 MiB. This is a per-block-device setting that should only be
applied to your HDD volumes.

To examine the current value of read-ahead for your block devices, use the following command:

[ec2-user ~]$ sudo blockdev --report /dev/<device>

Block device information is returned in the following format:

RO RA SSZ BSZ StartSec Size Device

rw 256 512 4096 4096 8587820544 /dev/<device>

The device shown reports a read-ahead value of 256 (the default). Multiply this number by the sector
size (512 bytes) to obtain the size of the read-ahead buffer, which in this case is 128 KiB. To set the
buffer value to 1 MiB, use the following command:

[ec2-user ~]$ sudo blockdev --setra 2048 /dev/<device>

Verify that the read-ahead setting now displays 2,048 by running the first command again.

Only use this setting when your workload consists of large, sequential I/Os. If it consists mostly of small,
random I/Os, this setting will actually degrade your performance. In general, if your workload consists
mostly of small or random I/Os, you should consider using a General Purpose SSD (gp2 and gp3) volume
rather than an st1 or sc1 volume.

Use a modern Linux kernel

Use a modern Linux kernel with support for indirect descriptors. Any Linux kernel 3.8 and above has
this support, as well as any current-generation EC2 instance. If your average I/O size is at or near 44
KiB, you may be using an instance or kernel without support for indirect descriptors. For information

1238
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

about deriving the average I/O size from Amazon CloudWatch metrics, see I/O characteristics and
monitoring (p. 1239).

To achieve maximum throughput on st1 or sc1 volumes, we recommend applying a value

of 256 to the xen_blkfront.max parameter (for Linux kernel versions below 4.6) or the
xen_blkfront.max_indirect_segments parameter (for Linux kernel version 4.6 and above). The
appropriate parameter can be set in your OS boot command line.

For example, in an Amazon Linux AMI with an earlier kernel, you can add it to the end of the kernel line
in the GRUB configuration found in /boot/grub/menu.lst:

kernel /boot/vmlinuz-4.4.5-15.26.amzn1.x86_64 root=LABEL=/ console=ttyS0

xen_blkfront.max=256

For a later kernel, the command would be similar to the following:

kernel /boot/vmlinuz-4.9.20-11.31.amzn1.x86_64 root=LABEL=/ console=tty1 console=ttyS0

xen_blkfront.max_indirect_segments=256

Reboot your instance for this setting to take effect.

For more information, see Configuring GRUB (p. 197). Other Linux distributions, especially those that do
not use the GRUB boot loader, may require a different approach to adjusting the kernel parameters.

For more information about EBS I/O characteristics, see the Amazon EBS: Designing for Performance
re:Invent presentation on this topic.

Use RAID 0 to maximize utilization of instance resources

Some instance types can drive more I/O throughput than what you can provision for a single EBS
volume. You can join multiple volumes together in a RAID 0 configuration to use the available bandwidth
for these instances. For more information, see RAID configuration on Linux (p. 1243).

Track performance using Amazon CloudWatch

Amazon Web Services provides performance metrics for Amazon EBS that you can analyze and view with
Amazon CloudWatch and status checks that you can use to monitor the health of your volumes. For more
information, see Monitor the status of your volumes (p. 1116).

I/O characteristics and monitoring

On a given volume configuration, certain I/O characteristics drive the performance behavior for your EBS
volumes. SSD-backed volumes—General Purpose SSD (gp2 and gp3) and Provisioned IOPS SSD (io1 and
io2)—deliver consistent performance whether an I/O operation is random or sequential. HDD-backed
volumes—Throughput Optimized HDD (st1) and Cold HDD (sc1)—deliver optimal performance only
when I/O operations are large and sequential. To understand how SSD and HDD volumes will perform in
your application, it is important to know the connection between demand on the volume, the quantity
of IOPS available to it, the time it takes for an I/O operation to complete, and the volume's throughput
limits.

IOPS
IOPS are a unit of measure representing input/output operations per second. The operations are
measured in KiB, and the underlying drive technology determines the maximum amount of data that a
volume type counts as a single I/O. I/O size is capped at 256 KiB for SSD volumes and 1,024 KiB for HDD
volumes because SSD volumes handle small or random I/O much more efficiently than HDD volumes.

When small I/O operations are physically contiguous, Amazon EBS attempts to merge them into a single
I/O operation up to the maximum size. For example, for SSD volumes, a single 1,024 KiB I/O operation

1239
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

counts as 4 operations (1,024÷256=4), while 8 contiguous I/O operations at 32 KiB each count as 1
operation (8×32=256). However, 8 random non-contiguous I/O operations at 32 KiB each count as 8
operations. In this case, each I/O operation under 32 KiB counts as 1 operation.

Similarly, for HDD-backed volumes, both a single 1,024 KiB I/O operation and 8 sequential 128 KiB
operations would count as one operation. However, 8 random 128 KiB I/O operations would count as 8
operations.

Consequently, when you create an SSD-backed volume supporting 3,000 IOPS (either by provisioning a
Provisioned IOPS SSD volume at 3,000 IOPS or by sizing a General Purpose SSD volume at 1,000 GiB),
and you attach it to an EBS-optimized instance that can provide sufficient bandwidth, you can transfer
up to 3,000 I/Os of data per second, with throughput determined by I/O size.

Volume queue length and latency

The volume queue length is the number of pending I/O requests for a device. Latency is the true end-to-
end client time of an I/O operation, in other words, the time elapsed between sending an I/O to EBS and
receiving an acknowledgement from EBS that the I/O read or write is complete. Queue length must be
correctly calibrated with I/O size and latency to avoid creating bottlenecks either on the guest operating
system or on the network link to EBS.

Optimal queue length varies for each workload, depending on your particular application's sensitivity to
IOPS and latency. If your workload is not delivering enough I/O requests to fully use the performance
available to your EBS volume, then your volume might not deliver the IOPS or throughput that you have
provisioned.

Transaction-intensive applications are sensitive to increased I/O latency and are well-suited for SSD-
backed volumes. You can maintain high IOPS while keeping latency down by maintaining a low queue
length and a high number of IOPS available to the volume. Consistently driving more IOPS to a volume
than it has available can cause increased I/O latency.

Throughput-intensive applications are less sensitive to increased I/O latency, and are well-suited for
HDD-backed volumes. You can maintain high throughput to HDD-backed volumes by maintaining a high
queue length when performing large, sequential I/O.

I/O size and volume throughput limits

For SSD-backed volumes, if your I/O size is very large, you may experience a smaller number of IOPS
than you provisioned because you are hitting the throughput limit of the volume. For example, a gp2
volume under 1,000 GiB with burst credits available has an IOPS limit of 3,000 and a volume throughput
limit of 250 MiB/s. If you are using a 256 KiB I/O size, your volume reaches its throughput limit at 1000
IOPS (1000 x 256 KiB = 250 MiB). For smaller I/O sizes (such as 16 KiB), this same volume can sustain
3,000 IOPS because the throughput is well below 250 MiB/s. (These examples assume that your volume's
I/O is not hitting the throughput limits of the instance.) For more information about the throughput
limits for each EBS volume type, see Amazon EBS volume types (p. 1084).

For smaller I/O operations, you may see a higher-than-provisioned IOPS value as measured from inside
your instance. This happens when the instance operating system merges small I/O operations into a
larger operation before passing them to Amazon EBS.

If your workload uses sequential I/Os on HDD-backed st1 and sc1 volumes, you may experience a
higher than expected number of IOPS as measured from inside your instance. This happens when the
instance operating system merges sequential I/Os and counts them in 1,024 KiB-sized units. If your
workload uses small or random I/Os, you may experience a lower throughput than you expect. This is
because we count each random, non-sequential I/O toward the total IOPS count, which can cause you to
hit the volume's IOPS limit sooner than expected.

Whatever your EBS volume type, if you are not experiencing the IOPS or throughput you expect in your
configuration, ensure that your EC2 instance bandwidth is not the limiting factor. You should always use
a current-generation, EBS-optimized instance (or one that includes 10 Gb/s network connectivity) for

1240
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

optimal performance. For more information, see Amazon EBS–optimized instances (p. 1217). Another
possible cause for not experiencing the expected IOPS is that you are not driving enough I/O to the EBS
volumes.

Monitor I/O characteristics using CloudWatch

You can monitor these I/O characteristics with each volume's CloudWatch volume metrics (p. 1252).
Important metrics to consider include the following:

• BurstBalance
• VolumeReadBytes
• VolumeWriteBytes
• VolumeReadOps
• VolumeWriteOps
• VolumeQueueLength

BurstBalance displays the burst bucket balance for gp2, st1, and sc1 volumes as a percentage of
the remaining balance. When your burst bucket is depleted, volume I/O (for gp2 volumes) or volume
throughput (for st1 and sc1 volumes) is throttled to the baseline. Check the BurstBalance value to
determine whether your volume is being throttled for this reason.

HDD-backed st1 and sc1 volumes are designed to perform best with workloads that take
advantage of the 1,024 KiB maximum I/O size. To determine your volume's average I/O size, divide
VolumeWriteBytes by VolumeWriteOps. The same calculation applies to read operations. If average
I/O size is below 64 KiB, increasing the size of the I/O operations sent to an st1 or sc1 volume should
improve performance.
Note
If average I/O size is at or near 44 KiB, you might be using an instance or kernel without support
for indirect descriptors. Any Linux kernel 3.8 and above has this support, as well as any current-
generation instance.

If your I/O latency is higher than you require, check VolumeQueueLength to make sure your application
is not trying to drive more IOPS than you have provisioned. If your application requires a greater number
of IOPS than your volume can provide, you should consider using a larger gp2 volume with a higher base
performance level or an io1 or io2 volume with more provisioned IOPS to achieve faster latencies.

Related resources
For more information about Amazon EBS I/O characteristics, see the following re:Invent presentation:
Amazon EBS: Designing for Performance.

Initialize Amazon EBS volumes

Empty EBS volumes receive their maximum performance the moment that they are created and do not
require initialization (formerly known as pre-warming).

For volumes that were created from snapshots, the storage blocks must be pulled down from Amazon
S3 and written to the volume before you can access them. This preliminary action takes time and can
cause a significant increase in the latency of I/O operations the first time each block is accessed. Volume
performance is achieved after all blocks have been downloaded and written to the volume.
Important
While initializing Provisioned IOPS SSD volumes that were created from snapshots, the
performance of the volume may drop below 50 percent of its expected level, which causes the
volume to display a warning state in the I/O Performance status check. This is expected, and
you can ignore the warning state on Provisioned IOPS SSD volumes while you are initializing
them. For more information, see EBS volume status checks (p. 1117).

1241
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

For most applications, amortizing the initialization cost over the lifetime of the volume is acceptable. To
avoid this initial performance hit in a production environment, you can use one of the following options:

• Force the immediate initialization of the entire volume. For more information, see Initialize Amazon
EBS volumes on Linux (p. 1242).
• Enable fast snapshot restore on a snapshot to ensure that the EBS volumes created from it are fully-
initialized at creation and instantly deliver all of their provisioned performance. For more information,
see Amazon EBS fast snapshot restore (p. 1210).

Initialize Amazon EBS volumes on Linux

Empty EBS volumes receive their maximum performance the moment that they are available and do
not require initialization (formerly known as pre-warming). For volumes that have been created from
snapshots, use the dd or fio utilities to read from all of the blocks on a volume. All existing data on the
volume will be preserved.

For information about initializing Amazon EBS volumes on Windows, see Initializing Amazon EBS
volumes on Windows.

To initialize a volume created from a snapshot on Linux

1. Attach the newly-restored volume to your Linux instance.

2. Use the lsblk command to list the block devices on your instance.

[ec2-user ~]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
xvdf 202:80 0 30G 0 disk
xvda1 202:1 0 8G 0 disk /

Here you can see that the new volume, /dev/xvdf, is attached, but not mounted (because there is
no path listed under the MOUNTPOINT column).
3. Use the dd or fio utilities to read all of the blocks on the device. The dd command is installed by
default on Linux systems, but fio is considerably faster because it allows multi-threaded reads.
Note
This step may take several minutes up to several hours, depending on your EC2 instance
bandwidth, the IOPS provisioned for the volume, and the size of the volume.

[dd] The if (input file) parameter should be set to the drive you wish to initialize. The of (output
file) parameter should be set to the Linux null virtual device, /dev/null. The bs parameter sets the
block size of the read operation; for optimal performance, this should be set to 1 MB.
Important
Incorrect use of dd can easily destroy a volume's data. Be sure to follow precisely the
example command below. Only the if=/dev/xvdf parameter will vary depending on the
name of the device you are reading.

[ec2-user ~]$ sudo dd if=/dev/xvdf of=/dev/null bs=1M

[fio] If you have fio installed on your system, use the following command to initialize your volume.
The --filename (input file) parameter should be set to the drive you wish to initialize.

[ec2-user ~]$ sudo fio --filename=/dev/xvdf --rw=read --bs=128k --iodepth=32 --

ioengine=libaio --direct=1 --name=volume-initialize

To install fio on Amazon Linux, use the following command:

1242
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

sudo yum install -y fio

To install fio on Ubuntu, use the following command:

sudo apt-get install -y fio

When the operation is finished, you will see a report of the read operation. Your volume is now ready
for use. For more information, see Make an Amazon EBS volume available for use on Linux (p. 1111).

RAID configuration on Linux

With Amazon EBS, you can use any of the standard RAID configurations that you can use with a
traditional bare metal server, as long as that particular RAID configuration is supported by the operating
system for your instance. This is because all RAID is accomplished at the software level. For greater I/O
performance than you can achieve with a single volume, RAID 0 can stripe multiple volumes together;
for on-instance redundancy, RAID 1 can mirror two volumes together.

Amazon EBS volume data is replicated across multiple servers in an Availability Zone to prevent the loss
of data from the failure of any single component. This replication makes Amazon EBS volumes ten times
more reliable than typical commodity disk drives. For more information, see Amazon EBS Availability and
Durability in the Amazon EBS product detail pages.
Note
You should avoid booting from a RAID volume. Grub is typically installed on only one device in
a RAID array, and if one of the mirrored devices fails, you may be unable to boot the operating
system.

If you need to create a RAID array on a Windows instance, see RAID Configuration on Windows in the
Amazon EC2 User Guide for Windows Instances.

Contents
• RAID configuration options (p. 1243)
• Create a RAID array on Linux (p. 1244)
• Create snapshots of volumes in a RAID array (p. 1247)

RAID configuration options

The following table compares the common RAID 0 and RAID 1 options.

Configuration Use Advantages Disadvantages

RAID 0 When I/O performance is I/O is distributed across Performance of the stripe
more important than fault the volumes in a stripe. If is limited to the worst
tolerance; for example, as you add a volume, you get performing volume in the
in a heavily used database the straight addition of set. Loss of a single volume
(where data replication is throughput and IOPS. results in a complete data
already set up separately). loss for the array.

RAID 1 When fault tolerance is Safer from the standpoint Does not provide a write
more important than I/O of data durability. performance improvement;
performance; for example, requires more Amazon
as in a critical application. EC2 to Amazon EBS
bandwidth than non-RAID
configurations because the

1243
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

Configuration Use Advantages Disadvantages

data is written to multiple
volumes simultaneously.

Important
RAID 5 and RAID 6 are not recommended for Amazon EBS because the parity write operations
of these RAID modes consume some of the IOPS available to your volumes. Depending on the
configuration of your RAID array, these RAID modes provide 20-30% fewer usable IOPS than
a RAID 0 configuration. Increased cost is a factor with these RAID modes as well; when using
identical volume sizes and speeds, a 2-volume RAID 0 array can outperform a 4-volume RAID 6
array that costs twice as much.

Creating a RAID 0 array allows you to achieve a higher level of performance for a file system than you
can provision on a single Amazon EBS volume. A RAID 1 array offers a "mirror" of your data for extra
redundancy. Before you perform this procedure, you need to decide how large your RAID array should be
and how many IOPS you want to provision.

The resulting size of a RAID 0 array is the sum of the sizes of the volumes within it, and the bandwidth is
the sum of the available bandwidth of the volumes within it. The resulting size and bandwidth of a RAID
1 array is equal to the size and bandwidth of the volumes in the array. For example, two 500 GiB io1
volumes with 4,000 provisioned IOPS each create a 1000 GiB RAID 0 array with an available bandwidth
of 8,000 IOPS and 1,000 MiB/s of throughput or a 500 GiB RAID 1 array with an available bandwidth of
4,000 IOPS and 500 MiB/s of throughput.

This documentation provides basic RAID setup examples. For more information about RAID
configuration, performance, and recovery, see the Linux RAID Wiki at https://raid.wiki.kernel.org/
index.php/Linux_Raid.

Create a RAID array on Linux

Use the following procedure to create the RAID array. Note that you can get directions for Windows
instances from Creating a RAID Array on Windows in the Amazon EC2 User Guide for Windows Instances.

To create a RAID array on Linux

1. Create the Amazon EBS volumes for your array. For more information, see Create an Amazon EBS
volume (p. 1104).
Important
Create volumes with identical size and IOPS performance values for your array. Make sure
you do not create an array that exceeds the available bandwidth of your EC2 instance. For
more information, see Amazon EBS–optimized instances (p. 1217).
2. Attach the Amazon EBS volumes to the instance that you want to host the array. For more
information, see Attach an Amazon EBS volume to an instance (p. 1106).
3. Use the mdadm command to create a logical RAID device from the newly attached Amazon EBS
volumes. Substitute the number of volumes in your array for number_of_volumes and the device
names for each volume in the array (such as /dev/xvdf) for device_name. You can also substitute
MY_RAID with your own unique name for the array.
Note
You can list the devices on your instance with the lsblk command to find the device names.

(RAID 0 only) To create a RAID 0 array, run the following command (note the --level=0 option to
stripe the array):

[ec2-user ~]$ sudo mdadm --create --verbose /dev/md0 --level=0 --name=MY_RAID --raid-
devices=number_of_volumes device_name1 device_name2

1244
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

(RAID 1 only) To create a RAID 1 array, run the following command (note the --level=1 option to
mirror the array):

[ec2-user ~]$ sudo mdadm --create --verbose /dev/md0 --level=1 --name=MY_RAID --raid-
devices=number_of_volumes device_name1 device_name2

4. Allow time for the RAID array to initialize and synchronize. You can track the progress of these
operations with the following command:

[ec2-user ~]$ sudo cat /proc/mdstat

The following is example output:

Personalities : [raid1]
md0 : active raid1 xvdg[1] xvdf[0]
20955008 blocks super 1.2 [2/2] [UU]
[=========>...........] resync = 46.8% (9826112/20955008) finish=2.9min
speed=63016K/sec

In general, you can display detailed information about your RAID array with the following command:

[ec2-user ~]$ sudo mdadm --detail /dev/md0

The following is example output:

/dev/md0:
Version : 1.2
Creation Time : Mon Jun 27 11:31:28 2016
Raid Level : raid1
Array Size : 20955008 (19.98 GiB 21.46 GB)
Used Dev Size : 20955008 (19.98 GiB 21.46 GB)
Raid Devices : 2
Total Devices : 2
Persistence : Superblock is persistent

Update Time : Mon Jun 27 11:37:02 2016

State : clean
...
...
...

Number Major Minor RaidDevice State

0 202 80 0 active sync /dev/sdf
1 202 96 1 active sync /dev/sdg

5. Create a file system on your RAID array, and give that file system a label to use when you mount
it later. For example, to create an ext4 file system with the label MY_RAID, run the following
command:

[ec2-user ~]$ sudo mkfs.ext4 -L MY_RAID /dev/md0

Depending on the requirements of your application or the limitations of your operating system, you
can use a different file system type, such as ext3 or XFS (consult your file system documentation for
the corresponding file system creation command).
6. To ensure that the RAID array is reassembled automatically on boot, create a configuration file to
contain the RAID information:

1245
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

[ec2-user ~]$ sudo mdadm --detail --scan | sudo tee -a /etc/mdadm.conf

Note
If you are using a Linux distribution other than Amazon Linux, this file may need to be
placed in different location. For more information, consult man mdadm.conf on your Linux
system..
7. Create a new ramdisk image to properly preload the block device modules for your new RAID
configuration:

[ec2-user ~]$ sudo dracut -H -f /boot/initramfs-$(uname -r).img $(uname -r)

8. Create a mount point for your RAID array.

[ec2-user ~]$ sudo mkdir -p /mnt/raid

9. Finally, mount the RAID device on the mount point that you created:

[ec2-user ~]$ sudo mount LABEL=MY_RAID /mnt/raid

Your RAID device is now ready for use.

10. (Optional) To mount this Amazon EBS volume on every system reboot, add an entry for the device
to the /etc/fstab file.

a. Create a backup of your /etc/fstab file that you can use if you accidentally destroy or delete
this file while you are editing it.

[ec2-user ~]$ sudo cp /etc/fstab /etc/fstab.orig

b. Open the /etc/fstab file using your favorite text editor, such as nano or vim.
c. Comment out any lines starting with "UUID=" and, at the end of the file, add a new line for your
RAID volume using the following format:

device_label mount_point file_system_type fs_mntops fs_freq fs_passno

The last three fields on this line are the file system mount options, the dump frequency of the
file system, and the order of file system checks done at boot time. If you don't know what these
values should be, then use the values in the example below for them (defaults,nofail 0
2). For more information about /etc/fstab entries, see the fstab manual page (by entering
man fstab on the command line). For example, to mount the ext4 file system on the device with
the label MY_RAID at the mount point /mnt/raid, add the following entry to /etc/fstab.
Note
If you ever intend to boot your instance without this volume attached (for example,
so this volume could move back and forth between different instances), you should
add the nofail mount option that allows the instance to boot even if there are
errors in mounting the volume. Debian derivatives, such as Ubuntu, must also add the
nobootwait mount option.

LABEL=MY_RAID /mnt/raid ext4 defaults,nofail 0 2

d. After you've added the new entry to /etc/fstab, you need to check that your entry works.
Run the sudo mount -a command to mount all file systems in /etc/fstab.

1246
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

[ec2-user ~]$ sudo mount -a

If the previous command does not produce an error, then your /etc/fstab file is OK and your
file system will mount automatically at the next boot. If the command does produce any errors,
examine the errors and try to correct your /etc/fstab.
Warning
Errors in the /etc/fstab file can render a system unbootable. Do not shut down a
system that has errors in the /etc/fstab file.
e. (Optional) If you are unsure how to correct /etc/fstab errors, you can always restore your
backup /etc/fstab file with the following command.

[ec2-user ~]$ sudo mv /etc/fstab.orig /etc/fstab

Create snapshots of volumes in a RAID array

If you want to back up the data on the EBS volumes in a RAID array using snapshots, you must
ensure that the snapshots are consistent. This is because the snapshots of these volumes are created
independently. To restore EBS volumes in a RAID array from snapshots that are out of sync would
degrade the integrity of the array.

To create a consistent set of snapshots for your RAID array, use EBS multi-volume snapshots. Multi-
volume snapshots allow you to take point-in-time, data coordinated, and crash-consistent snapshots
across multiple EBS volumes attached to an EC2 instance. You do not have to stop your instance to
coordinate between volumes to ensure consistency because snapshots are automatically taken across
multiple EBS volumes. For more information, see the steps for creating multi-volume snapshots under
Creating Amazon EBS snapshots.

Benchmark EBS volumes

You can test the performance of Amazon EBS volumes by simulating I/O workloads. The process is as
follows:

1. Launch an EBS-optimized instance.

2. Create new EBS volumes.
3. Attach the volumes to your EBS-optimized instance.
4. Configure and mount the block device.
5. Install a tool to benchmark I/O performance.
6. Benchmark the I/O performance of your volumes.
7. Delete your volumes and terminate your instance so that you don't continue to incur charges.

Important
Some of the procedures result in the destruction of existing data on the EBS volumes you
benchmark. The benchmarking procedures are intended for use on volumes specially created for
testing purposes, not production volumes.

Set up your instance

To get optimal performance from EBS volumes, we recommend that you use an EBS-optimized instance.
EBS-optimized instances deliver dedicated throughput between Amazon EC2 and Amazon EBS, with
instance. EBS-optimized instances deliver dedicated bandwidth between Amazon EC2 and Amazon EBS,
with specifications depending on the instance type. For more information, see Amazon EBS–optimized
instances (p. 1217).

1247
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

To create an EBS-optimized instance, choose Launch as an EBS-Optimized instance when launching the
instance using the Amazon EC2 console, or specify --ebs-optimized when using the command line. Be
sure that you launch a current-generation instance that supports this option. For more information, see
Amazon EBS–optimized instances (p. 1217).

Set up Provisioned IOPS SSD volumes

To create Provisioned IOPS SSD (io1 and io2) volumes using the Amazon EC2 console, choose the
Provisioned IOPS SSD (io1) or Provisioned IOPS SSD (io2) volume type. At the command line, specify
io1 or io2 for the --volume-type parameter and an integer between 100 and 64,000 for the --iops
parameter. For more information, see Amazon EBS volume types (p. 1084) and Create an Amazon EBS
volume (p. 1104).

For the example tests, we recommend that you create a RAID array with 6 volumes, which offers a high
level of performance. Because you are charged by gigabytes provisioned (and the number of provisioned
IOPS for Provisioned IOPS SSD volumes), not the number of volumes, there is no additional cost for
creating multiple, smaller volumes and using them to create a stripe set. If you're using Oracle Orion to
benchmark your volumes, it can simulate striping the same way that Oracle ASM does, so we recommend
that you let Orion do the striping. If you are using a different benchmarking tool, you need to stripe the
volumes yourself.

To create a six-volume stripe set on Amazon Linux, use a command such as the following:

[ec2-user ~]$ sudo mdadm --create /dev/md0 --level=0 --chunk=64 --raid-devices=6 /dev/sdf /
dev/sdg /dev/sdh /dev/sdi /dev/sdj /dev/sdk

For this example, the file system is XFS. Use the file system that meets your requirements. Use the
following command to install XFS file system support:

[ec2-user ~]$ sudo yum install -y xfsprogs

Then, use these commands to create, mount, and assign ownership to the XFS file system:

[ec2-user ~]$ sudo mkdir -p /mnt/p_iops_vol0 && sudo mkfs.xfs /dev/md0

[ec2-user ~]$ sudo mount -t xfs /dev/md0 /mnt/p_iops_vol0
[ec2-user ~]$ sudo chown ec2-user:ec2-user /mnt/p_iops_vol0/

Set up Throughput Optimized HDD (st1) or Cold HDD (sc1) volumes

To create an st1 volume, choose Throughput Optimized HDD when creating the volume using the
Amazon EC2 console, or specify --type st1 when using the command line. To create an sc1 volume,
choose Cold HDD when creating the volume using the Amazon EC2 console, or specify --type sc1
when using the command line. For information about creating EBS volumes, see Create an Amazon
EBS volume (p. 1104). For information about attaching these volumes to your instance, see Attach an
Amazon EBS volume to an instance (p. 1106).

AWS provides a JSON template for use with AWS CloudFormation that simplifies this setup procedure.
Access the template and save it as a JSON file. AWS CloudFormation allows you to configure your own
SSH keys and offers an easier way to set up a performance test environment to evaluate st1 volumes.
The template creates a current-generation instance and a 2 TiB st1 volume, and attaches the volume to
the instance at /dev/xvdf.

To create an HDD volume using the template

1. Open the AWS CloudFormation console at https://console.aws.amazon.com/cloudformation.

2. Choose Create Stack.
3. Choose Upload a Template to Amazon S3 and select the JSON template you previously obtained.

1248
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

4. Give your stack a name like “ebs-perf-testing”, and select an instance type (the default is r3.8xlarge)
and SSH key.
5. Choose Next twice, and then choose Create Stack.
6. After the status for your new stack moves from CREATE_IN_PROGRESS to COMPLETE, choose
Outputs to get the public DNS entry for your new instance, which will have a 2 TiB st1 volume
attached to it.
7. Connect using SSH to your new stack as user ec2-user, with the hostname obtained from the DNS
entry in the previous step.
8. Proceed to Install benchmark tools (p. 1249).

Install benchmark tools

The following table lists some of the possible tools you can use to benchmark the performance of EBS
volumes.

Tool Description

fio For benchmarking I/O performance. (Note that fio has a dependency on libaio-

devel.)

To install fio on Amazon Linux, run the following command:

[ec2-user ~]$ sudo yum install -y fio

To install fio on Ubuntu, run the following command:

sudo apt-get install -y fio

Oracle Orion For calibrating the I/O performance of storage systems to be used with Oracle
Calibration Tool databases.

These benchmarking tools support a wide variety of test parameters. You should use commands that
approximate the workloads your volumes will support. These commands provided below are intended as
examples to help you get started.

Choose the volume queue length

Choosing the best volume queue length based on your workload and volume type.

Queue length on SSD-backed volumes

To determine the optimal queue length for your workload on SSD-backed volumes, we recommend that
you target a queue length of 1 for every 1000 IOPS available (baseline for General Purpose SSD volumes
and the provisioned amount for Provisioned IOPS SSD volumes). Then you can monitor your application
performance and tune that value based on your application requirements.

Increasing the queue length is beneficial until you achieve the provisioned IOPS, throughput or optimal
system queue length value, which is currently set to 32. For example, a volume with 3,000 provisioned
IOPS should target a queue length of 3. You should experiment with tuning these values up or down to
see what performs best for your application.

Queue length on HDD-backed volumes

To determine the optimal queue length for your workload on HDD-backed volumes, we recommend that
you target a queue length of at least 4 while performing 1MiB sequential I/Os. Then you can monitor

1249
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

your application performance and tune that value based on your application requirements. For example,
a 2 TiB st1 volume with burst throughput of 500 MiB/s and IOPS of 500 should target a queue length
of 4, 8, or 16 while performing 1,024 KiB, 512 KiB, or 256 KiB sequential I/Os respectively. You should
experiment with tuning these values value up or down to see what performs best for your application.

Disable C-states
Before you run benchmarking, you should disable processor C-states. Temporarily idle cores in a
supported CPU can enter a C-state to save power. When the core is called on to resume processing, a
certain amount of time passes until the core is again fully operational. This latency can interfere with
processor benchmarking routines. For more information about C-states and which EC2 instance types
support them, see Processor State Control for Your EC2 Instance.

Disable C-states on Linux

You can disable C-states on Amazon Linux, RHEL, and CentOS as follows:

1. Get the number of C-states.

$ cpupower idle-info | grep "Number of idle states:"

2. Disable the C-states from c1 to cN. Ideally, the cores should be in state c0.

$ for i in `seq 1 $((N-1))`; do cpupower idle-set -d $i; done

Perform benchmarking
The following procedures describe benchmarking commands for various EBS volume types.

Run the following commands on an EBS-optimized instance with attached EBS volumes. If the EBS
volumes were created from snapshots, be sure to initialize them before benchmarking. For more
information, see Initialize Amazon EBS volumes (p. 1241).

When you are finished testing your volumes, see the following topics for help cleaning up: Delete an
Amazon EBS volume (p. 1126) and Terminate your instance (p. 643).

Benchmark Provisioned IOPS SSD volumes

Run fio on the stripe set that you created.

The following command performs 16 KB random write operations.

[ec2-user ~]$ sudo fio --directory=/mnt/p_iops_vol0 --name fio_test_file --direct=1 --

rw=randwrite --bs=16k --size=1G --numjobs=16 --time_based --runtime=180 --group_reporting
--norandommap

The following command performs 16 KB random read operations.

[ec2-user ~]$ sudo fio --directory=/mnt/p_iops_vol0 --name fio_test_file --direct=1 --

rw=randread --bs=16k --size=1G --numjobs=16 --time_based --runtime=180 --group_reporting --
norandommap

For more information about interpreting the results, see this tutorial: Inspecting disk IO performance
with fio.

Benchmark st1 and sc1 volumes

Run fio on your st1 or sc1 volume.

1250
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS performance

Note
Prior to running these tests, set buffered I/O on your instance as described in Increase read-
ahead for high-throughput, read-heavy workloads on st1 and sc1 (p. 1238).

The following command performs 1 MiB sequential read operations against an attached st1 block
device (e.g., /dev/xvdf):

[ec2-user ~]$ sudo fio --filename=/dev/<device> --direct=1 --rw=read --randrepeat=0

--ioengine=libaio --bs=1024k --iodepth=8 --time_based=1 --runtime=180 --
name=fio_direct_read_test

The following command performs 1 MiB sequential write operations against an attached st1 block
device:

[ec2-user ~]$ sudo fio --filename=/dev/<device> --direct=1 --rw=write --randrepeat=0

--ioengine=libaio --bs=1024k --iodepth=8 --time_based=1 --runtime=180 --
name=fio_direct_write_test

Some workloads perform a mix of sequential reads and sequential writes to different parts of the block
device. To benchmark such a workload, we recommend that you use separate, simultaneous fio jobs for
reads and writes, and use the fio offset_increment option to target different block device locations
for each job.

Running this workload is a bit more complicated than a sequential-write or sequential-read workload.
Use a text editor to create a fio job file, called fio_rw_mix.cfg in this example, that contains the
following:

[global]
clocksource=clock_gettime
randrepeat=0
runtime=180
offset_increment=100g

[sequential-write]
bs=1M
ioengine=libaio
direct=1
iodepth=8
filename=/dev/<device>
do_verify=0
rw=write
rwmixread=0
rwmixwrite=100

[sequential-read]
bs=1M
ioengine=libaio
direct=1
iodepth=8
filename=/dev/<device>
do_verify=0
rw=read
rwmixread=100
rwmixwrite=0

Then run the following command:

[ec2-user ~]$ sudo fio fio_rw_mix.cfg

1251
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch metrics

For more information about interpreting the results, see this tutorial: Inspecting disk I/O performance
with fio.

Multiple fio jobs for direct I/O, even though using sequential read or write operations, can result in lower
than expected throughput for st1 and sc1 volumes. We recommend that you use one direct I/O job and
use the iodepth parameter to control the number of concurrent I/O operations.

Amazon CloudWatch metrics for Amazon EBS

Amazon CloudWatch metrics are statistical data that you can use to view, analyze, and set alarms on the
operational behavior of your volumes.

Data is available automatically in 1-minute periods at no charge.

When you get data from CloudWatch, you can include a Period request parameter to specify the
granularity of the returned data. This is different than the period that we use when we collect the data
(1-minute periods). We recommend that you specify a period in your request that is equal to or larger
than the collection period to ensure that the returned data is valid.

You can get the data using either the CloudWatch API or the Amazon EC2 console. The console takes the
raw data from the CloudWatch API and displays a series of graphs based on the data. Depending on your
needs, you might prefer to use either the data from the API or the graphs in the console.

Amazon EBS metrics

Amazon Elastic Block Store (Amazon EBS) sends data points to CloudWatch for several metrics. All EBS
volume types automatically send 1-minute metrics to CloudWatch. Data is reported to CloudWatch only
when the volume is attached to an instance

Some of these metrics have differences on instances built on the Nitro System. For a list of these instance
types, see Instances built on the Nitro System (p. 209).

The AWS/EBS namespace includes the following metrics.

Metrics
• Volume metrics (p. 1252)
• Fast snapshot restore metrics (p. 1256)

Volume metrics
The AWS/EBS namespace includes the following metrics for EBS volumes. To get information about the
available disk space from the operating system on an instance, see View free disk space (p. 1116).

Metric Description

VolumeReadBytes Provides information on the read operations in a specified period

of time. The Sum statistic reports the total number of bytes
transferred during the period. The Average statistic reports
the average size of each read operation during the period,
except on volumes attached to a Nitro-based instance, where
the average represents the average over the specified period.
The SampleCount statistic reports the total number of read
operations during the period, except on volumes attached to a
Nitro-based instance, where the sample count represents the
number of data points used in the statistical calculation. For Xen
instances, data is reported only when there is read activity on the
volume.

1252
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch metrics

Metric Description
The Minimum and Maximum statistics on this metric are
supported only by volumes attached to Nitro-based instances.

Units: Bytes

VolumeWriteBytes Provides information on the write operations in a specified

period of time. The Sum statistic reports the total number of
bytes transferred during the period. The Average statistic
reports the average size of each write operation during the
period, except on volumes attached to a Nitro-based instance,
where the average represents the average over the specified
period. The SampleCount statistic reports the total number of
write operations during the period, except on volumes attached
to a Nitro-based instance, where the sample count represents the
number of data points used in the statistical calculation. For Xen
instances, data is reported only when there is write activity on
the volume.

The Minimum and Maximum statistics on this metric are

supported only by volumes attached to Nitro-based instances.

Units: Bytes

VolumeReadOps The total number of read operations in a specified period of time.

To calculate the average read operations per second (read IOPS)

for the period, divide the total read operations in the period by
the number of seconds in that period.

The Minimum and Maximum statistics on this metric are

supported only by volumes attached to Nitro-based instances.

Units: Count

VolumeWriteOps The total number of write operations in a specified period of

time.

To calculate the average write operations per second (write IOPS)

for the period, divide the total write operations in the period by
the number of seconds in that period.

The Minimum and Maximum statistics on this metric are

supported only by volumes attached to Nitro-based instances.

Units: Count

1253
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch metrics

Metric Description

VolumeTotalReadTime Note
This metric is not supported with Multi-Attach enabled
volumes.

The total number of seconds spent by all read operations that

completed in a specified period of time. If multiple requests are
submitted at the same time, this total could be greater than the
length of the period. For example, for a period of 1 minutes (60
seconds): if 150 operations completed during that period, and
each operation took 1 second, the value would be 150 seconds.
For Xen instances, data is reported only when there is read
activity on the volume.

The Average statistic on this metric is not relevant for volumes

attached to Nitro-based instances.

The Minimum and Maximum statistics on this metric are

supported only by volumes attached to Nitro-based instances.

Units: Seconds

VolumeTotalWriteTime Note
This metric is not supported with Multi-Attach enabled
volumes.

The total number of seconds spent by all write operations that

completed in a specified period of time. If multiple requests are
submitted at the same time, this total could be greater than the
length of the period. For example, for a period of 1 minute (60
seconds): if 150 operations completed during that period, and
each operation took 1 second, the value would be 150 seconds.
For Xen instances, data is reported only when there is write
activity on the volume.

The Average statistic on this metric is not relevant for volumes

attached to Nitro-based instances.

The Minimum and Maximum statistics on this metric are

supported only by volumes attached to Nitro-based instances.

Units: Seconds

VolumeIdleTime Note
This metric is not supported with Multi-Attach enabled
volumes.

The total number of seconds in a specified period of time when

no read or write operations were submitted.

The Average statistic on this metric is not relevant for volumes

attached to Nitro-based instances.

The Minimum and Maximum statistics on this metric are

supported only by volumes attached to Nitro-based instances.

Units: Seconds

1254
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch metrics

Metric Description

VolumeQueueLength The number of read and write operation requests waiting to be

completed in a specified period of time.

The Sum statistic on this metric is not relevant for volumes

attached to Nitro-based instances.

The Minimum and Maximum statistics on this metric are

supported only by volumes attached to Nitro-based instances.

Units: Count

VolumeThroughputPercentage Note
This metric is not supported with Multi-Attach enabled
volumes.

Used with Provisioned IOPS SSD volumes only. The percentage

of I/O operations per second (IOPS) delivered of the total IOPS
provisioned for an Amazon EBS volume. Provisioned IOPS SSD
volumes deliver their provisioned performance 99.9 percent of
the time.

During a write, if there are no other pending I/O requests in a

minute, the metric value will be 100 percent. Also, a volume's I/O
performance may become degraded temporarily due to an action
you have taken (for example, creating a snapshot of a volume
during peak usage, running the volume on a non-EBS-optimized
instance, or accessing data on the volume for the first time).

Units: Percent

VolumeConsumedReadWriteOps Used with Provisioned IOPS SSD volumes only. The total amount
of read and write operations (normalized to 256K capacity units)
consumed in a specified period of time.

I/O operations that are smaller than 256K each count as 1

consumed IOPS. I/O operations that are larger than 256K are
counted in 256K capacity units. For example, a 1024K I/O would
count as 4 consumed IOPS.

Units: Count

1255
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch metrics

Metric Description

BurstBalance Used with General Purpose SSD (gp2), Throughput Optimized

HDD (st1), and Cold HDD (sc1) volumes only. Provides
information about the percentage of I/O credits (for gp2) or
throughput credits (for st1 and sc1) remaining in the burst
bucket. Data is reported to CloudWatch only when the volume is
active. If the volume is not attached, no data is reported.

The Sum statistic on this metric is not relevant for volumes

attached to instances built on the Nitro System.

If the baseline performance of the volume exceeds the maximum

burst performance, credits are never spent. If the volume is
attached to an instance built on the Nitro System, the burst
balance is not reported. For other instances, the reported burst
balance is 100%. For more information, see I/O Credits and burst
performance (p. 1087).

Units: Percent

Fast snapshot restore metrics

AWS/EBS namespace includes the following metrics for fast snapshot restore (p. 1210).

Metric Description

The maximum number of volume create credits that can be

FastSnapshotRestoreCreditsBucketSize
accumulated. This metric is reported per snapshot per Availability
Zone.

The most meaningful statistic is Average. The results for the

Minimum and Maximum statistics are the same as for Average
and could be used instead.

The number of volume create credits available. This metric is

FastSnapshotRestoreCreditsBalance
reported per snapshot per Availability Zone.

The most meaningful statistic is Average. The results for the

Minimum and Maximum statistics are the same as for Average
and could be used instead.

Dimensions for Amazon EBS metrics

The supported dimension is the volume ID (VolumeId). All available statistics are filtered by volume ID.

For the volume metrics (p. 1252), the supported dimension is the volume ID (VolumeId). All available
statistics are filtered by volume ID.

For the fast snapshot restore metrics (p. 1256), the supported dimensions are the snapshot ID
(SnapshotId) and the Availability Zone (AvailabilityZone).

Graphs in the Amazon EC2 console

After you create a volume, you can view the volume's monitoring graphs in the Amazon EC2 console.
Select a volume on the Volumes page in the console and choose Monitoring. The following table lists

1256
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch metrics

the graphs that are displayed. The column on the right describes how the raw data metrics from the
CloudWatch API are used to produce each graph. The period for all the graphs is 5 minutes.

Graph Description using raw metrics

Read Bandwidth (KiB/s) Sum(VolumeReadBytes) / Period / 1024

Write Bandwidth (KiB/s) Sum(VolumeWriteBytes) / Period / 1024

Read Throughput (IOPS) Sum(VolumeReadOps) / Period

Write Throughput (IOPS) Sum(VolumeWriteOps) / Period

Avg Queue Length (Operations) Avg(VolumeQueueLength)

% Time Spent Idle Sum(VolumeIdleTime) / Period × 100

Avg Read Size (KiB/Operation) Avg(VolumeReadBytes) / 1024

For Nitro-based instances, the following formula derives

Average Read Size using CloudWatch Metric Math:

(Sum(VolumeReadBytes) / Sum(VolumeReadOps)) /
1024

The VolumeReadBytes and VolumeReadOps metrics are

available in the EBS CloudWatch console.

Avg Write Size (KiB/Operation) Avg(VolumeWriteBytes) / 1024

For Nitro-based instances, the following formula derives

Average Write Size using CloudWatch Metric Math:

(Sum(VolumeWriteBytes) / Sum(VolumeWriteOps)) /
1024

The VolumeWriteBytes and VolumeWriteOps metrics are

available in the EBS CloudWatch console.

Avg Read Latency (ms/Operation) Avg(VolumeTotalReadTime) × 1000

For Nitro-based instances, the following formula derives

Average Read Latency using CloudWatch Metric Math:

(Sum(VolumeTotalReadTime) / Sum(VolumeReadOps))
× 1000

The VolumeTotalReadTime and VolumeReadOps metrics are

available in the EBS CloudWatch console.

Avg Write Latency (ms/Operation) Avg(VolumeTotalWriteTime) × 1000

For Nitro-based instances, the following formula derives

Average Write Latency using CloudWatch Metric Math:

(Sum(VolumeTotalWriteTime) /
Sum(VolumeWriteOps)) * 1000

The VolumeTotalWriteTime and VolumeWriteOps metrics

are available in the EBS CloudWatch console.

1257
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch events

For the average latency graphs and average size graphs, the average is calculated over the total number
of operations (read or write, whichever is applicable to the graph) that completed during the period.

Amazon CloudWatch Events for Amazon EBS

Amazon EBS emits notifications based on Amazon CloudWatch Events for a variety of volume,
snapshot, and encryption status changes. With CloudWatch Events, you can establish rules that trigger
programmatic actions in response to a change in volume, snapshot, or encryption key state. For example,
when a snapshot is created, you can trigger an AWS Lambda function to share the completed snapshot
with another account or copy it to another Region for disaster-recovery purposes.

Events in CloudWatch are represented as JSON objects. The fields that are unique to the event are
contained in the "detail" section of the JSON object. The "event" field contains the event name. The
"result" field contains the completed status of the action that triggered the event. For more information,
see Event Patterns in CloudWatch Events in the Amazon CloudWatch Events User Guide.

For more information, see Using Events in the Amazon CloudWatch User Guide.

Contents
• EBS volume events (p. 1258)
• EBS snapshot events (p. 1261)
• EBS volume modification events (p. 1265)
• EBS fast snapshot restore events (p. 1265)
• Using AWS Lambda to handle CloudWatch events (p. 1266)

EBS volume events

Amazon EBS sends events to CloudWatch Events when the following volume events occur.

Events
• Create volume (createVolume) (p. 1258)
• Delete volume (deleteVolume) (p. 1259)
• Volume attach or reattach (attachVolume, reattachVolume) (p. 1260)

Create volume (createVolume)

The createVolume event is sent to your AWS account when an action to create a volume completes.
However it is not saved, logged, or archived. This event can have a result of either available or
failed. Creation will fail if an invalid KMS key was provided, as shown in the examples below.

Event data

The listing below is an example of a JSON object emitted by EBS for a successful createVolume event.

{
"version": "0",
"id": "01234567-0123-0123-0123-012345678901",
"detail-type": "EBS Volume Notification",
"source": "aws.ec2",
"account": "012345678901",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:012345678901:volume/vol-01234567"

1258
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch events

],
"detail": {
"result": "available",
"cause": "",
"event": "createVolume",
"request-id": "01234567-0123-0123-0123-0123456789ab"
}
}

The listing below is an example of a JSON object emitted by EBS after a failed createVolume event.
The cause for the failure was a disabled KMS key.

{
"version": "0",
"id": "01234567-0123-0123-0123-0123456789ab",
"detail-type": "EBS Volume Notification",
"source": "aws.ec2",
"account": "012345678901",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "sa-east-1",
"resources": [
"arn:aws:ec2:sa-east-1:0123456789ab:volume/vol-01234567",
],
"detail": {
"event": "createVolume",
"result": "failed",
"cause": "arn:aws:kms:sa-east-1:0123456789ab:key/01234567-0123-0123-0123-0123456789ab
is disabled.",
"request-id": "01234567-0123-0123-0123-0123456789ab",
}
}

The following is an example of a JSON object that is emitted by EBS after a failed createVolume event.
The cause for the failure was a KMS key pending import.

{
"version": "0",
"id": "01234567-0123-0123-0123-0123456789ab",
"detail-type": "EBS Volume Notification",
"source": "aws.ec2",
"account": "012345678901",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "sa-east-1",
"resources": [
"arn:aws:ec2:sa-east-1:0123456789ab:volume/vol-01234567",
],
"detail": {
"event": "createVolume",
"result": "failed",
"cause": "arn:aws:kms:sa-east-1:0123456789ab:key/01234567-0123-0123-0123-0123456789ab
is pending import.",
"request-id": "01234567-0123-0123-0123-0123456789ab",
}
}

Delete volume (deleteVolume)

The deleteVolume event is sent to your AWS account when an action to delete a volume completes.
However it is not saved, logged, or archived. This event has the result deleted. If the deletion does not
complete, the event is never sent.

Event data

1259
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch events

The listing below is an example of a JSON object emitted by EBS for a successful deleteVolume event.

{
"version": "0",
"id": "01234567-0123-0123-0123-012345678901",
"detail-type": "EBS Volume Notification",
"source": "aws.ec2",
"account": "012345678901",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:012345678901:volume/vol-01234567"
],
"detail": {
"result": "deleted",
"cause": "",
"event": "deleteVolume",
"request-id": "01234567-0123-0123-0123-0123456789ab"
}
}

Volume attach or reattach (attachVolume, reattachVolume)

The attachVolume or reattachVolume event is sent to your AWS account if a volume fails to attach
or reattach to an instance. However it is not saved, logged, or archived. If you use a KMS key to encrypt
an EBS volume and the key becomes invalid, EBS will emit an event if that key is later used to attach or
reattach to an instance, as shown in the examples below.

Event data

The listing below is an example of a JSON object emitted by EBS after a failed attachVolume event.
The cause for the failure was a KMS key pending deletion.
Note
AWS may attempt to reattach to a volume following routine server maintenance.

{
"version": "0",
"id": "01234567-0123-0123-0123-0123456789ab",
"detail-type": "EBS Volume Notification",
"source": "aws.ec2",
"account": "012345678901",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:0123456789ab:volume/vol-01234567",
"arn:aws:kms:us-east-1:0123456789ab:key/01234567-0123-0123-0123-0123456789ab"
],
"detail": {
"event": "attachVolume",
"result": "failed",
"cause": "arn:aws:kms:us-east-1:0123456789ab:key/01234567-0123-0123-0123-0123456789ab
is pending deletion.",
"request-id": ""
}
}

The listing below is an example of a JSON object emitted by EBS after a failed reattachVolume event.
The cause for the failure was a KMS key pending deletion.

1260
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch events

"version": "0",
"id": "01234567-0123-0123-0123-0123456789ab",
"detail-type": "EBS Volume Notification",
"source": "aws.ec2",
"account": "012345678901",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:0123456789ab:volume/vol-01234567",
"arn:aws:kms:us-east-1:0123456789ab:key/01234567-0123-0123-0123-0123456789ab"
],
"detail": {
"event": "reattachVolume",
"result": "failed",
"cause": "arn:aws:kms:us-east-1:0123456789ab:key/01234567-0123-0123-0123-0123456789ab
is pending deletion.",
"request-id": ""
}
}

EBS snapshot events

Amazon EBS sends events to CloudWatch Events when the following volume events occur.

Events
• Create snapshot (createSnapshot) (p. 1261)
• Create snapshots (createSnapshots) (p. 1262)
• Copy snapshot (copySnapshot) (p. 1263)
• Share snapshot (shareSnapshot) (p. 1264)

Create snapshot (createSnapshot)

The createSnapshot event is sent to your AWS account when an action to create a snapshot
completes. However it is not saved, logged, or archived. This event can have a result of either
succeeded or failed.

Event data

The listing below is an example of a JSON object emitted by EBS for a successful createSnapshot
event. In the detail section, the source field contains the ARN of the source volume. The startTime
and endTime fields indicate when creation of the snapshot started and completed.

{
"version": "0",
"id": "01234567-0123-0123-0123-012345678901",
"detail-type": "EBS Snapshot Notification",
"source": "aws.ec2",
"account": "012345678901",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-west-2::snapshot/snap-01234567"
],
"detail": {
"event": "createSnapshot",
"result": "succeeded",
"cause": "",
"request-id": "",
"snapshot_id": "arn:aws:ec2:us-west-2::snapshot/snap-01234567",
"source": "arn:aws:ec2:us-west-2::volume/vol-01234567",

1261
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch events

"startTime": "yyyy-mm-ddThh:mm:ssZ",
"endTime": "yyyy-mm-ddThh:mm:ssZ" }
}

Create snapshots (createSnapshots)

The createSnapshots event is sent to your AWS account when an action to create a multi-volume
snapshot completes. This event can have a result of either succeeded or failed.

Event data

The listing below is an example of a JSON object emitted by EBS for a successful createSnapshots
event. In the detail section, the source field contains the ARNs of the source volumes of the multi-
volume snapshot set. The startTime and endTime fields indicate when creation of the snapshot
started and completed.

{
"version": "0",
"id": "01234567-0123-0123-0123-012345678901",
"detail-type": "EBS Multi-Volume Snapshots Completion Status",
"source": "aws.ec2",
"account": "012345678901",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-1",
"resources": [
"arn:aws:ec2::us-east-1:snapshot/snap-01234567",
"arn:aws:ec2::us-east-1:snapshot/snap-012345678"
],
"detail": {
"event": "createSnapshots",
"result": "succeeded",
"cause": "",
"request-id": "",
"startTime": "yyyy-mm-ddThh:mm:ssZ",
"endTime": "yyyy-mm-ddThh:mm:ssZ",
"snapshots": [
{
"snapshot_id": "arn:aws:ec2::us-east-1:snapshot/snap-01234567",
"source": "arn:aws:ec2::us-east-1:volume/vol-01234567",
"status": "completed"
},
{
"snapshot_id": "arn:aws:ec2::us-east-1:snapshot/snap-012345678",
"source": "arn:aws:ec2::us-east-1:volume/vol-012345678",
"status": "completed"
}
]
}
}

The listing below is an example of a JSON object emitted by EBS after a failed createSnapshots event.
The cause for the failure was one or more snapshots failed to complete. The values of snapshot_id are
the ARNs of the failed snapshots. startTime and endTime represent when the create-snapshots action
started and ended.

{
"version": "0",
"id": "01234567-0123-0123-0123-012345678901",
"detail-type": "EBS Multi-Volume Snapshots Completion Status",
"source": "aws.ec2",
"account": "012345678901",
"time": "yyyy-mm-ddThh:mm:ssZ",

1262
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch events

"region": "us-east-1",
"resources": [
"arn:aws:ec2::us-east-1:snapshot/snap-01234567",
"arn:aws:ec2::us-east-1:snapshot/snap-012345678"
],
"detail": {
"event": "createSnapshots",
"result": "failed",
"cause": "Snapshot snap-01234567 is in status deleted",
"request-id": "",
"startTime": "yyyy-mm-ddThh:mm:ssZ",
"endTime": "yyyy-mm-ddThh:mm:ssZ",
"snapshots": [
{
"snapshot_id": "arn:aws:ec2::us-east-1:snapshot/snap-01234567",
"source": "arn:aws:ec2::us-east-1:volume/vol-01234567",
"status": "error"
},
{
"snapshot_id": "arn:aws:ec2::us-east-1:snapshot/snap-012345678",
"source": "arn:aws:ec2::us-east-1:volume/vol-012345678",
"status": "deleted"
}
]
}
}

Copy snapshot (copySnapshot)

The copySnapshot event is sent to your AWS account when an action to copy a snapshot completes.
However it is not saved, logged, or archived. This event can have a result of either succeeded or
failed.

Event data

The listing below is an example of a JSON object emitted by EBS after a successful copySnapshot
event. The value of snapshot_id is the ARN of the newly created snapshot. In the detail section, the
value of source is the ARN of the source snapshot. startTime and endTime represent when the copy-
snapshot action started and ended.

{
"version": "0",
"id": "01234567-0123-0123-0123-012345678901",
"detail-type": "EBS Snapshot Notification",
"source": "aws.ec2",
"account": "123456789012",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-west-2::snapshot/snap-01234567"
],
"detail": {
"event": "copySnapshot",
"result": "succeeded",
"cause": "",
"request-id": "",
"snapshot_id": "arn:aws:ec2:us-west-2::snapshot/snap-01234567",
"source": "arn:aws:ec2:eu-west-1::snapshot/snap-76543210",
"startTime": "yyyy-mm-ddThh:mm:ssZ",
"endTime": "yyyy-mm-ddThh:mm:ssZ",
"Incremental": "True"
}
}

1263
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch events

The listing below is an example of a JSON object emitted by EBS after a failed copySnapshot event.
The cause for the failure was an invalid source snapshot ID. The value of snapshot_id is the ARN of
the failed snapshot. In the detail section, the value of source is the ARN of the source snapshot.
startTime and endTime represent when the copy-snapshot action started and ended.

{
"version": "0",
"id": "01234567-0123-0123-0123-012345678901",
"detail-type": "EBS Snapshot Notification",
"source": "aws.ec2",
"account": "123456789012",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-west-2::snapshot/snap-01234567"
],
"detail": {
"event": "copySnapshot",
"result": "failed",
"cause": "Source snapshot ID is not valid",
"request-id": "",
"snapshot_id": "arn:aws:ec2:us-west-2::snapshot/snap-01234567",
"source": "arn:aws:ec2:eu-west-1::snapshot/snap-76543210",
"startTime": "yyyy-mm-ddThh:mm:ssZ",
"endTime": "yyyy-mm-ddThh:mm:ssZ"
}
}

Share snapshot (shareSnapshot)

The shareSnapshot event is sent to your AWS account when another account shares a snapshot with it.
However it is not saved, logged, or archived. The result is always succeeded.

Event data

The following is an example of a JSON object emitted by EBS after a completed shareSnapshot event.
In the detail section, the value of source is the AWS account number of the user that shared the
snapshot with you. startTime and endTime represent when the share-snapshot action started and
ended. The shareSnapshot event is emitted only when a private snapshot is shared with another user.
Sharing a public snapshot does not trigger the event.

{
"version": "0",
"id": "01234567-01234-0123-0123-012345678901",
"detail-type": "EBS Snapshot Notification",
"source": "aws.ec2",
"account": "012345678901",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-west-2::snapshot/snap-01234567"
],
"detail": {
"event": "shareSnapshot",
"result": "succeeded",
"cause": "",
"request-id": "",
"snapshot_id": "arn:aws:ec2:us-west-2::snapshot/snap-01234567",
"source": 012345678901,
"startTime": "yyyy-mm-ddThh:mm:ssZ",
"endTime": "yyyy-mm-ddThh:mm:ssZ"
}

1264
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch events

EBS volume modification events

Amazon EBS sends modifyVolume events to CloudWatch Events when a volume is modified. However it
is not saved, logged, or archived.

{
"version": "0",
"id": "01234567-0123-0123-0123-012345678901",
"detail-type": "EBS Volume Notification",
"source": "aws.ec2",
"account": "012345678901",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1:012345678901:volume/vol-03a55cf56513fa1b6"
],
"detail": {
"result": "optimizing",
"cause": "",
"event": "modifyVolume",
"request-id": "01234567-0123-0123-0123-0123456789ab"
}
}

EBS fast snapshot restore events

Amazon EBS sends events to CloudWatch Events when the state of fast snapshot restore for a snapshot
changes. Events are emitted on a best effort basis.

The following is example data for this event.

{
"version": "0",
"id": "01234567-0123-0123-0123-012345678901",
"detail-type": "EBS Fast Snapshot Restore State-change Notification",
"source": "aws.ec2",
"account": "123456789012",
"time": "yyyy-mm-ddThh:mm:ssZ",
"region": "us-east-1",
"resources": [
"arn:aws:ec2:us-east-1::snapshot/snap-03a55cf56513fa1b6"
],
"detail": {
"snapshot-id": "snap-1234567890abcdef0",
"state": "optimizing",
"zone": "us-east-1a",
"message": "Client.UserInitiated - Lifecycle state transition",
}
}

The possible values for state are enabling, optimizing, enabled, disabling, and disabled.

The possible values for message are as follows:

Client.InvalidSnapshot.InvalidState - The requested snapshot transitioned to an

invalid state (Error)

A request to enable fast snapshot restore failed and the state transitioned to disabling or
disabled. Fast snapshot restore cannot be enabled for this snapshot.

1265
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch events

Client.UserInitiated

The state successfully transitioned to enabling or disabling.

Client.UserInitiated - Lifecycle state transition

The state successfully transitioned to optimizing, enabled, or disabled.

Server.InsufficientCapacity - There was insufficient capacity available to
satisfy the request

A request to enable fast snapshot restore failed due to insufficient capacity, and the state
transitioned to disabling or disabled. Wait and then try again.
Server.InternalError - An internal error caused the operation to fail

A request to enable fast snapshot restore failed due to an internal error, and the state transitioned
to disabling or disabled. Wait and then try again.
Client.InvalidSnapshot.InvalidState - The requested snapshot was deleted or
access permissions were revoked

The fast snapshot restore state for the snapshot has transitioned to disabling or disabled
because the snapshot was deleted or unshared by the snapshot owner. Fast snapshot restore cannot
be enabled for a snapshot that has been deleted or is no longer shared with you.

Using AWS Lambda to handle CloudWatch events

You can use Amazon EBS and CloudWatch Events to automate your data-backup workflow. This requires
you to create an IAM policy, a AWS Lambda function to handle the event, and an Amazon CloudWatch
Events rule that matches incoming events and routes them to the Lambda function.

The following procedure uses the createSnapshot event to automatically copy a completed snapshot
to another Region for disaster recovery.

To copy a completed snapshot to another Region

1. Create an IAM policy, such as the one shown in the following example, to provide permissions to use
the CopySnapshot action and write to the CloudWatch Events log. Assign the policy to the IAM user
that will handle the CloudWatch event.

{
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"logs:CreateLogGroup",
"logs:CreateLogStream",
"logs:PutLogEvents"
],
"Resource": "arn:aws:logs:*:*:*"
},
{
"Effect": "Allow",
"Action": [
"ec2:CopySnapshot"
],
"Resource": "*"
}
]
}

1266
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS CloudWatch events

2. Define a function in Lambda that will be available from the CloudWatch console. The sample
Lambda function below, written in Node.js, is invoked by CloudWatch when a matching
createSnapshot event is emitted by Amazon EBS (signifying that a snapshot was completed).
When invoked, the function copies the snapshot from us-east-2 to us-east-1.

// Sample Lambda function to copy an EBS snapshot to a different region

var AWS = require('aws-sdk');

var ec2 = new AWS.EC2();

// define variables
var destinationRegion = 'us-east-1';
var sourceRegion = 'us-east-2';
console.log ('Loading function');

//main function
exports.handler = (event, context, callback) => {

// Get the EBS snapshot ID from the CloudWatch event details

var snapshotArn = event.detail.snapshot_id.split('/');
const snapshotId = snapshotArn[1];
const description = `Snapshot copy from ${snapshotId} in ${sourceRegion}.`;
console.log ("snapshotId:", snapshotId);

// Load EC2 class and update the configuration to use destination Region to
initiate the snapshot.
AWS.config.update({region: destinationRegion});
var ec2 = new AWS.EC2();

// Prepare variables for ec2.modifySnapshotAttribute call

const copySnapshotParams = {
Description: description,
DestinationRegion: destinationRegion,
SourceRegion: sourceRegion,
SourceSnapshotId: snapshotId
};

// Execute the copy snapshot and log any errors

ec2.copySnapshot(copySnapshotParams, (err, data) => {
if (err) {
const errorMessage = `Error copying snapshot ${snapshotId} to Region
${destinationRegion}.`;
console.log(errorMessage);
console.log(err);
callback(errorMessage);
} else {
const successMessage = `Successfully started copy of snapshot ${snapshotId}
to Region ${destinationRegion}.`;
console.log(successMessage);
console.log(data);
callback(null, successMessage);
}
});
};

To ensure that your Lambda function is available from the CloudWatch console, create it in the
Region where the CloudWatch event will occur. For more information, see the AWS Lambda
Developer Guide.
3. Open the CloudWatch console at https://console.aws.amazon.com/cloudwatch/.
4. Choose Events, Create rule, Select event source, and Amazon EBS Snapshots.
5. For Specific Event(s), choose createSnapshot and for Specific Result(s), choose succeeded.
6. For Rule target, find and choose the sample function that you previously created.

1267
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EBS quotas

7. Choose Target, Add Target.

8. For Lambda function, select the Lambda function that you previously created and choose Configure
details.
9. On the Configure rule details page, type values for Name and Description. Select the State check
box to activate the function (setting it to Enabled).
10. Choose Create rule.

Your rule should now appear on the Rules tab. In the example shown, the event that you configured
should be emitted by EBS the next time you copy a snapshot.

Amazon EBS quotas

To view the quotas for your Amazon EBS resources, open the Service Quotas console at https://
console.aws.amazon.com/servicequotas/. In the navigation pane, choose AWS services, and select
Amazon Elastic Block Store (Amazon EBS).

For a list of Amazon EBS service quotas, see Amazon Elastic Block Store endpoints and quotas in the AWS
General Reference.

Amazon EC2 instance store

An instance store provides temporary block-level storage for your instance. This storage is located on
disks that are physically attached to the host computer. Instance store is ideal for temporary storage of
information that changes frequently, such as buffers, caches, scratch data, and other temporary content,
or for data that is replicated across a fleet of instances, such as a load-balanced pool of web servers.

An instance store consists of one or more instance store volumes exposed as block devices. The size of an
instance store as well as the number of devices available varies by instance type.

The virtual devices for instance store volumes are ephemeral[0-23]. Instance types that support one
instance store volume have ephemeral0. Instance types that support two instance store volumes have
ephemeral0 and ephemeral1, and so on.

1268
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance store lifetime

Contents
• Instance store lifetime (p. 1269)
• Instance store volumes (p. 1270)
• Add instance store volumes to your EC2 instance (p. 1277)
• SSD instance store volumes (p. 1280)
• Instance store swap volumes (p. 1282)
• Optimize disk performance for instance store volumes (p. 1284)

Instance store lifetime

You can specify instance store volumes for an instance only when you launch it. You can't detach an
instance store volume from one instance and attach it to a different instance.

The data in an instance store persists only during the lifetime of its associated instance. If an instance
reboots (intentionally or unintentionally), data in the instance store persists. However, data in the
instance store is lost under any of the following circumstances:

• The underlying disk drive fails

• The instance stops
• The instance hibernates
• The instance terminates

Therefore, do not rely on instance store for valuable, long-term data. Instead, use more durable data
storage, such as Amazon S3, Amazon EBS, or Amazon EFS.

When you stop, hibernate, or terminate an instance, every block of storage in the instance store is reset.
Therefore, your data cannot be accessed through the instance store of another instance.

1269
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance store volumes

If you create an AMI from an instance, the data on its instance store volumes isn't preserved and isn't
present on the instance store volumes of the instances that you launch from the AMI.

If you change the instance type, an instance store will not be attached to the new instance type. For
more information, see Change the instance type (p. 311).

Instance store volumes

The instance type determines the size of the instance store available and the type of hardware used for
the instance store volumes. Instance store volumes are included as part of the instance's usage cost.
You must specify the instance store volumes that you'd like to use when you launch the instance (except
for NVMe instance store volumes, which are available by default). Then format and mount the instance
store volumes before using them. You can't make an instance store volume available after you launch the
instance. For more information, see Add instance store volumes to your EC2 instance (p. 1277).

Some instance types use NVMe or SATA-based solid state drives (SSD) to deliver high random I/O
performance. This is a good option when you need storage with very low latency, but you don't need the
data to persist when the instance terminates or you can take advantage of fault-tolerant architectures.
For more information, see SSD instance store volumes (p. 1280).

The following table provides the quantity, size, type, and performance optimizations of instance store
volumes available on each supported instance type. For a complete list of instance types, including EBS-
only types, see Amazon EC2 Instance Types.

Instance type Instance store volumes Type Needs TRIM support**

initialization*

c1.medium 1 x 350 GB† HDD ✔  

c1.xlarge 4 x 420 GB (1.6 TB) HDD ✔  

c3.large 2 x 16 GB (32 GB) SSD ✔  

c3.xlarge 2 x 40 GB (80 GB) SSD ✔  

c3.2xlarge 2 x 80 GB (160 GB) SSD ✔  

c3.4xlarge 2 x 160 GB (320 GB) SSD ✔  

c3.8xlarge 2 x 320 GB (640 GB) SSD ✔  

c5ad.large 1 x 75 GB NVMe SSD   ✔

c5ad.xlarge 1 x 150 GB NVMe SSD   ✔

c5ad.2xlarge 1 x 300 GB NVMe SSD   ✔

c5ad.4xlarge 2 x 300 GB (600 GB) NVMe SSD   ✔

c5ad.8xlarge 2 x 600 GB (1.2 TB) NVMe SSD   ✔

c5ad.12xlarge 2 x 900 GB (1.8 TB) NVMe SSD   ✔

c5ad.16xlarge 2 x 1,200 GB (2.4 TB) NVMe SSD   ✔

c5ad.24xlarge 2 x 1,900 GB (3.8 TB) NVMe SSD   ✔

c5d.large 1 x 50 GB NVMe SSD   ✔

c5d.xlarge 1 x 100 GB NVMe SSD   ✔

1270
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance store volumes

Instance type Instance store volumes Type Needs TRIM support**

initialization*

c5d.2xlarge 1 x 200 GB NVMe SSD   ✔

c5d.4xlarge 1 x 400 GB NVMe SSD   ✔

c5d.9xlarge 1 x 900 GB NVMe SSD   ✔

c5d.12xlarge 2 x 900 GB (1.8 TB) NVMe SSD   ✔

c5d.18xlarge 2 x 900 GB (1.8 TB) NVMe SSD   ✔

c5d.24xlarge 4 x 900 GB (3.6 TB) NVMe SSD   ✔

c5d.metal 4 x 900 GB (3.6 TB) NVMe SSD   ✔

c6gd.medium 1 x 59 GB NVMe SSD   ✔

c6gd.large 1 x 118 GB NVMe SSD   ✔

c6gd.xlarge 1 x 237 GB NVMe SSD   ✔

c6gd.2xlarge 1 x 474 GB NVMe SSD   ✔

c6gd.4xlarge 1 x 950 GB NVMe SSD   ✔

c6gd.8xlarge 1 x 1,900 GB NVMe SSD   ✔

c6gd.12xlarge 2 x 1,425 GB (2.85 TB) NVMe SSD   ✔

c6gd.16xlarge 2 x 1,900 GB (3.8 TB) NVMe SSD   ✔

c6gd.metal 2 x 1,900 GB (3.8 TB) NVMe SSD   ✔

cc2.8xlarge 4 x 840 GB (3.36 TB) HDD ✔  

cr1.8xlarge 2 x 120 GB (240 GB) SSD ✔  

d2.xlarge 3 x 2,000 GB (6 TB) HDD    

d2.2xlarge 6 x 2,000 GB (12 TB) HDD    

d2.4xlarge 12 x 2,000 GB (24 TB) HDD    

d2.8xlarge 24 x 2,000 GB (48 TB) HDD    

d3.xlarge 3 x 1,980 GB HDD    

d3.2xlarge 6 x 1,980 GB HDD    

d3.4xlarge 12 x 1,980 GB HDD    

d3.8xlarge 24 x 1,980 GB HDD    

d3en.large 1 x 13,980 GB HDD    

d3en.xlarge 2 x 13,980 GB HDD    

d3en.2xlarge 4 x 13,980 GB HDD    

d3en.4xlarge 8 x 13,980 GB HDD    

1271
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance store volumes

Instance type Instance store volumes Type Needs TRIM support**

initialization*

d3en.6xlarge 12 x 13,980 GB HDD    

d3en.8xlarge 16 x 13,980 GB HDD    

d3en.12xlarge 24 x 13,980 GB HDD    

f1.2xlarge 1 x 470 GB NVMe SSD   ✔

f1.4xlarge 1 x 940 GB NVMe SSD   ✔

f1.16xlarge 4 x 940 GB (3.76 TB) NVMe SSD   ✔

g2.2xlarge 1 x 60 GB SSD ✔  

g2.8xlarge 2 x 120 GB (240 GB) SSD ✔  

g4ad.4xlarge 1 x 600 GB NVMe SSD   ✔

g4ad.8xlarge 1 x 1,200 GB NVMe SSD   ✔

g4ad.16xlarge 2 x 1,200 GB (2.4 TB) NVMe SSD   ✔

g4dn.xlarge 1 x 125 GB NVMe SSD   ✔

g4dn.2xlarge 1 x 225 GB NVMe SSD   ✔

g4dn.4xlarge 1 x 225 GB NVMe SSD   ✔

g4dn.8xlarge 1 x 900 GB NVMe SSD   ✔

g4dn.12xlarge 1 x 900 GB NVMe SSD   ✔

g4dn.16xlarge 1 x 900 GB NVMe SSD   ✔

g4dn.metal 2 x 900 GB (1.8 TB) NVMe SSD   ✔

h1.2xlarge 1 x 2,000 GB (2 TB) HDD    

h1.4xlarge 2 x 2,000 GB (4 TB) HDD    

h1.8xlarge 4 x 2,000 GB (8 TB) HDD    

h1.16xlarge 8 x 2,000 GB (16 TB) HDD    

hs1.8xlarge 24 x 2,000 GB (48 TB) HDD ✔  

i2.xlarge 1 x 800 GB SSD   ✔

i2.2xlarge 2 x 800 GB (1.6 TB) SSD   ✔

i2.4xlarge 4 x 800 GB (3.2 TB) SSD   ✔

i2.8xlarge 8 x 800 GB (6.4 TB) SSD   ✔

i3.large 1 x 475 GB NVMe SSD   ✔

i3.xlarge 1 x 950 GB NVMe SSD   ✔

i3.2xlarge 1 x 1,900 GB NVMe SSD   ✔

1272
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance store volumes

Instance type Instance store volumes Type Needs TRIM support**

initialization*

i3.4xlarge 2 x 1,900 GB (3.8 TB) NVMe SSD   ✔

i3.8xlarge 4 x 1,900 GB (7.6 TB) NVMe SSD   ✔

i3.16xlarge 8 x 1,900 GB (15.2 TB) NVMe SSD   ✔

i3.metal 8 x 1,900 GB (15.2 TB) NVMe SSD   ✔

i3en.large 1 x 1,250 GB NVMe SSD   ✔

i3en.xlarge 1 x 2,500 GB NVMe SSD   ✔

i3en.2xlarge 2 x 2,500 GB (5 TB) NVMe SSD   ✔

i3en.3xlarge 1 x 7,500 GB NVMe SSD   ✔

i3en.6xlarge 2 x 7,500 GB (15 TB) NVMe SSD   ✔

i3en.12xlarge 4 x 7,500 GB (30 TB) NVMe SSD   ✔

i3en.24xlarge 8 x 7,500 GB (60 TB) NVMe SSD   ✔

i3en.metal 8 x 7,500 GB (60 TB) NVMe SSD   ✔

m1.small 1 x 160 GB† HDD ✔  

m1.medium 1 x 410 GB HDD ✔  

m1.large 2 x 420 GB (840 GB) HDD ✔  

m1.xlarge 4 x 420 GB (1.6 TB) HDD ✔  

m2.xlarge 1 x 420 GB HDD ✔  

m2.2xlarge 1 x 850 GB HDD ✔  

m2.4xlarge 2 x 840 GB (1.68 TB) HDD ✔  

m3.medium 1 x 4 GB SSD ✔  

m3.large 1 x 32 GB SSD ✔  

m3.xlarge 2 x 40 GB (80 GB) SSD ✔  

m3.2xlarge 2 x 80 GB (160 GB) SSD ✔  

m5ad.large 1 x 75 GB NVMe SSD   ✔

m5ad.xlarge 1 x 150 GB NVMe SSD   ✔

m5ad.2xlarge 1 x 300 GB NVMe SSD   ✔

m5ad.4xlarge 2 x 300 GB (600 GB) NVMe SSD   ✔

m5ad.8xlarge 2 x 600 GB (1.2 TB) NVMe SSD   ✔

m5ad.12xlarge 2 x 900 GB (1.8 TB) NVMe SSD   ✔

m5ad.16xlarge 4 x 600 GB (2.4 TB) NVMe SSD   ✔

1273
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance store volumes

Instance type Instance store volumes Type Needs TRIM support**

initialization*

m5ad.24xlarge 4 x 900 GB (3.6 TB) NVMe SSD   ✔

m5d.large 1 x 75 GB NVMe SSD   ✔

m5d.xlarge 1 x 150 GB NVMe SSD   ✔

m5d.2xlarge 1 x 300 GB NVMe SSD   ✔

m5d.4xlarge 2 x 300 GB (600 GB) NVMe SSD   ✔

m5d.8xlarge 2 x 600 GB (1.2 TB) NVMe SSD   ✔

m5d.12xlarge 2 x 900 GB (1.8 TB) NVMe SSD   ✔

m5d.16xlarge 4 x 600 GB (2.4 TB) NVMe SSD   ✔

m5d.24xlarge 4 x 900 GB (3.6 TB) NVMe SSD   ✔

m5d.metal 4 x 900 GB (3.6 TB) NVMe SSD   ✔

m5dn.large 1 x 75 GB NVMe SSD   ✔

m5dn.xlarge 1 x 150 GB NVMe SSD   ✔

m5dn.2xlarge 1 x 300 GB NVMe SSD   ✔

m5dn.4xlarge 2 x 300 GB (600 GB) NVMe SSD   ✔

m5dn.8xlarge 2 x 600 GB (1.2 TB) NVMe SSD   ✔

m5dn.12xlarge 2 x 900 GB (1.8 TB) NVMe SSD   ✔

m5dn.16xlarge 4 x 600 GB (2.4 TB) NVMe SSD   ✔

m5dn.24xlarge 4 x 900 GB (3.6 TB) NVMe SSD   ✔

m6gd.medium 1 x 59 GB NVMe SSD   ✔

m6gd.large 1 x 118 GB NVMe SSD   ✔

m6gd.xlarge 1 x 237 GB NVMe SSD   ✔

m6gd.2xlarge 1 x 474 GB NVMe SSD   ✔

m6gd.4xlarge 1 x 950 GB NVMe SSD   ✔

m6gd.8xlarge 1 x 1,900 GB NVMe SSD   ✔

m6gd.12xlarge 2 x 1,425 GB (2.85 TB) NVMe SSD   ✔

m6gd.16xlarge 2 x 1,900 GB (3.8 TB) NVMe SSD   ✔

m6gd.metal 2 x 1,900 GB (3.8 TB) NVMe SSD   ✔

p3dn.24xlarge 2 x 900 GB (1.8 TB) NVMe SSD   ✔

p4d.24xlarge 8 x 1,000 GB (8 TB) NVMe SSD   ✔

r3.large 1 x 32 GB SSD   ✔

1274
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance store volumes

Instance type Instance store volumes Type Needs TRIM support**

initialization*

r3.xlarge 1 x 80 GB SSD   ✔

r3.2xlarge 1 x 160 GB SSD   ✔

r3.4xlarge 1 x 320 GB SSD   ✔

r3.8xlarge 2 x 320 GB (640 GB) SSD   ✔

r5ad.large 1 x 75 GB NVMe SSD   ✔

r5ad.xlarge 1 x 150 GB NVMe SSD   ✔

r5ad.2xlarge 1 x 300 GB NVMe SSD   ✔

r5ad.4xlarge 2 x 300 GB (600 GB) NVMe SSD   ✔

r5ad.8xlarge 2 x 600 GB (1.2 TB) NVMe SSD   ✔

r5ad.12xlarge 2 x 900 GB (1.8 TB) NVMe SSD   ✔

r5ad.16xlarge 4 x 600 GB (2.4 TB) NVMe SSD   ✔

r5ad.24xlarge 4 x 900 GB (3.6 TB) NVMe SSD   ✔

r5d.large 1 x 75 GB NVMe SSD   ✔

r5d.xlarge 1 x 150 GB NVMe SSD   ✔

r5d.2xlarge 1 x 300 GB NVMe SSD   ✔

r5d.4xlarge 2 x 300 GB (600 GB) NVMe SSD   ✔

r5d.8xlarge 2 x 600 GB (1.2 TB) NVMe SSD   ✔

r5d.12xlarge 2 x 900 GB (1.8 TB) NVMe SSD   ✔

r5d.16xlarge 4 x 600 GB (2.4 TB) NVMe SSD   ✔

r5d.24xlarge 4 x 900 GB (3.6 TB) NVMe SSD   ✔

r5d.metal 4 x 900 GB (3.6 TB) NVMe SSD   ✔

r5dn.large 1 x 75 GB NVMe SSD   ✔

r5dn.xlarge 1 x 150 GB NVMe SSD   ✔

r5dn.2xlarge 1 x 300 GB NVMe SSD   ✔

r5dn.4xlarge 2 x 300 GB (600 GB) NVMe SSD   ✔

r5dn.8xlarge 2 x 600 GB (1.2 TB) NVMe SSD   ✔

r5dn.12xlarge 2 x 900 GB (1.8 TB) NVMe SSD   ✔

r5dn.16xlarge 4 x 600 GB (2.4 TB) NVMe SSD   ✔

r5dn.24xlarge 4 x 900 GB (3.6 TB) NVMe SSD   ✔

r6gd.medium 1 x 59 GB NVMe SSD   ✔

1275
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance store volumes

Instance type Instance store volumes Type Needs TRIM support**

initialization*

r6gd.large 1 x 118 GB NVMe SSD   ✔

r6gd.xlarge 1 x 237 GB NVMe SSD   ✔

r6gd.2xlarge 1 x 474 GB NVMe SSD   ✔

r6gd.4xlarge 1 x 950 GB NVMe SSD   ✔

r6gd.8xlarge 1 x 1900 GB NVMe SSD   ✔

r6gd.12xlarge 2 x 1,425 GB (2.85 TB) NVMe SSD   ✔

r6gd.16xlarge 2 x 1,900 GB (3.8 TB) NVMe SSD   ✔

r6gd.metal 2 x 1,900 GB (3.8 TB) NVMe SSD   ✔

x1.16xlarge 1 x 1,920 GB SSD    

x1.32xlarge 2 x 1,920 GB (3.84 TB) SSD    

x1e.xlarge 1 x 120 GB SSD    

x1e.2xlarge 1 x 240 GB SSD    

x1e.4xlarge 1 x 480 GB SSD    

x1e.8xlarge 1 x 960 GB SSD    

x1e.16xlarge 1 x 1,920 GB SSD    

x1e.32xlarge 2 x 1,920 GB (3.84 TB) SSD    

z1d.large 1 x 75 GB NVMe SSD   ✔

z1d.xlarge 1 x 150 GB NVMe SSD   ✔

z1d.2xlarge 1 x 300 GB NVMe SSD   ✔

z1d.3xlarge 1 x 450 GB NVMe SSD   ✔

z1d.6xlarge 1 x 900 GB NVMe SSD   ✔

z1d.12xlarge 2 x 900 GB (1.8 TB) NVMe SSD   ✔

z1d.metal 2 x 900 GB (1.8 TB) NVMe SSD   ✔

* Volumes attached to certain instances suffer a first-write penalty unless initialized. For more
information, see Optimize disk performance for instance store volumes (p. 1284).

** For more information, see Instance store volume TRIM support (p. 1282).

† The c1.medium and m1.small instance types also include a 900 MB instance store swap volume,
which may not be automatically enabled at boot time. For more information, see Instance store swap
volumes (p. 1282).

1276
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Add instance store volumes

Add instance store volumes to your EC2 instance

You specify the EBS volumes and instance store volumes for your instance using a block device mapping.
Each entry in a block device mapping includes a device name and the volume that it maps to. The default
block device mapping is specified by the AMI you use. Alternatively, you can specify a block device
mapping for the instance when you launch it.

All the NVMe instance store volumes supported by an instance type are automatically enumerated and
assigned a device name on instance launch; including them in the block device mapping for the AMI or
the instance has no effect. For more information, see Block device mapping (p. 1294).

A block device mapping always specifies the root volume for the instance. The root volume is either
an Amazon EBS volume or an instance store volume. For more information, see Storage for the root
device (p. 102). The root volume is mounted automatically. For instances with an instance store volume
for the root volume, the size of this volume varies by AMI, but the maximum size is 10 GB.

You can use a block device mapping to specify additional EBS volumes when you launch your instance, or
you can attach additional EBS volumes after your instance is running. For more information, see Amazon
EBS volumes (p. 1082).

You can specify the instance store volumes for your instance only when you launch it. You can't attach
instance store volumes to an instance after you've launched it.

If you change the instance type, an instance store will not be attached to the new instance type. For
more information, see Change the instance type (p. 311).

The number and size of available instance store volumes for your instance varies by instance type. Some
instance types do not support instance store volumes. If the number of instance store volumes in a block
device mapping exceeds the number of instance store volumes available to an instance, the additional
volumes are ignored. For more information about the instance store volumes supported by each instance
type, see Instance store volumes (p. 1270).

If the instance type you choose for your instance supports non-NVMe instance store volumes, you must
add them to the block device mapping for the instance when you launch it. NVMe instance store volumes
are available by default. After you launch an instance, you must ensure that the instance store volumes
for your instance are formatted and mounted before you can use them. The root volume of an instance
store-backed instance is mounted automatically.

Contents
• Add instance store volumes to an AMI (p. 1277)
• Add instance store volumes to an instance (p. 1278)
• Make instance store volumes available on your instance (p. 1279)

Add instance store volumes to an AMI

You can create an AMI with a block device mapping that includes instance store volumes. If you launch
an instance with an instance type that supports instance store volumes and an AMI that specifies
instance store volumes in its block device mapping, the instance includes these instance store volumes. If
the number of instance store volumes in the block device mapping exceeds the number of instance store
volumes available to the instance, the additional instance store volumes are ignored.

Considerations

• For M3 instances, specify instance store volumes in the block device mapping of the instance, not
the AMI. Amazon EC2 might ignore instance store volumes that are specified only in the block device
mapping of the AMI.

1277
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Add instance store volumes

• When you launch an instance, you can omit non-NVMe instance store volumes specified in the AMI
block device mapping or add instance store volumes.

New console

To add instance store volumes to an Amazon EBS-backed AMI using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and select the instance.
3. Choose Actions, Image and templates, Create image.
4. On the Create image page, enter a meaningful name and description for your image.
5. For each instance store volume to add, choose Add volume, from Volume type select an
instance store volume, and from Device select a device name. (For more information, see Name
devices on Linux instances (p. 1292).) The number of available instance store volumes depends
on the instance type. For instances with NVMe instance store volumes, the device mapping of
these volumes depends on the order in which the operating system enumerates the volumes.
6. Choose Create image.

Old console

To add instance store volumes to an Amazon EBS-backed AMI using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and select the instance.
3. Choose Actions, Image, Create Image.
4. In the Create Image dialog box, type a meaningful name and description for your image.
5. For each instance store volume to add, choose Add New Volume, from Volume Type select an
instance store volume, and from Device select a device name. (For more information, see Name
devices on Linux instances (p. 1292).) The number of available instance store volumes depends
on the instance type. For instances with NVMe instance store volumes, the device mapping of
these volumes depends on the order in which the operating system enumerates the volumes.
6. Choose Create Image.

To add instance store volumes to an AMI using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• create-image or register-image (AWS CLI)

• New-EC2Image and Register-EC2Image (AWS Tools for Windows PowerShell)

Add instance store volumes to an instance

When you launch an instance, the default block device mapping is provided by the specified AMI. If you
need additional instance store volumes, you must add them to the instance as you launch it. You can also
omit devices specified in the AMI block device mapping.

Considerations

• For M3 instances, you might receive instance store volumes even if you do not specify them in the
block device mapping for the instance.

1278
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Add instance store volumes

• For HS1 instances, no matter how many instance store volumes you specify in the block device
mapping of an AMI, the block device mapping for an instance launched from the AMI automatically
includes the maximum number of supported instance store volumes. You must explicitly remove the
instance store volumes that you don't want from the block device mapping for the instance before you
launch it.

To update the block device mapping for an instance using the console

1. Open the Amazon EC2 console.

2. From the dashboard, choose Launch instance.
3. In Step 1: Choose an Amazon Machine Image (AMI), select the AMI to use and choose Select.
4. Follow the wizard to complete Step 1: Choose an Amazon Machine Image (AMI), Step 2: Choose an
Instance Type, and Step 3: Configure Instance Details.
5. In Step 4: Add Storage, modify the existing entries as needed. For each instance store volume to
add, choose Add New Volume, from Volume Type select an instance store volume, and from Device
select a device name. The number of available instance store volumes depends on the instance type.
6. Complete the wizard and launch the instance.
7. (Optional) To view the instance store volumes available on your instance, run the lsblk command.

To update the block device mapping for an instance using the command line

You can use one of the following options commands with the corresponding command. For more
information about these command line interfaces, see Access Amazon EC2 (p. 3).

• --block-device-mappings with run-instances (AWS CLI)

• -BlockDeviceMapping with New-EC2Instance (AWS Tools for Windows PowerShell)

Make instance store volumes available on your instance

After you launch an instance, the instance store volumes are available to the instance, but you can't
access them until they are mounted. For Linux instances, the instance type determines which instance
store volumes are mounted for you and which are available for you to mount yourself. For Windows
instances, the EC2Config service mounts the instance store volumes for an instance. The block device
driver for the instance assigns the actual volume name when mounting the volume, and the name
assigned can be different than the name that Amazon EC2 recommends.

Many instance store volumes are pre-formatted with the ext3 file system. SSD-based instance store
volumes that support TRIM instruction are not pre-formatted with any file system. However, you can
format volumes with the file system of your choice after you launch your instance. For more information,
see Instance store volume TRIM support (p. 1282). For Windows instances, the EC2Config service
reformats the instance store volumes with the NTFS file system.

You can confirm that the instance store devices are available from within the instance itself using
instance metadata. For more information, see View the instance block device mapping for instance store
volumes (p. 1302).

For Windows instances, you can also view the instance store volumes using Windows Disk Management.
For more information, see Listing disks using Windows Disk Management.

For Linux instances, you can view and mount the instance store volumes as described in the following
procedure.

1279
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
SSD instance store volumes

To make an instance store volume available on Linux

1. Connect to the instance using an SSH client. For more information, see Connect to your Linux
instance (p. 598).
2. Use the df -h command to view the volumes that are formatted and mounted.

[ec2-user ~]$ df -h
Filesystem Size Used Avail Use% Mounted on
devtmpfs 3.8G 72K 3.8G 1% /dev
tmpfs 3.8G 0 3.8G 0% /dev/shm
/dev/nvme0n1p1 7.9G 1.2G 6.6G 15% /

3. Use the lsblk to view any volumes that were mapped at launch but not formatted and mounted.

[ec2-user ~]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
nvme0n1 259:1 0 8G 0 disk
##nvme0n1p1 259:2 0 8G 0 part /
##nvme0n1p128 259:3 0 1M 0 part
nvme1n1 259:0 0 69.9G 0 disk

4. To format and mount an instance store volume that was mapped only, do the following:

a. Create a file system on the device using the mkfs command.

[ec2-user ~]$ sudo mkfs -t xfs /dev/nvme1n1

b. Create a directory on which to mount the device using the mkdir command.

[ec2-user ~]$ sudo mkdir /data

c. Mount the device on the newly created directory using the mount command.

[ec2-user ~]$ sudo mount /dev/nvme1n1 /data

For instructions on how to mount an attached volume automatically after reboot, see Automatically
mount an attached volume after reboot (p. 1113).

SSD instance store volumes

To ensure the best IOPS performance from your SSD instance store volumes on Linux, we recommend
that you use the most recent version of Amazon Linux, or another Linux AMI with a kernel version of 3.8
or later. If you do not use a Linux AMI with a kernel version of 3.8 or later, your instance won't achieve
the maximum IOPS performance available for these instance types.

Like other instance store volumes, you must map the SSD instance store volumes for your instance when
you launch it. The data on an SSD instance volume persists only for the life of its associated instance. For
more information, see Add instance store volumes to your EC2 instance (p. 1277).

NVMe SSD volumes

Some instances offer non-volatile memory express (NVMe) solid state drives (SSD) instance store
volumes. For more information about the type of instance store volume supported by each instance
type, see Instance store volumes (p. 1270).

To access NVMe volumes, the NVMe drivers (p. 1214) must be installed. The following AMIs meet this
requirement:

1280
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
SSD instance store volumes

• Amazon Linux 2
• Amazon Linux AMI 2018.03
• Ubuntu 14.04 (with linux-aws kernel) or later
• Red Hat Enterprise Linux 7.4 or later
• SUSE Linux Enterprise Server 12 SP2 or later
• CentOS 7.4.1708 or later
• FreeBSD 11.1 or later
• Debian GNU/Linux 9 or later

After you connect to your instance, you can list the NVMe devices using the lspci command. The
following is example output for an i3.8xlarge instance, which supports four NVMe devices.

[ec2-user ~]$ lspci

00:00.0 Host bridge: Intel Corporation 440FX - 82441FX PMC [Natoma] (rev 02)
00:01.0 ISA bridge: Intel Corporation 82371SB PIIX3 ISA [Natoma/Triton II]
00:01.1 IDE interface: Intel Corporation 82371SB PIIX3 IDE [Natoma/Triton II]
00:01.3 Bridge: Intel Corporation 82371AB/EB/MB PIIX4 ACPI (rev 01)
00:02.0 VGA compatible controller: Cirrus Logic GD 5446
00:03.0 Ethernet controller: Device 1d0f:ec20
00:17.0 Non-Volatile memory controller: Device 1d0f:cd01
00:18.0 Non-Volatile memory controller: Device 1d0f:cd01
00:19.0 Non-Volatile memory controller: Device 1d0f:cd01
00:1a.0 Non-Volatile memory controller: Device 1d0f:cd01
00:1f.0 Unassigned class [ff80]: XenSource, Inc. Xen Platform Device (rev 01)

If you are using a supported operating system but you do not see the NVMe devices, verify that the
NVMe module is loaded using the following command.

• Amazon Linux, Amazon Linux 2, Ubuntu 14/16, Red Hat Enterprise Linux, SUSE Linux Enterprise
Server, CentOS 7

$ lsmod | grep nvme

nvme 48813 0

• Ubuntu 18

$ cat /lib/modules/$(uname -r)/modules.builtin | grep nvme

s/nvme/host/nvme-core.ko
kernel/drivers/nvme/host/nvme.ko
kernel/drivers/nvmem/nvmem_core.ko

The NVMe volumes are compliant with the NVMe 1.0e specification. You can use the NVMe commands
with your NVMe volumes. With Amazon Linux, you can install the nvme-cli package from the repo
using the yum install command. With other supported versions of Linux, you can download the nvme-
cli package if it's not available in the image.

The data on NVMe instance storage is encrypted using an XTS-AES-256 block cipher implemented in a
hardware module on the instance. The encryption keys are generated using the hardware module and
are unique to each NVMe instance storage device. All encryption keys are destroyed when the instance
is stopped or terminated and cannot be recovered. You cannot disable this encryption and you cannot
provide your own encryption key.

1281
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance store swap volumes

Non-NVMe SSD volumes

The following instances support instance store volumes that use non-NVMe SSDs to deliver high random
I/O performance: C3, G2, I2, M3, R3, and X1. For more information about the instance store volumes
supported by each instance type, see Instance store volumes (p. 1270).

Instance store volume TRIM support

Some instance types support SSD volumes with TRIM. For more information, see Instance store
volumes (p. 1270).

Instance store volumes that support TRIM are fully trimmed before they are allocated to your instance.
These volumes are not formatted with a file system when an instance launches, so you must format
them before they can be mounted and used. For faster access to these volumes, you should skip the
TRIM operation when you format them.

With instance store volumes that support TRIM, you can use the TRIM command to notify the SSD
controller when you no longer need data that you've written. This provides the controller with more
free space, which can reduce write amplification and increase performance. On Linux, use the fstrim
command to enable periodic TRIM.

Instance store swap volumes

Swap space in Linux can be used when a system requires more memory than it has been physically
allocated. When swap space is enabled, Linux systems can swap infrequently used memory pages from
physical memory to swap space (either a dedicated partition or a swap file in an existing file system) and
free up that space for memory pages that require high-speed access.
Note
Using swap space for memory paging is not as fast or efficient as using RAM. If your workload
is regularly paging memory into swap space, you should consider migrating to a larger instance
type with more RAM. For more information, see Change the instance type (p. 311).

The c1.medium and m1.small instance types have a limited amount of physical memory to work with,
and they are given a 900 MiB swap volume at launch time to act as virtual memory for Linux AMIs.
Although the Linux kernel sees this swap space as a partition on the root device, it is actually a separate
instance store volume, regardless of your root device type.

Amazon Linux automatically enables and uses this swap space, but your AMI may require some
additional steps to recognize and use this swap space. To see if your instance is using swap space, you
can use the swapon -s command.

[ec2-user ~]$ swapon -s

Filename Type Size Used Priority
/dev/xvda3 partition 917500 0 -1

The above instance has a 900 MiB swap volume attached and enabled. If you don't see a swap volume
listed with this command, you may need to enable swap space for the device. Check your available disks
using the lsblk command.

[ec2-user ~]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
xvda1 202:1 0 8G 0 disk /
xvda3 202:3 0 896M 0 disk

Here, the swap volume xvda3 is available to the instance, but it is not enabled (notice that the
MOUNTPOINT field is empty). You can enable the swap volume with the swapon command.

1282
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance store swap volumes

Note
You must prepend /dev/ to the device name listed by lsblk. Your device may be named
differently, such as sda3, sde3, or xvde3. Use the device name for your system in the command
below.

[ec2-user ~]$ sudo swapon /dev/xvda3

Now the swap space should show up in lsblk and swapon -s output.

[ec2-user ~]$ lsblk

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
xvda1 202:1 0 8G 0 disk /
xvda3 202:3 0 896M 0 disk [SWAP]
[ec2-user ~]$ swapon -s
Filename Type Size Used Priority
/dev/xvda3 partition 917500 0 -1

You also need to edit your /etc/fstab file so that this swap space is automatically enabled at every
system boot.

[ec2-user ~]$ sudo vim /etc/fstab

Append the following line to your /etc/fstab file (using the swap device name for your system):

/dev/xvda3 none swap sw 0 0

To use an instance store volume as swap space

Any instance store volume can be used as swap space. For example, the m3.medium instance type
includes a 4 GB SSD instance store volume that is appropriate for swap space. If your instance store
volume is much larger (for example, 350 GB), you may consider partitioning the volume with a smaller
swap partition of 4-8 GB and the rest for a data volume.
Note
This procedure applies only to instance types that support instance storage. For a list of
supported instance types, see Instance store volumes (p. 1270).

1. List the block devices attached to your instance to get the device name for your instance store
volume.

[ec2-user ~]$ lsblk -p

NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
/dev/xvdb 202:16 0 4G 0 disk /media/ephemeral0
/dev/xvda1 202:1 0 8G 0 disk /

In this example, the instance store volume is /dev/xdvb. Because this is an Amazon Linux instance,
the instance store volume is formatted and mounted at /media/ephemeral0; not all Linux
operating systems do this automatically.
2. (Optional) If your instance store volume is mounted (it lists a MOUNTPOINT in the lsblk command
output), unmount it with the following command.

[ec2-user ~]$ sudo umount /dev/xvdb

3. Set up a Linux swap area on the device with the mkswap command.

[ec2-user ~]$ sudo mkswap /dev/xvdb

1283
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Optimize disk performance

mkswap: /dev/xvdb: warning: wiping old ext3 signature.

Setting up swapspace version 1, size = 4188668 KiB
no label, UUID=b4f63d28-67ed-46f0-b5e5-6928319e620b

4. Enable the new swap space.

[ec2-user ~]$ sudo swapon /dev/xvdb

5. Verify that the new swap space is being used.

[ec2-user ~]$ swapon -s

Filename Type Size Used Priority
/dev/xvdb partition 4188668 0 -1

6. Edit your /etc/fstab file so that this swap space is automatically enabled at every system boot.

[ec2-user ~]$ sudo vim /etc/fstab

If your /etc/fstab file has an entry for /dev/xvdb (or /dev/sdb) change it to match the line
below; if it does not have an entry for this device, append the following line to your /etc/fstab
file (using the swap device name for your system):

/dev/xvdb none swap sw 0 0

Important
Instance store volume data is lost when an instance is stopped or hibernated; this includes
the instance store swap space formatting created in Step 3 (p. 1283). If you stop and restart
an instance that has been configured to use instance store swap space, you must repeat
Step 1 (p. 1283) through Step 5 (p. 1284) on the new instance store volume.

Optimize disk performance for instance store

volumes
Because of the way that Amazon EC2 virtualizes disks, the first write to any location on some instance
store volumes performs more slowly than subsequent writes. For most applications, amortizing this
cost over the lifetime of the instance is acceptable. However, if you require high disk performance, we
recommend that you initialize your drives by writing once to every drive location before production use.
Note
Some instance types with direct-attached solid state drives (SSD) and TRIM support provide
maximum performance at launch time, without initialization. For information about the instance
store for each instance type, see Instance store volumes (p. 1270).

If you require greater flexibility in latency or throughput, we recommend using Amazon EBS.

To initialize the instance store volumes, use the following dd commands, depending on the store to
initialize (for example, /dev/sdb or /dev/nvme1n1).
Note
Make sure to unmount the drive before performing this command.
Initialization can take a long time (about 8 hours for an extra large instance).

To initialize the instance store volumes, use the following commands on the m1.large, m1.xlarge,
c1.xlarge, m2.xlarge, m2.2xlarge, and m2.4xlarge instance types:

dd if=/dev/zero of=/dev/sdb bs=1M

1284
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
File storage

dd if=/dev/zero of=/dev/sdc bs=1M

dd if=/dev/zero of=/dev/sdd bs=1M
dd if=/dev/zero of=/dev/sde bs=1M

To perform initialization on all instance store volumes at the same time, use the following command:

dd if=/dev/zero bs=1M|tee /dev/sdb|tee /dev/sdc|tee /dev/sde > /dev/sdd

Configuring drives for RAID initializes them by writing to every drive location. When configuring
software-based RAID, make sure to change the minimum reconstruction speed:

echo $((30*1024)) > /proc/sys/dev/raid/speed_limit_min

File storage
Cloud file storage is a method for storing data in the cloud that provides servers and applications access
to data through shared file systems. This compatibility makes cloud file storage ideal for workloads that
rely on shared file systems and provides simple integration without code changes.

There are many file storage solutions that exist, ranging from a single node file server on a compute
instance using block storage as the underpinnings with no scalability or few redundancies to protect
the data, to a do-it-yourself clustered solution, to a fully-managed solution. The following content
introduces some of the storage services provided by AWS for use with Linux.

Contents
• Use Amazon S3 with Amazon EC2 (p. 1285)
• Use Amazon EFS with Amazon EC2 (p. 1287)

Use Amazon S3 with Amazon EC2

Amazon S3 is a repository for internet data. Amazon S3 provides access to reliable, fast, and inexpensive
data storage infrastructure. It is designed to make web-scale computing easier by enabling you to store
and retrieve any amount of data, at any time, from within Amazon EC2 or anywhere on the web. Amazon
S3 stores data objects redundantly on multiple devices across multiple facilities and allows concurrent
read or write access to these data objects by many separate clients or application threads. You can use
the redundant data stored in Amazon S3 to recover quickly and reliably from instance or application
failures.

Amazon EC2 uses Amazon S3 for storing Amazon Machine Images (AMIs). You use AMIs for launching
EC2 instances. In case of instance failure, you can use the stored AMI to immediately launch another
instance, thereby allowing for fast recovery and business continuity.

Amazon EC2 also uses Amazon S3 to store snapshots (backup copies) of the data volumes. You can use
snapshots for recovering data quickly and reliably in case of application or system failures. You can
also use snapshots as a baseline to create multiple new data volumes, expand the size of an existing
data volume, or move data volumes across multiple Availability Zones, thereby making your data usage
highly scalable. For more information about using data volumes and snapshots, see Amazon Elastic Block
Store (p. 1080).

Objects are the fundamental entities stored in Amazon S3. Every object stored in Amazon S3 is
contained in a bucket. Buckets organize the Amazon S3 namespace at the highest level and identify
the account responsible for that storage. Amazon S3 buckets are similar to internet domain names.

1285
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Amazon S3

Objects stored in the buckets have a unique key value and are retrieved using a URL. For example, if an
object with a key value /photos/mygarden.jpg is stored in the DOC-EXAMPLE-BUCKET1 bucket, then
it is addressable using the URL https://DOC-EXAMPLE-BUCKET1.s3.amazonaws.com/photos/
mygarden.jpg.

For more information about the features of Amazon S3, see the Amazon S3 product page.

Usage examples
Given the benefits of Amazon S3 for storage, you might decide to use this service to store files and data
sets for use with EC2 instances. There are several ways to move data to and from Amazon S3 to your
instances. In addition to the examples discussed below, there are a variety of tools that people have
written that you can use to access your data in Amazon S3 from your computer or your instance. Some of
the common ones are discussed in the AWS forums.

If you have permission, you can copy a file to or from Amazon S3 and your instance using one of the
following methods.

GET or wget

The wget utility is an HTTP and FTP client that allows you to download public objects from Amazon S3.
It is installed by default in Amazon Linux and most other distributions, and available for download on
Windows. To download an Amazon S3 object, use the following command, substituting the URL of the
object to download.

[ec2-user ~]$ wget https://my_bucket.s3.amazonaws.com/path-to-file

This method requires that the object you request is public; if the object is not public, you receive an
"ERROR 403: Forbidden" message. If you receive this error, open the Amazon S3 console and change
the permissions of the object to public. For more information, see the Amazon Simple Storage Service
Developer Guide.

AWS Command Line Interface

The AWS Command Line Interface (AWS CLI) is a unified tool to manage your AWS services. The AWS
CLI enables users to authenticate themselves and download restricted items from Amazon S3 and also
to upload items. For more information, such as how to install and configure the tools, see the AWS
Command Line Interface detail page.

The aws s3 cp command is similar to the Unix cp command. You can copy files from Amazon S3 to your
instance, copy files from your instance to Amazon S3, and copy files from one Amazon S3 location to
another.

Use the following command to copy an object from Amazon S3 to your instance.

[ec2-user ~]$ aws s3 cp s3://my_bucket/my_folder/my_file.ext my_copied_file.ext

Use the following command to copy an object from your instance back into Amazon S3.

[ec2-user ~]$ aws s3 cp my_copied_file.ext s3://my_bucket/my_folder/my_file.ext

The aws s3 sync command can synchronize an entire Amazon S3 bucket to a local directory location. This
can be helpful for downloading a data set and keeping the local copy up-to-date with the remote set. If
you have the proper permissions on the Amazon S3 bucket, you can push your local directory back up to
the cloud when you are finished by reversing the source and destination locations in the command.

1286
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Amazon EFS

Use the following command to download an entire Amazon S3 bucket to a local directory on your
instance.

[ec2-user ~]$ aws s3 sync s3://remote_S3_bucket local_directory

Amazon S3 API

If you are a developer, you can use an API to access data in Amazon S3. For more information, see the
Amazon Simple Storage Service Developer Guide. You can use this API and its examples to help develop
your application and integrate it with other APIs and SDKs, such as the boto Python interface.

Use Amazon EFS with Amazon EC2

Amazon EFS provides scalable file storage for use with Amazon EC2. You can use an EFS file system
as a common data source for workloads and applications running on multiple instances. For more
information, see the Amazon Elastic File System product page.
Important
Amazon EFS is not supported on Windows instances.

You can mount an EFS file system to your instance in the following ways:

Topics
• Create an EFS file system using Amazon EFS Quick Create (p. 1287)
• Create an EFS file system and mount it to your instance (p. 1288)

Create an EFS file system using Amazon EFS Quick Create

You can create an EFS file system and mount it to your instance at the time of launch using the Amazon
EFS Quick Create feature of the Instance Launch Wizard.

When you create an EFS file system using EFS Quick Create, the file system is created with the following
service recommended settings:

• Automatic backups turned on. For more information, see Using AWS Backup with Amazon EFS in the
Amazon Elastic File System User Guide.
• Mount targets in each default subnet in the selected VPC, using the VPC's default security group. For
more information, see Managing file system network accessibility in the Amazon Elastic File System
User Guide.
• General Purpose performance mode. For more information, see Performance Modes in the Amazon
Elastic File System User Guide.
• Bursting throughput mode. For more information, see Throughput Modes in the Amazon Elastic File
System User Guide.
• Encryption of data at rest enabled using your default key for Amazon EFS (aws/
elasticfilesystem). For more information, see Encrypting Data at Rest in the Amazon Elastic File
System User Guide.
• Amazon EFS lifecycle management enabled with a 30-day policy. For more information, see EFS
lifecycle management in the Amazon Elastic File System User Guide.

To create an EFS file system using Amazon EFS Quick Create

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

1287
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Amazon EFS

2. Choose Launch Instance.

3. On the Choose an AMI page, choose a Linux AMI.
4. On the Choose an Instance Type page, select an instance type and then choose Next: Configure
Instance Details.
5. On the Configure Instance Details page, for File systems, choose Create new file system, enter a
name for the new file system, and then choose Create.

To enable access to the file system, the following security groups are automatically created and
attached to the instance and the mount targets of the file system.

• Instance security group—Includes no inbound rules and an outbound rule that allows traffic over
the NFS 2049 port.
• File system mount targets security group—Includes an inbound rule that allows traffic over the
NFS 2049 port from the instance security group (described above), and an outbound rule that
allows traffic over the NFS 2049 port.

You can also choose to manually create and attach the security groups. To do this, clear
Automatically create and attach the required security groups.

Configure the remaining settings as needed and choose Next: Add Storage.
6. On the Add Storage page, specify the volumes to attach to the instances, in addition to the volumes
specified by the AMI (such as the root device volume). Ensure that you provision enough storage for
the Nvidia CUDA Toolkit. Then choose Next: Add Tags.
7. On the Add Tags page, specify a tag that you can use to identify the temporary instance, and then
choose Next: Configure Security Group.
8. On the Configure Security Group page, review the security groups and then choose Review and
Launch.
9. On the Review Instance Launch page, review the settings, and then choose Launch to choose a key
pair and to launch your instance.

Create an EFS file system and mount it to your instance

In this tutorial, you create an EFS file system and two Linux instances that can share data using the file
system.

Tasks
• Prerequisites (p. 1288)
• Step 1: Create an EFS file system (p. 1289)
• Step 2: Mount the file system (p. 1289)
• Step 3: Test the file system (p. 1290)
• Step 4: Clean up (p. 1291)

Prerequisites
• Create a security group (for example, efs-sg) to associate with the EC2 instances and EFS mount target,
and add the following rules:
• Allow inbound SSH connections to the EC2 instances from your computer (the source is the CIDR
block for your network).
• Allow inbound NFS connections to the file system via the EFS mount target from the EC2 instances
that are associated with this security group (the source is the security group itself). For more

1288
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Amazon EFS

information, see Amazon EFS rules (p. 1076), and Creating security Groups in the Amazon Elastic File
System User Guide.
• Create a key pair. You must specify a key pair when you configure your instances or you can't connect
to them. For more information, see Create a key pair (p. 28).

Step 1: Create an EFS file system

Amazon EFS enables you to create a file system that multiple instances can mount and access at the
same time. For more information, see Creating Resources for Amazon EFS in the Amazon Elastic File
System User Guide.

To create a file system

1. Open the Amazon Elastic File System console at https://console.aws.amazon.com/efs/.

2. Choose Create file system.
3. (Optional) For Name, enter a name for the file system. This creates a tag with Name as the key and
the name of the file system as the value.
4. For Virtual Private Cloud (VPC), select the VPC to use for your instances.
5. Choose Create.
6. After the file system is created, note the file system ID. It is used later in this tutorial.
7. Choose the file system ID.
8. On the file systems page, choose Network, Manage. View the mount targets that Amazon EFS
creates in each Availability Zone in the Region in which your VPC resides. For each Availability Zone
for your instances, ensure that the value for Security groups is the security group that you created
in Prerequisites (p. 1288).
9. Choose Save.

Step 2: Mount the file system

Use the following procedure to launch two t2.micro instances. Note that T2 instances must be
launched in a subnet. You can use a default VPC or a nondefault VPC.
Note
There are other ways that you can mount the volume (for example, on an already running
instance). For more information, see Mounting File Systems in the Amazon Elastic File System
User Guide.

To launch two instances and mount an EFS file system

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. Choose Launch Instance.
3. For Step 1: Choose an Amazon Machine Image (AMI), select an Amazon Linux AMI.
4. For Step 2: Choose an Instance Type, keep the default instance type, t2.micro, and choose Next:
Configure Instance Details.
5. For Step 3: Configure Instance Details, do the following:

a. For Number of instances, enter 2.

b. [Default VPC] If you have a default VPC, it is the default value for Network. Keep the default
VPC and the default value for Subnet to use the default subnet in the Availability Zone that
Amazon EC2 chooses for your instances.

[Nondefault VPC] Select your VPC for Network, and a public subnet from Subnet.

1289
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Amazon EFS

c. [Nondefault VPC] For Auto-assign Public IP, choose Enable. Otherwise, your instances do not
get public IP addresses or public DNS names.
d. For File systems, choose Add file system. Ensure that the value matches the file system ID
that you created in Step 1: Create an EFS file system (p. 1289). The path shown next to the file
system ID is the mount point that the instance will use, which you can change. Under Advanced
Details, the User data is automatically generated, and includes the commands needed to mount
the file system.
e. Advance to Step 6 of the wizard.
6. On the Configure Security Group page, choose Select an existing security group and select the
security group that you created in Prerequisites (p. 1288). Then choose Review and Launch.
7. On the Review Instance Launch page, choose Launch.
8. In the Select an existing key pair or create a new key pair dialog box, select Choose an existing
key pair and choose your key pair. Select the acknowledgment check box, and choose Launch
Instances.
9. In the navigation pane, choose Instances to see the status of your instances. Initially, their status is
pending. After the status changes to running, your instances are ready for use.

Your instance is now configured to mount the Amazon EFS file system at launch and whenever it's
rebooted.

Step 3: Test the file system

You can connect to your instances and verify that the file system is mounted to the directory that you
specified (for example, /mnt/efs).

To verify that the file system is mounted

1. Connect to your instances. For more information, see Connect to your Linux instance (p. 598).
2. From the terminal window for each instance, run the df -T command to verify that the EFS file
system is mounted.

$ df -T
Filesystem Type 1K-blocks Used Available Use% Mounted on
/dev/xvda1 ext4 8123812 1949800 6073764 25% /
devtmpfs devtmpfs 4078468 56 4078412 1% /dev
tmpfs tmpfs 4089312 0 4089312 0% /dev/shm
efs-dns nfs4 9007199254740992 0 9007199254740992 0% /mnt/efs

Note that the name of the file system, shown in the example output as efs-dns, has the following
form.

file-system-id.efs.aws-region.amazonaws.com:/

3. (Optional) Create a file in the file system from one instance, and then verify that you can view the
file from the other instance.

a. From the first instance, run the following command to create the file.

$ sudo touch /mnt/efs/test-file.txt

b. From the second instance, run the following command to view the file.

$ ls /mnt/efs
test-file.txt

1290
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance volume limits

Step 4: Clean up
When you are finished with this tutorial, you can terminate the instances and delete the file system.

To terminate the instances

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances.
3. Select the instances to terminate.
4. Choose Instance state, Terminate instance.
5. Choose Terminate when prompted for confirmation.

To delete the file system

1. Open the Amazon Elastic File System console at https://console.aws.amazon.com/efs/.

2. Select the file system to delete.
3. Choose Actions, Delete file system.
4. When prompted for confirmation, enter the file system ID and choose Delete file system.

Instance volume limits

The maximum number of volumes that your instance can have depends on the operating system and
instance type. When considering how many volumes to add to your instance, you should consider
whether you need increased I/O bandwidth or increased storage capacity.

Contents
• Nitro System volume limits (p. 1291)
• Linux-specific volume limits (p. 1292)
• Bandwidth versus capacity (p. 1292)

Nitro System volume limits

Instances built on the Nitro System (p. 209) support a maximum number of attachments, which are
shared between network interfaces, EBS volumes, and NVMe instance store volumes. Every instance has
at least one network interface attachment. NVMe instance store volumes are automatically attached. For
more information, see Elastic network interfaces (p. 839) and Instance store volumes (p. 1270).

Most of these instances support a maximum of 28 attachments. For example, if you have no additional
network interface attachments on an EBS-only instance, you can attach up to 27 EBS volumes to it. If
you have one additional network interface on an instance with 2 NVMe instance store volumes, you can
attach 24 EBS volumes to it.

For other instances, the following limits apply:

• d3.8xlarge and d3en.12xlarge instances support a maximum of 3 EBS volumes.

• inf1.xlarge and inf1.2xlarge instances support a maximum of 26 EBS volumes.
• inf1.6xlarge instances support a maximum of 23 EBS volumes.
• inf1.24xlarge instances support a maximum of 11 EBS volumes.
• Most bare metal instances support a maximum of 31 EBS volumes.

1291
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Linux-specific volume limits

• u-6tb1.metal, u-9tb1.metal, and u-12tb1.metal instances support a maximum of 19 EBS

volumes if launched after March 12, 2020 and a maximum of 14 EBS volumes otherwise. To attach
more than 14 EBS volumes to an instance launched before March 12, 2020, contact your account team
to upgrade the instance at no additional cost.
• u-18tb1.metal and u-24tb1.metal instances support a maximum of 19 EBS volumes.

Linux-specific volume limits

Attaching more than 40 volumes can cause boot failures. This number includes the root volume, plus
any attached instance store volumes and EBS volumes. If you experience boot problems on an instance
with a large number of volumes, stop the instance, detach any volumes that are not essential to the boot
process, and then reattach the volumes after the instance is running.
Important
Attaching more than 40 volumes to a Linux instance is supported on a best effort basis only and
is not guaranteed.

Bandwidth versus capacity

For consistent and predictable bandwidth use cases, use EBS-optimized or 10 Gigabit network
connectivity instances and General Purpose SSD or Provisioned IOPS SSD volumes. Follow the guidance
in Amazon EBS–optimized instances (p. 1217) to match the IOPS you have provisioned for your volumes
to the bandwidth available from your instances for maximum performance. For RAID configurations,
many administrators find that arrays larger than 8 volumes have diminished performance returns due to
increased I/O overhead. Test your individual application performance and tune it as required.

Name devices on Linux instances

When you attach a volume to your instance, you include a device name for the volume. This device name
is used by Amazon EC2. The block device driver for the instance assigns the actual volume name when
mounting the volume, and the name assigned can be different from the name that Amazon EC2 uses.

The number of volumes that your instance can support is determined by the operating system. For more
information, see Instance volume limits (p. 1291).

Contents
• Available device names (p. 1292)
• Device name considerations (p. 1293)

For information about device names on Windows instances, see Device naming on Windows instances in
the Amazon EC2 User Guide for Windows Instances.

Available device names

There are two types of virtualization available for Linux instances: paravirtual (PV) and hardware virtual
machine (HVM). The virtualization type of an instance is determined by the AMI used to launch the
instance. All instance types support HVM AMIs. Some previous generation instance types support PV
AMIs. Be sure to note the virtualization type of your AMI because the recommended and available device
names that you can use depend on the virtualization type of your instance. For more information, see
Linux AMI virtualization types (p. 104).

The following table lists the available device names that you can specify in a block device mapping or
when attaching an EBS volume.

1292
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Device name considerations

Virtualization Available Reserved for root Recommended Instance store

type for EBS volumes volumes

Paravirtual /dev/sd[a-z] /dev/sda1 /dev/sd[f-p] /dev/sd[b-e]

/dev/sd[a-z][1-15] /dev/sd[f-p][1-6]

/dev/hd[a-z]

/dev/hd[a-z]
[1-15]

HVM /dev/sd[a-z] Differs by AMI /dev/sd[f-p] * /dev/sd[b-e]

/dev/xvd[b-c][a-z] /dev/sda1 or / /dev/sd[b-h]

dev/xvda (h1.16xlarge)

/dev/sd[b-y]
(d2.8xlarge)

/dev/sd[b-i]
(i2.8xlarge)

**

* The device names that you specify for NVMe EBS volumes in a block device mapping are renamed using
NVMe device names (/dev/nvme[0-26]n1). The block device driver can assign NVMe device names in a
different order than you specified for the volumes in the block device mapping.

** NVMe instance store volumes are automatically enumerated and assigned an NVMe device name.

For more information about instance store volumes, see Amazon EC2 instance store (p. 1268). For more
information about NVMe EBS volumes (Nitro-based instances), including how to identify the EBS device,
see Amazon EBS and NVMe on Linux instances (p. 1214).

Device name considerations

Keep the following in mind when selecting a device name:

• Although you can attach your EBS volumes using the device names used to attach instance store
volumes, we strongly recommend that you don't because the behavior can be unpredictable.
• The number of NVMe instance store volumes for an instance depends on the size of the instance.
NVMe instance store volumes are automatically enumerated and assigned an NVMe device name (/
dev/nvme[0-26]n1).
• Depending on the block device driver of the kernel, the device could be attached with a different
name than you specified. For example, if you specify a device name of /dev/sdh, your device could
be renamed /dev/xvdh or /dev/hdh. In most cases, the trailing letter remains the same. In some
versions of Red Hat Enterprise Linux (and its variants, such as CentOS), the trailing letter could change
(/dev/sda could become /dev/xvde). In these cases, the trailing letter of each device name is
incremented the same number of times. For example, if /dev/sdb is renamed /dev/xvdf, then /
dev/sdc is renamed /dev/xvdg. Amazon Linux creates a symbolic link for the name you specified to
the renamed device. Other operating systems could behave differently.
• HVM AMIs do not support the use of trailing numbers on device names, except for /dev/sda1,
which is reserved for the root device, and /dev/sda2. While using /dev/sda2 is possible, we do not
recommend using this device mapping with HVM instances.

1293
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Block device mapping

• When using PV AMIs, you cannot attach volumes that share the same device letters both with and
without trailing digits. For example, if you attach a volume as /dev/sdc and another volume as /
dev/sdc1, only /dev/sdc is visible to the instance. To use trailing digits in device names, you must
use trailing digits on all device names that share the same base letters (such as /dev/sdc1, /dev/
sdc2, /dev/sdc3).
• Some custom kernels might have restrictions that limit use to /dev/sd[f-p] or /dev/sd[f-p]
[1-6]. If you're having trouble using /dev/sd[q-z] or /dev/sd[q-z][1-6], try switching to /
dev/sd[f-p] or /dev/sd[f-p][1-6].

Block device mapping

Each instance that you launch has an associated root device volume, which is either an Amazon
EBS volume or an instance store volume. You can use block device mapping to specify additional
EBS volumes or instance store volumes to attach to an instance when it's launched. You can also
attach additional EBS volumes to a running instance; see Attach an Amazon EBS volume to an
instance (p. 1106). However, the only way to attach instance store volumes to an instance is to use block
device mapping to attach the volumes as the instance is launched.

For more information about root device volumes, see Change the root volume to persist (p. 24).

Contents
• Block device mapping concepts (p. 1294)
• AMI block device mapping (p. 1297)
• Instance block device mapping (p. 1299)

Block device mapping concepts

A block device is a storage device that moves data in sequences of bytes or bits (blocks). These devices
support random access and generally use buffered I/O. Examples include hard disks, CD-ROM drives, and
flash drives. A block device can be physically attached to a computer or accessed remotely as if it were
physically attached to the computer.

Amazon EC2 supports two types of block devices:

• Instance store volumes (virtual devices whose underlying hardware is physically attached to the host
computer for the instance)
• EBS volumes (remote storage devices)

A block device mapping defines the block devices (instance store volumes and EBS volumes) to attach
to an instance. You can specify a block device mapping as part of creating an AMI so that the mapping
is used by all instances launched from the AMI. Alternatively, you can specify a block device mapping
when you launch an instance, so this mapping overrides the one specified in the AMI from which you
launched the instance. Note that all NVMe instance store volumes supported by an instance type are
automatically enumerated and assigned a device name on instance launch; including them in your block
device mapping has no effect.

Contents
• Block device mapping entries (p. 1295)
• Block device mapping instance store caveats (p. 1295)
• Example block device mapping (p. 1296)
• How devices are made available in the operating system (p. 1296)

1294
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Block device mapping concepts

Block device mapping entries

When you create a block device mapping, you specify the following information for each block device
that you need to attach to the instance:

• The device name used within Amazon EC2. The block device driver for the instance assigns the actual
volume name when mounting the volume. The name assigned can be different from the name that
Amazon EC2 recommends. For more information, see Name devices on Linux instances (p. 1292).

For Instance store volumes, you also specify the following information:

• The virtual device: ephemeral[0-23]. Note that the number and size of available instance store
volumes for your instance varies by instance type.

For NVMe instance store volumes, the following information also applies:

• These volumes are automatically enumerated and assigned a device name; including them in your
block device mapping has no effect.

For EBS volumes, you also specify the following information:

• The ID of the snapshot to use to create the block device (snap-xxxxxxxx). This value is optional as long
as you specify a volume size.
• The size of the volume, in GiB. The specified size must be greater than or equal to the size of the
specified snapshot.
• Whether to delete the volume on instance termination (true or false). The default value is true
for the root device volume and false for attached volumes. When you create an AMI, its block device
mapping inherits this setting from the instance. When you launch an instance, it inherits this setting
from the AMI.
• The volume type, which can be gp2 and gp3 for General Purpose SSD, io1 and io2 for Provisioned
IOPS SSD, st1 for Throughput Optimized HDD, sc1 for Cold HDD, or standard for Magnetic. The
default value is gp2.
• The number of input/output operations per second (IOPS) that the volume supports. (Used only with
io1 and io2 volumes.)

Block device mapping instance store caveats

There are several caveats to consider when launching instances with AMIs that have instance store
volumes in their block device mappings.

• Some instance types include more instance store volumes than others, and some instance types
contain no instance store volumes at all. If your instance type supports one instance store volume, and
your AMI has mappings for two instance store volumes, then the instance launches with one instance
store volume.
• Instance store volumes can only be mapped at launch time. You cannot stop an instance without
instance store volumes (such as the t2.micro), change the instance to a type that supports instance
store volumes, and then restart the instance with instance store volumes. However, you can create an
AMI from the instance and launch it on an instance type that supports instance store volumes, and
map those instance store volumes to the instance.
• If you launch an instance with instance store volumes mapped, and then stop the instance and change
it to an instance type with fewer instance store volumes and restart it, the instance store volume
mappings from the initial launch still show up in the instance metadata. However, only the maximum
number of supported instance store volumes for that instance type are available to the instance.

1295
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Block device mapping concepts

Note
When an instance is stopped, all data on the instance store volumes is lost.
• Depending on instance store capacity at launch time, M3 instances may ignore AMI instance store
block device mappings at launch unless they are specified at launch. You should specify instance
store block device mappings at launch time, even if the AMI you are launching has the instance store
volumes mapped in the AMI, to ensure that the instance store volumes are available when the instance
launches.

Example block device mapping

This figure shows an example block device mapping for an EBS-backed instance. It maps /dev/sdb to
ephemeral0 and maps two EBS volumes, one to /dev/sdh and the other to /dev/sdj. It also shows
the EBS volume that is the root device volume, /dev/sda1.

Note that this example block device mapping is used in the example commands and APIs in this
topic. You can find example commands and APIs that create block device mappings in Specify a
block device mapping for an AMI (p. 1297) and Update the block device mapping when launching an
instance (p. 1299).

How devices are made available in the operating system

Device names like /dev/sdh and xvdh are used by Amazon EC2 to describe block devices. The block
device mapping is used by Amazon EC2 to specify the block devices to attach to an EC2 instance. After

1296
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI block device mapping

a block device is attached to an instance, it must be mounted by the operating system before you can
access the storage device. When a block device is detached from an instance, it is unmounted by the
operating system and you can no longer access the storage device.

With a Linux instance, the device names specified in the block device mapping are mapped to their
corresponding block devices when the instance first boots. The instance type determines which instance
store volumes are formatted and mounted by default. You can mount additional instance store volumes
at launch, as long as you don't exceed the number of instance store volumes available for your instance
type. For more information, see Amazon EC2 instance store (p. 1268). The block device driver for the
instance determines which devices are used when the volumes are formatted and mounted. For more
information, see Attach an Amazon EBS volume to an instance (p. 1106).

AMI block device mapping

Each AMI has a block device mapping that specifies the block devices to attach to an instance when it
is launched from the AMI. An AMI that Amazon provides includes a root device only. To add more block
devices to an AMI, you must create your own AMI.

Contents
• Specify a block device mapping for an AMI (p. 1297)
• View the EBS volumes in an AMI block device mapping (p. 1298)

Specify a block device mapping for an AMI

There are two ways to specify volumes in addition to the root volume when you create an AMI. If you've
already attached volumes to a running instance before you create an AMI from the instance, the block
device mapping for the AMI includes those same volumes. For EBS volumes, the existing data is saved
to a new snapshot, and it's this new snapshot that's specified in the block device mapping. For instance
store volumes, the data is not preserved.

For an EBS-backed AMI, you can add EBS volumes and instance store volumes using a block device
mapping. For an instance store-backed AMI, you can add instance store volumes only by modifying the
block device mapping entries in the image manifest file when registering the image.
Note
For M3 instances, you must specify instance store volumes in the block device mapping for the
instance when you launch it. When you launch an M3 instance, instance store volumes specified
in the block device mapping for the AMI may be ignored if they are not specified as part of the
instance block device mapping.

To add volumes to an AMI using the console

1. Open the Amazon EC2 console.

2. In the navigation pane, choose Instances.
3. Select an instance and choose Actions, Image and templates, Create image.
4. Enter a name and a description for the image.
5. The instance volumes appear under Instance volumes. To add another volume, choose Add volume.
6. For Volume type, choose the volume type. For Device choose the device name. For an EBS volume,
you can specify additional details, such as a snapshot, volume size, volume type, IOPS, and
encryption state.
7. Choose Create image.

To add volumes to an AMI using the command line

1297
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
AMI block device mapping

Use the create-image AWS CLI command to specify a block device mapping for an EBS-backed AMI. Use
the register-image AWS CLI command to specify a block device mapping for an instance store-backed
AMI.

Specify the block device mapping using the --block-device-mappings parameter. Arguments
encoded in JSON can be supplied either directly on the command line or by reference to a file:

--block-device-mappings [mapping, ...]

--block-device-mappings [file://mapping.json]

To add an instance store volume, use the following mapping.

{
"DeviceName": "/dev/sdf",
"VirtualName": "ephemeral0"
}

To add an empty 100 GiB gp2 volume, use the following mapping.

{
"DeviceName": "/dev/sdg",
"Ebs": {
"VolumeSize": 100
}
}

To add an EBS volume based on a snapshot, use the following mapping.

{
"DeviceName": "/dev/sdh",
"Ebs": {
"SnapshotId": "snap-xxxxxxxx"
}
}

To omit a mapping for a device, use the following mapping.

{
"DeviceName": "/dev/sdj",
"NoDevice": ""
}

Alternatively, you can use the -BlockDeviceMapping parameter with the following commands (AWS
Tools for Windows PowerShell):

• New-EC2Image
• Register-EC2Image

View the EBS volumes in an AMI block device mapping

You can easily enumerate the EBS volumes in the block device mapping for an AMI.

To view the EBS volumes for an AMI using the console

1. Open the Amazon EC2 console.

1298
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance block device mapping

2. In the navigation pane, choose AMIs.

3. Choose EBS images from the Filter list to get a list of EBS-backed AMIs.
4. Select the desired AMI, and look at the Details tab. At a minimum, the following information is
available for the root device:

• Root Device Type (ebs)

• Root Device Name (for example, /dev/sda1)
• Block Devices (for example, /dev/sda1=snap-1234567890abcdef0:8:true)

If the AMI was created with additional EBS volumes using a block device mapping, the Block Devices
field displays the mapping for those additional volumes as well. (This screen doesn't display instance
store volumes.)

To view the EBS volumes for an AMI using the command line

Use the describe-images (AWS CLI) command or Get-EC2Image (AWS Tools for Windows PowerShell)
command to enumerate the EBS volumes in the block device mapping for an AMI.

Instance block device mapping

By default, an instance that you launch includes any storage devices specified in the block device
mapping of the AMI from which you launched the instance. You can specify changes to the block device
mapping for an instance when you launch it, and these updates overwrite or merge with the block device
mapping of the AMI.

Limitations

• For the root volume, you can only modify the following: volume size, volume type, and the Delete on
Termination flag.
• When you modify an EBS volume, you can't decrease its size. Therefore, you must specify a snapshot
whose size is equal to or greater than the size of the snapshot specified in the block device mapping of
the AMI.

Contents
• Update the block device mapping when launching an instance (p. 1299)
• Update the block device mapping of a running instance (p. 1301)
• View the EBS volumes in an instance block device mapping (p. 1301)
• View the instance block device mapping for instance store volumes (p. 1302)

Update the block device mapping when launching an instance

You can add EBS volumes and instance store volumes to an instance when you launch it. Note that
updating the block device mapping for an instance doesn't make a permanent change to the block
device mapping of the AMI from which it was launched.

To add volumes to an instance using the console

1. Open the Amazon EC2 console.

2. From the dashboard, choose Launch Instance.
3. On the Choose an Amazon Machine Image (AMI) page, select the AMI to use and choose Select.

1299
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance block device mapping

4. Follow the wizard to complete the Choose an Instance Type and Configure Instance Details pages.
5. On the Add Storage page, you can modify the root volume, EBS volumes, and instance store
volumes as follows:

• To change the size of the root volume, locate the Root volume under the Type column, and
change its Size field.
• To suppress an EBS volume specified by the block device mapping of the AMI used to launch the
instance, locate the volume and click its Delete icon.
• To add an EBS volume, choose Add New Volume, choose EBS from the Type list, and fill in the
fields (Device, Snapshot, and so on).
• To suppress an instance store volume specified by the block device mapping of the AMI used to
launch the instance, locate the volume, and choose its Delete icon.
• To add an instance store volume, choose Add New Volume, select Instance Store from the Type
list, and select a device name from Device.
6. Complete the remaining wizard pages, and choose Launch.

To add volumes to an instance using the AWS CLI

Use the run-instances AWS CLI command with the --block-device-mappings option to specify a
block device mapping for an instance at launch.

For example, suppose that an EBS-backed AMI specifies the following block device mapping:

• /dev/sdb=ephemeral0
• /dev/sdh=snap-1234567890abcdef0
• /dev/sdj=:100

To prevent /dev/sdj from attaching to an instance launched from this AMI, use the following mapping.

{
"DeviceName": "/dev/sdj",
"NoDevice": ""
}

To increase the size of /dev/sdh to 300 GiB, specify the following mapping. Notice that you don't need
to specify the snapshot ID for /dev/sdh, because specifying the device name is enough to identify the
volume.

{
"DeviceName": "/dev/sdh",
"Ebs": {
"VolumeSize": 300
}
}

To increase the size of the root volume at instance launch, first call describe-images with the ID of the
AMI to verify the device name of the root volume. For example, "RootDeviceName": "/dev/xvda".
To override the size of the root volume, specify the device name of the root device used by the AMI and
the new volume size.

{
"DeviceName": "/dev/xvda",
"Ebs": {
"VolumeSize": 100

1300
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance block device mapping

}
}

To attach an additional instance store volume, /dev/sdc, specify the following mapping. If the instance
type doesn't support multiple instance store volumes, this mapping has no effect. If the instance
supports NVMe instance store volumes, they are automatically enumerated and assigned an NVMe
device name.

{
"DeviceName": "/dev/sdc",
"VirtualName": "ephemeral1"
}

To add volumes to an instance using the AWS Tools for Windows PowerShell

Use the -BlockDeviceMapping parameter with the New-EC2Instance command (AWS Tools for
Windows PowerShell).

Update the block device mapping of a running instance

You can use the modify-instance-attribute AWS CLI command to update the block device mapping of a
running instance. You do not need to stop the instance before changing this attribute.

aws ec2 modify-instance-attribute --instance-id i-1a2b3c4d --block-device-mappings file://

mapping.json

For example, to preserve the root volume at instance termination, specify the following in
mapping.json.

[
{
"DeviceName": "/dev/sda1",
"Ebs": {
"DeleteOnTermination": false
}
}
]

Alternatively, you can use the -BlockDeviceMapping parameter with the Edit-EC2InstanceAttribute
command (AWS Tools for Windows PowerShell).

View the EBS volumes in an instance block device mapping

You can easily enumerate the EBS volumes mapped to an instance.
Note
For instances launched before the release of the 2009-10-31 API, AWS can't display the block
device mapping. You must detach and reattach the volumes so that AWS can display the block
device mapping.

To view the EBS volumes for an instance using the console

1. Open the Amazon EC2 console.

2. In the navigation pane, choose Instances.
3. In the search box, enter Root device type, and then choose EBS. This displays a list of EBS-backed
instances.

1301
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance block device mapping

4. Select the desired instance and look at the details displayed in the Storage tab. At a minimum, the
following information is available for the root device:

• Root device type (for example, EBS)

• Root device name (for example, /dev/xvda)
• Block devices (for example, /dev/xvda, /dev/sdf, and /dev/sdj)

If the instance was launched with additional EBS volumes using a block device mapping, they appear
under Block devices. Any instance store volumes do not appear on this tab.
5. To display additional information about an EBS volume, choose its volume ID to go to the volume
page. For more information, see View information about an Amazon EBS volume (p. 1114).

To view the EBS volumes for an instance using the command line

Use the describe-instances (AWS CLI) command or Get-EC2Instance (AWS Tools for Windows PowerShell)
command to enumerate the EBS volumes in the block device mapping for an instance.

View the instance block device mapping for instance store

volumes
When you view the block device mapping for your instance, you can see only the EBS volumes, not the
instance store volumes. You can use instance metadata to query the non-NVMe instance store volumes in
the block device mapping. NVMe instance store volumes are not included.

The base URI for all requests for instance metadata is http://169.254.169.254/latest/. For more
information, see Instance metadata and user data (p. 704).

First, connect to your running instance. From the instance, use this query to get its block device mapping.

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" –v http://169.254.169.254/latest/meta-
data/block-device-mapping/

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/block-device-mapping/

The response includes the names of the block devices for the instance. For example, the output for an
instance store–backed m1.small instance looks like this.

ami
ephemeral0
root
swap

The ami device is the root device as seen by the instance. The instance store volumes are named
ephemeral[0-23]. The swap device is for the page file. If you've also mapped EBS volumes, they
appear as ebs1, ebs2, and so on.

To get details about an individual block device in the block device mapping, append its name to the
previous query, as shown here.

1302
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance block device mapping

IMDSv2

[ec2-user ~]$ TOKEN=`curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-

ec2-metadata-token-ttl-seconds: 21600"` \
&& curl -H "X-aws-ec2-metadata-token: $TOKEN" –v http://169.254.169.254/latest/meta-
data/block-device-mapping/ephemeral0

IMDSv1

[ec2-user ~]$ curl http://169.254.169.254/latest/meta-data/block-device-mapping/

ephemeral0

The instance type determines the number of instance store volumes that are available to the instance. If
the number of instance store volumes in a block device mapping exceeds the number of instance store
volumes available to an instance, the additional volumes are ignored. To view the instance store volumes
for your instance, run the lsblk command. To learn how many instance store volumes are supported by
each instance type, see Instance store volumes (p. 1270).

1303
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Resource locations

Resources and tags

Amazon EC2 provides different resources that you can create and use. Some of these resources include
images, instances, volumes, and snapshots. When you create a resource, we assign the resource a unique
resource ID.

Some resources can be tagged with values that you define, to help you organize and identify them.

The following topics describe resources and tags, and how you can work with them.

Contents
• Resource locations (p. 1304)
• Resource IDs (p. 1305)
• List and filter your resources (p. 1306)
• Tag your Amazon EC2 resources (p. 1311)
• Amazon EC2 service quotas (p. 1323)
• Amazon EC2 usage reports (p. 1325)

Resource locations
Some resources can be used in all AWS Regions (global), and some resources are specific to the AWS
Region or Availability Zone in which they reside.

Resource Type Description

AWS account Global You can use the same AWS account in all Regions.

Key pairs Global or The key pairs that you create using Amazon EC2 are
Regional tied to the Region where you created them. You can
create your own RSA key pair and upload it to the
Region in which you want to use it; therefore, you can
make your key pair globally available by uploading it
to each Region.

For more information, see Amazon EC2 key pairs and

Linux instances (p. 1044).

Amazon EC2 resource Regional Each resource identifier, such as an AMI ID, instance ID,
identifiers EBS volume ID, or EBS snapshot ID, is tied to its Region
and can be used only in the Region where you created
the resource.

User-supplied resource Regional Each resource name, such as a security group name
names or key pair name, is tied to its Region and can be used
only in the Region where you created the resource.
Although you can create resources with the same name
in multiple Regions, they aren't related to each other.

AMIs Regional An AMI is tied to the Region where its files are located
within Amazon S3. You can copy an AMI from one
Region to another. For more information, see Copy an
AMI (p. 166).

1304
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Resource IDs

Resource Type Description

Elastic IP addresses Regional An Elastic IP address is tied to a Region and can be

associated only with an instance in the same Region.

Security groups Regional A security group is tied to a Region and can be

assigned only to instances in the same Region. You
can't enable an instance to communicate with an
instance outside its Region using security group rules.
Traffic from an instance in another Region is seen as
WAN bandwidth.

EBS snapshots Regional An EBS snapshot is tied to its Region and can only
be used to create volumes in the same Region. You
can copy a snapshot from one Region to another.
For more information, see Copy an Amazon EBS
snapshot (p. 1134).

EBS volumes Availability Zone An Amazon EBS volume is tied to its Availability Zone
and can be attached only to instances in the same
Availability Zone.

Instances Availability Zone An instance is tied to the Availability Zones in which

you launched it. However, its instance ID is tied to the
Region.

Resource IDs
When resources are created, we assign each resource a unique resource ID. A resource ID takes the form
of a resource identifier (such as snap for a snapshot) followed by a hyphen and a unique combination of
letters and numbers.

You can use resource IDs to find your resources in the Amazon EC2 console. If you are using a command
line tool or the Amazon EC2 API to work with Amazon EC2, resource IDs are required for certain
commands. For example, if you are using the stop-instances AWS CLI command to stop an instance, you
must specify the instance ID in the command.

Resource ID length
Prior to January 2016, the IDs assigned to newly created resources of certain resource types used
8 characters after the hyphen (for example, i-1a2b3c4d). From January 2016 to June 2018, we
changed the IDs of these resource types to use 17 characters after the hyphen (for example,
i-1234567890abcdef0). Depending on when your account was created, you might have resources of
the following resource types with short IDs, though any new resources of these types receive the longer
IDs:

• bundle
• conversion-task
• customer-gateway
• dhcp-options
• elastic-ip-allocation
• elastic-ip-association
• export-task
• flow-log
• image

1305
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List and filter your resources

• import-task
• instance
• internet-gateway
• network-acl
• network-acl-association
• network-interface
• network-interface-attachment
• prefix-list
• route-table
• route-table-association
• security-group
• snapshot
• subnet
• subnet-cidr-block-association
• reservation
• volume
• vpc
• vpc-cidr-block-association
• vpc-endpoint
• vpc-peering-connection
• vpn-connection
• vpn-gateway

List and filter your resources

You can get a list of some types of resources using the Amazon EC2 console. You can get a list of each
type of resource using its corresponding command or API action. If you have many resources, you can
filter the results to include only the resources that match certain criteria.

Contents
• List and filter resources using the console (p. 1306)
• List and filter using the CLI and API (p. 1309)

List and filter resources using the console

Contents
• List resources using the console (p. 1306)
• Filter resources using the console (p. 1307)

List resources using the console

You can view the most common Amazon EC2 resource types using the console. To view additional
resources, use the command line interface or the API actions.

To list EC2 resources using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

1306
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List and filter resources using the console

2. In the navigation pane, choose the option that corresponds to the resource type. For example, to list
your instances, choose Instances.
3. The page displays all resources of the selected resource type.

Filter resources using the console

Search functionality differs slightly between the old and new Amazon EC2 console.

New console

The new console supports two types of filtering.

• API filtering happens on the server side. The filtering is applied on the API call and it reduces the
number of resources returned by the server. It allows for quick filtering across large sets of resources,
and it can reduce data transfer time and cost between the server and the browser.
• Client filtering happens on the client side. It enables you to filter down on data that is already available
in the browser (in other words, data that has already been returned by the API). Client filtering works
well in conjunction with an API filter to filter down to smaller data sets in the browser.

The new Amazon EC2 console supports the following types of searches:

Search by keyword

Searching by keyword is a free text search that lets you search for a value across all of your
resources' attributes, without specifying an attribute to search.
Note
All keyword searches use client filtering.

To search by keyword, enter or paste what you’re looking for in the search field, and then choose
Enter. For example, searching for 123 matches all instances that have 123 in any of their attributes,
such as an IP address, instance ID, VPC ID, or AMI ID. If your free text search returns unexpected
matches, apply additional filters.
Search by attributes

Searching by an attribute lets you search a specific attribute across all of your resources.
Note
Attribute searches use either API filtering or client filtering, depending on the selected
attribute. When performing an attribute search, the attributes are grouped accordingly.

For example, you can search the Instance state attribute for all of your instances to return only
instances that are in the stopped state. To do this:
1. In the search field on the Instances screen, start entering Instance state. As you enter
characters, a list of matching attributes appears.
2. Select Instance state from the list. A list of possible values for the selected attribute appears.
3. Select Stopped from the list.

You can use the following techniques to enhance or refine your searches:

Inverse search

Inverse searches let you search for resources that do not match a specified value. Inverse searches
are performed by prefixing the search keyword with the exclamation mark (!) character. For example,
to list all instances that are not assigned the security group named launch-wizard-1, search by
the Security group name attribute, and for the keyword, enter !launch-wizard-1.

1307
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List and filter resources using the console

Note
Inverse search is supported with keyword searches and attribute searches on client filters
only. It is not supported with attribute searches on API filters.
Partial search

With partial searches, you can search for partial string values. To perform a partial search, enter
only a part of the keyword that you want to search for. For example, to search for all t2.micro,
t2.small, and t2.medium instances, search by the Instance Type attribute, and for the keyword,
enter t2.
Note
Partial search is supported with keyword searches and attribute searches on client filters
only. It is not supported with attribute searches on API filters.
Regular expression search

To use regular expression searches, you must enable Use regular expression matching in the
Preferences.

Regular expressions are useful when you need to match the values in a field with a specific pattern.
For example, to search for a value that starts with s, search for ^s. To search for a value that ends
with xyz, search for xyz$. Or to search for a value that starts with a number that is followed by one
or more characters, search for [0-9]+.*. Regular expression searches are not case-sensitive.
Note
Regular expression search is supported with keyword searches and attribute searches on
client filters only. It is not supported with attribute searches on API filters.
Wildcard search

Use the * wildcard to match zero or more characters. Use the ? wildcard to match zero or one
character. For example, if you have a data set with the following values: prod, prods, and production;
"prod*" matches all values, whereas "prod?" matches only prod and prods. To use the literals
values, escape them with a backslash (\). For example, "prod\*" would match prod*.
Note
Wildcard search is supported with attribute searches on API filters only. It is not supported
with keyword searches and attribute searches on client filters only.
Combining searches

In general, multiple filters with the same attribute are automatically joined with OR. For example,
searching for Instance State : Running and Instance State : Stopped returns all
instances that are either running OR stopped. To join search with AND, search across different
attributes. For example, searching for Instance State : Running and Instance Type :
c4.large returns only instances that are of type c4.large AND that are in the stopped state.

Old console
The old Amazon EC2 console supports the following types of searches:

Search by keyword

Searching by keyword is a free text search that lets you search for a value across all of your
resources' attributes. To search by keyword, enter or paste what you’re looking for in the search field,
and then choose Enter. For example, searching for 123 matches all instances that have 123 in any of
their attributes, such as an IP address, instance ID, VPC ID, or AMI ID. If your free text search returns
unexpected matches, apply additional filters.
Search by attributes

Searching by an attribute lets you search a specific attribute across all of your resources. For
example, you can search the State attribute for all of your instances to return only instances that are
in the stopped state. To do this:

1308
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List and filter using the CLI and API

1. In the search field on the Instances screen, start entering Instance State. As you enter
characters, a list of matching attributes appears.
2. Select Instance State from the list. A list of possible values for the selected attribute appears.
3. Select Stopped from the list.

You can use the following techniques to enhance or refine your searches:

Inverse search

Inverse searches let you search for resources that do not match a specified value. Inverse searches
are performed by prefixing the search keyword with the exclamation mark (!) character. For example,
to list all instances that are not terminated, search by the Instance State attribute, and for the
keyword, enter !Terminated.
Partial search

With partial searches, you can search for partial string values. To perform a partial search, enter only
a part of the keyword you want to search for. For example, to search for all t2.micro, t2.small,
and t2.medium instances, search by the Instance Type attribute, and for the keyword, enter t2.
Regular expression search

Regular expressions are useful when you need to match the values in a field with a specific pattern.
For example, to search for all instances that have an attribute value that starts with s, search for
^s. Or to search for all instances that have an attribute value that ends with xyz, search for xyz$.
Regular expression searches are not case-sensitive.
Combining searches

In general, multiple filters with the same attribute are automatically joined with OR. For example,
searching for Instance State : Running and Instance State : Stopped returns all
instances that are either running OR stopped. To join search with AND, search across different
attributes. For example, searching for Instance State : Running and Instance Type :
c4.large returns only instances that are of type c4.large AND that are in the stopped state.

To filter a list of resources

1. In the navigation pane, select a resource type (for example, Instances).

2. Choose the search field.
3. Choose the filter from in the list.
4. Specify a filter value.
5. When you are finished, remove the filter.

List and filter using the CLI and API

Each resource type has a corresponding CLI command and API action that you use to list resources of
that type. The resulting lists of resources can be long, so it can be faster and more useful to filter the
results to include only the resources that match specific criteria.

Filtering considerations

• You can specify multiple filters and multiple filter values in a single request.
• You can use wildcards with the filter values. An asterisk (*) matches zero or more characters, and a
question mark (?) matches zero or one character.
• Filter values are case sensitive.

1309
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
List and filter using the CLI and API

• Your search can include the literal values of the wildcard characters; you just need to escape them with
a backslash before the character. For example, a value of \*amazon\?\\ searches for the literal string
*amazon?\.

Supported filters
To see the supported filters for each Amazon EC2 resource, see the following documentation:

• AWS CLI: The describe commands in the AWS CLI Command Reference-Amazon EC2.
• Tools for Windows PowerShell: The Get commands in the AWS Tools for PowerShell Cmdlet
Reference-Amazon EC2.
• Query API: The Describe API actions in the Amazon EC2 API Reference.

Example Example: Specify a single filter

You can list your Amazon EC2 instances using describe-instances. Without filters, the response contains
information for all of your resources. You can use the following command to include only the running
instances in your output.

aws ec2 describe-instances --filters Name=instance-state-name,Values=running

To list only the instance IDs for your running instances, add the --query parameter as follows.

aws ec2 describe-instances --filters Name=instance-state-name,Values=running --query

"Reservations[*].Instances[*].InstanceId" --output text

The following is example output.

i-0ef1f57f78d4775a4
i-0626d4edd54f1286d
i-04a636d18e83cfacb

Example Example: Specify multiple filters or filter values

If you specify multiple filters or multiple filter values, the resource must match all filters to be included
in the results.

You can use the following command to list all instances whose type is either m5.large or m5d.large.

aws ec2 describe-instances --filters Name=instance-type,Values=m5.large,m5d.large

You can use the following command to list all stopped instances whose type is t2.micro.

aws ec2 describe-instances --filters Name=instance-state-name,Values=stopped Name=instance-

type,Values=t2.micro

Example Example: Use wildcards in a filter value

If you specify database as the filter value for the description filter when describing EBS snapshots
using describe-snapshots, the command returns only the snapshots whose description is "database".

aws ec2 describe-snapshots --filters Name=description,Values=database

The * wildcard matches zero or more characters. If you specify *database* as the filter value, the
command returns only snapshots whose description includes the word database.

1310
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tag your resources

aws ec2 describe-snapshots --filters Name=description,Values=*database*

The ? wildcard matches exactly 1 character. If you specify database? as the filter value, the command
returns only snapshots whose description is "database" or "database" followed by one character.

aws ec2 describe-snapshots --filters Name=description,Values=database?

If you specify database????, the command returns only snapshots whose description is "database"
followed by up to four characters. It excludes descriptions with "database" followed by five or more
characters.

aws ec2 describe-snapshots --filters Name=description,Values=database????

Example Example: Filter based on date

With the AWS CLI, you can use JMESPath to filter results using expressions. For example, the
following describe-snapshots command displays the IDs of all snapshots created by your AWS account
(represented by 123456789012) before the specified date (represented by 2020-03-31). If you do not
specify the owner, the results include all public snapshots.

aws ec2 describe-snapshots --filters Name=owner-id,Values=123456789012 --query "Snapshots[?

(StartTime<='2020-03-31')].[SnapshotId]" --output text

The following command displays the IDs of all snapshots created in the specified date range.

aws ec2 describe-snapshots --filters Name=owner-id,Values=123456789012 --query "Snapshots[?

(StartTime>='2019-01-01') && (StartTime<='2019-12-31')].[SnapshotId]" --output text

Filter based on tags

For examples of how to filter a list of resources according to their tags, see Work with tags using the
command line (p. 1319).

Tag your Amazon EC2 resources

To help you manage your instances, images, and other Amazon EC2 resources, you can assign your own
metadata to each resource in the form of tags. Tags enable you to categorize your AWS resources in
different ways, for example, by purpose, owner, or environment. This is useful when you have many
resources of the same type—you can quickly identify a specific resource based on the tags that you've
assigned to it. This topic describes tags and shows you how to create them.
Warning
Tag keys and their values are returned by many different API calls. Denying access to
DescribeTags doesn’t automatically deny access to tags returned by other APIs. As a best
practice, we recommend that you do not include sensitive data in your tags.

Contents
• Tag basics (p. 1312)
• Tag your resources (p. 1313)
• Tag restrictions (p. 1315)
• Tag your resources for billing (p. 1316)
• Work with tags using the console (p. 1316)

1311
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tag basics

• Work with tags using the command line (p. 1319)

• Add tags to a resource using CloudFormation (p. 1322)

Tag basics
A tag is a label that you assign to an AWS resource. Each tag consists of a key and an optional value, both
of which you define.

Tags enable you to categorize your AWS resources in different ways, for example, by purpose, owner, or
environment. For example, you could define a set of tags for your account's Amazon EC2 instances that
helps you track each instance's owner and stack level.

The following diagram illustrates how tagging works. In this example, you've assigned two tags to each
of your instances—one tag with the key Owner and another with the key Stack. Each tag also has an
associated value.

We recommend that you devise a set of tag keys that meets your needs for each resource type. Using
a consistent set of tag keys makes it easier for you to manage your resources. You can search and filter
the resources based on the tags you add. For more information about how to implement an effective
resource tagging strategy, see the AWS whitepaper Tagging Best Practices.

Tags don't have any semantic meaning to Amazon EC2 and are interpreted strictly as a string of
characters. Also, tags are not automatically assigned to your resources. You can edit tag keys and values,
and you can remove tags from a resource at any time. You can set the value of a tag to an empty string,
but you can't set the value of a tag to null. If you add a tag that has the same key as an existing tag on
that resource, the new value overwrites the old value. If you delete a resource, any tags for the resource
are also deleted.

You can work with tags using the AWS Management Console, the AWS CLI, and the Amazon EC2 API.

1312
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tag your resources

If you're using AWS Identity and Access Management (IAM), you can control which users in your AWS
account have permission to create, edit, or delete tags. For more information, see Identity and access
management for Amazon EC2 (p. 977).

Tag your resources

You can tag most Amazon EC2 resources that already exist in your account. The table (p. 1313) below
lists the resources that support tagging.

If you're using the Amazon EC2 console, you can apply tags to resources by using the Tags tab on the
relevant resource screen, or you can use the Tags screen. Some resource screens enable you to specify
tags for a resource when you create the resource; for example, a tag with a key of Name and a value that
you specify. In most cases, the console applies the tags immediately after the resource is created (rather
than during resource creation). The console may organize resources according to the Name tag, but this
tag doesn't have any semantic meaning to the Amazon EC2 service.

If you're using the Amazon EC2 API, the AWS CLI, or an AWS SDK, you can use the CreateTags EC2
API action to apply tags to existing resources. Additionally, some resource-creating actions enable you
to specify tags for a resource when the resource is created. If tags cannot be applied during resource
creation, we roll back the resource creation process. This ensures that resources are either created with
tags or not created at all, and that no resources are left untagged at any time. By tagging resources at
the time of creation, you can eliminate the need to run custom tagging scripts after resource creation.

The following table describes the Amazon EC2 resources that can be tagged, and the resources that can
be tagged on creation using the Amazon EC2 API, the AWS CLI, or an AWS SDK.

Tagging support for Amazon EC2 resources

Resource Supports tags Supports tagging on creation

AFI Yes Yes

AMI Yes Yes

Bundle task No No

Capacity Reservation Yes Yes

Carrier gateway Yes Yes

Client VPN endpoint Yes Yes

Client VPN route No No

Customer gateway Yes Yes

Dedicated Host Yes Yes

Dedicated Host Reservation Yes Yes

DHCP option Yes Yes

EBS snapshot Yes Yes

EBS volume Yes Yes

EC2 Fleet Yes Yes

Egress-only internet gateway Yes Yes

Elastic IP address Yes Yes

1313
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tag your resources

Resource Supports tags Supports tagging on creation

Elastic Graphics accelerator Yes No

Instance Yes Yes

Instance store volume N/A N/A

Internet gateway Yes Yes

IP address pool (BYOIP) Yes Yes

Key pair Yes Yes

Launch template Yes Yes

Launch template version No No

Local gateway Yes No

Local gateway route table Yes No

Local gateway virtual interface Yes No

Local gateway virtual interface Yes No

group

Local gateway route table VPC Yes Yes

association

Local gateway route table virtual Yes No

interface group association

NAT gateway Yes Yes

Network ACL Yes Yes

Network interface Yes Yes

Placement group Yes Yes

Prefix list Yes Yes

Reserved Instance Yes No

Reserved Instance listing No No

Route table Yes Yes

Spot Fleet request Yes Yes

Spot Instance request Yes Yes

Security group Yes Yes

Subnet Yes Yes

Traffic Mirror filter Yes Yes

Traffic Mirror session Yes Yes

Traffic Mirror target Yes Yes

1314
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tag restrictions

Resource Supports tags Supports tagging on creation

Transit gateway Yes Yes

Transit gateway route table Yes Yes

Transit gateway VPC attachment Yes Yes

Virtual private gateway Yes Yes

VPC Yes Yes

VPC endpoint Yes Yes

VPC endpoint service Yes Yes

VPC endpoint service Yes Yes

configuration

VPC flow log Yes Yes

VPC peering connection Yes Yes

VPN connection Yes Yes

You can tag instances and volumes on creation using the Amazon EC2 Launch Instances wizard in the
Amazon EC2 console. You can tag your EBS volumes on creation using the Volumes screen, or EBS
snapshots using the Snapshots screen. Alternatively, use the resource-creating Amazon EC2 APIs (for
example, RunInstances) to apply tags when creating your resource.

You can apply tag-based resource-level permissions in your IAM policies to the Amazon EC2 API actions
that support tagging on creation to implement granular control over the users and groups that can tag
resources on creation. Your resources are properly secured from creation—tags are applied immediately
to your resources, therefore any tag-based resource-level permissions controlling the use of resources are
immediately effective. Your resources can be tracked and reported on more accurately. You can enforce
the use of tagging on new resources, and control which tag keys and values are set on your resources.

You can also apply resource-level permissions to the CreateTags and DeleteTags Amazon EC2 API
actions in your IAM policies to control which tag keys and values are set on your existing resources. For
more information, see Example: Tag resources (p. 1016).

For more information about tagging your resources for billing, see Using Cost Allocation Tags in the AWS
Billing and Cost Management User Guide.

Tag restrictions
The following basic restrictions apply to tags:

• Maximum number of tags per resource – 50

• For each resource, each tag key must be unique, and each tag key can have only one value.
• Maximum key length – 128 Unicode characters in UTF-8
• Maximum value length – 256 Unicode characters in UTF-8
• Although EC2 allows for any character in its tags, other services are more restrictive. The allowed
characters across services are: letters, numbers, and spaces representable in UTF-8, and the following
characters: + - = . _ : / @.
• Tag keys and values are case-sensitive.

1315
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Tag your resources for billing

• The aws: prefix is reserved for AWS use. If a tag has a tag key with this prefix, then you can't edit or
delete the tag's key or value. Tags with the aws: prefix do not count against your tags per resource
limit.

You can't terminate, stop, or delete a resource based solely on its tags; you must specify the resource
identifier. For example, to delete snapshots that you tagged with a tag key called DeleteMe, you
must use the DeleteSnapshots action with the resource identifiers of the snapshots, such as
snap-1234567890abcdef0.

You can tag public or shared resources, but the tags you assign are available only to your AWS account
and not to the other accounts sharing the resource.

You can't tag all resources. For more information, see Tagging support for Amazon EC2
resources (p. 1313).

Tag your resources for billing

You can use tags to organize your AWS bill to reflect your own cost structure. To do this, sign up to
get your AWS account bill with tag key values included. For more information about setting up a cost
allocation report with tags, see The Monthly Cost Allocation Report in AWS Billing and Cost Management
User Guide. To see the cost of your combined resources, you can organize your billing information
based on resources that have the same tag key values. For example, you can tag several resources with
a specific application name, and then organize your billing information to see the total cost of that
application across several services. For more information, see Using Cost Allocation Tags in the AWS
Billing and Cost Management User Guide.
Note
If you've just enabled reporting, data for the current month is available for viewing after 24
hours.

Cost allocation tags can indicate which resources are contributing to costs, but deleting or deactivating
resources doesn't always reduce costs. For example, snapshot data that is referenced by another
snapshot is preserved, even if the snapshot that contains the original data is deleted. For more
information, see Amazon Elastic Block Store Volumes and Snapshots in the AWS Billing and Cost
Management User Guide.
Note
Elastic IP addresses that are tagged do not appear on your cost allocation report.

Work with tags using the console

Using the Amazon EC2 console, you can see which tags are in use across all of your Amazon EC2
resources in the same Region. You can view tags by resource and by resource type, and you can also view
how many items of each resource type are associated with a specified tag. You can also use the Amazon
EC2 console to apply or remove tags from one or more resources at a time.

For more information about using filters when listing your resources, see List and filter your
resources (p. 1306).

For ease of use and best results, use Tag Editor in the AWS Management Console, which provides a
central, unified way to create and manage your tags. For more information, see Working with Tag Editor
in Getting Started with the AWS Management Console.

Tasks
• Display tags (p. 1317)
• Add and delete tags on an individual resource (p. 1317)

1316
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with tags using the console

• Add and delete tags to a group of resources (p. 1318)

• Add a tag when you launch an instance (p. 1319)
• Filter a list of resources by tag (p. 1319)

Display tags
You can display tags in two different ways in the Amazon EC2 console. You can display the tags for an
individual resource or for all resources.

Display tags for individual resources

When you select a resource-specific page in the Amazon EC2 console, it displays a list of those resources.
For example, if you select Instances from the navigation pane, the console displays your Amazon EC2
instances. When you select a resource from one of these lists (for example, an instance), if the resource
supports tags, you can view and manage its tags. On most resource pages, you can view the tags by
selecting the Tags tab.

You can add a column to the resource list that displays all values for tags with the same key. This column
enables you to sort and filter the resource list by the tag. There are two ways to add a new column to the
resource list to display your tags:

• On the Tags tab, select Show Column. A new column is added to the console.
• Choose the Show/Hide Columns gear-shaped icon, and in the Show/Hide Columns dialog box, select
the tag key under Your Tag Keys.

Display tags for all resources

You can display tags across all resources by selecting Tags from the navigation pane in the Amazon EC2
console. The following image shows the Tags pane, which lists all tags in use by resource type.

Add and delete tags on an individual resource

You can manage tags for an individual resource directly from the resource's page.

1317
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with tags using the console

To add a tag to an individual resource

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region that meets your needs. This choice is important
because some Amazon EC2 resources can be shared between Regions, while others can't. For more
information, see Resource locations (p. 1304).
3. In the navigation pane, select a resource type (for example, Instances).
4. Select the resource from the resource list and choose the Tags tab.
5. Choose Manage tags, Add tag. Enter the key and value for the tag. When you are finished adding
tags, choose Save.

To delete a tag from an individual resource

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region that meets your needs. This choice is important
because some Amazon EC2 resources can be shared between Regions, while others can't. For more
information, see Resource locations (p. 1304).
3. In the navigation pane, choose a resource type (for example, Instances).
4. Select the resource from the resource list and choose the Tags tab.
5. Choose Manage tags. For each tag, choose Remove. When you are finished removing tags, choose
Save.

Add and delete tags to a group of resources

To add a tag to a group of resources

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region that meets your needs. This choice is important
because some Amazon EC2 resources can be shared between Regions, while others can't. For more
information, see Resource locations (p. 1304).
3. In the navigation pane, choose Tags.
4. At the top of the content pane, choose Manage Tags.
5. For Filter, select the type of resource (for example, instances).
6. In the resources list, select the check box next to each resource.
7. Under Add Tag, enter the tag key and value and choose Add Tag.
Note
If you add a new tag with the same tag key as an existing tag, the new tag overwrites the
existing tag.

To remove a tag from a group of resources

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select the Region that meets your needs. This choice is important
because some Amazon EC2 resources can be shared between Regions, while others can't. For more
information, see Resource locations (p. 1304).
3. In the navigation pane, choose Tags, Manage Tags.
4. To view the tags in use, select the Show/Hide Columns gear-shaped icon, and in the Show/Hide
Columns dialog box, select the tag keys to view and choose Close.
5. For Filter, select the type of resource (for example, instances).
6. In the resource list, select the check box next to each resource.

1318
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with tags using the command line

7. Under Remove Tag, enter the tag key and choose Remove Tag.

Add a tag when you launch an instance

To add a tag using the Launch Wizard

1. From the navigation bar, select the Region for the instance. This choice is important because some
Amazon EC2 resources can be shared between Regions, while others can't. Select the Region that
meets your needs. For more information, see Resource locations (p. 1304).
2. Choose Launch Instance.
3. The Choose an Amazon Machine Image (AMI) page displays a list of basic configurations called
Amazon Machine Images (AMIs). Select the AMI to use and choose Select. For more information
about selecting an AMI, see Finding an AMI.
4. On the Configure Instance Details page, configure the instance settings as necessary, and then
choose Next: Add Storage.
5. On the Add Storage page, you can specify additional storage volumes for your instance. Choose
Next: Add Tags when done.
6. On the Add Tags page, specify tags for the instance, the volumes, or both. Choose Add another tag
to add more than one tag to your instance. Choose Next: Configure Security Group when you are
done.
7. On the Configure Security Group page, you can choose from an existing security group that you
own, or let the wizard create a new security group for you. Choose Review and Launch when you are
done.
8. Review your settings. When you're satisfied with your selections, choose Launch. Select an existing
key pair or create a new one, select the acknowledgment check box, and then choose Launch
Instances.

Filter a list of resources by tag

You can filter your list of resources based on one or more tag keys and tag values.

To filter a list of resources by tag

1. In the navigation pane, select a resource type (for example, Instances).

2. Choose the search field.
3. Choose the tag key from in the list.
4. Choose the corresponding tag value from the list.
5. When you are finished, remove the filter.

For more information about filters, see List and filter your resources (p. 1306).

Work with tags using the command line

You can add tags to many EC2 resource when you create them, using the tag specifications parameter for
the create command. You can view the tags for a resource using the describe command for the resource.
You can also add, update, or delete tags for your existing resources using the following commands.

Task AWS CLI AWS Tools for Windows

PowerShell

Add or overwrite one or more tags create-tags New-EC2Tag

1319
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with tags using the command line

Task AWS CLI AWS Tools for Windows

PowerShell

Delete one or more tags delete-tags Remove-EC2Tag

Describe one or more tags describe-tags Get-EC2Tag

Tasks
• Add tags on resource creation (p. 1320)
• Add tags to an existing resource (p. 1321)
• Describe tagged resources (p. 1322)

Add tags on resource creation

The following examples demonstrate how to apply tags when you create resources.

The way you enter JSON-formatted parameters on the command line differs depending on your
operating system. Linux, macOS, or Unix and Windows PowerShell use single quotes (') to enclose the
JSON data structure. Omit the single quotes when using the commands with the Windows command
line. For more information, see Specifying Parameter Values for the AWS Command Line Interface.

Example Example: Launch an instance and apply tags to the instance and volume

The following run-instances command launches an instance and applies a tag with the key webserver
and the value production to the instance. The command also applies a tag with the key cost-center
and the value cc123 to any EBS volume that's created (in this case, the root volume).

aws ec2 run-instances \

--image-id ami-abc12345 \
--count 1 \
--instance-type t2.micro \
--key-name MyKeyPair \
--subnet-id subnet-6e7f829e \
--tag-specifications 'ResourceType=instance,Tags=[{Key=webserver,Value=production}]'
'ResourceType=volume,Tags=[{Key=cost-center,Value=cc123}]'

You can apply the same tag keys and values to both instances and volumes during launch. The following
command launches an instance and applies a tag with a key of cost-center and a value of cc123 to
both the instance and any EBS volume that's created.

aws ec2 run-instances \

--image-id ami-abc12345 \
--count 1 \
--instance-type t2.micro \
--key-name MyKeyPair \
--subnet-id subnet-6e7f829e \
--tag-specifications 'ResourceType=instance,Tags=[{Key=cost-center,Value=cc123}]'
'ResourceType=volume,Tags=[{Key=cost-center,Value=cc123}]'

Example Example: Create a volume and apply a tag

The following create-volume command creates a volume and applies two tags: purpose=production
and cost-center=cc123.

aws ec2 create-volume \

1320
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with tags using the command line

--availability-zone us-east-1a \
--volume-type gp2 \
--size 80 \
--tag-specifications 'ResourceType=volume,Tags=[{Key=purpose,Value=production},
{Key=cost-center,Value=cc123}]'

Add tags to an existing resource

The following examples demonstrate how to add tags to an existing resource using the create-tags
command.

Example Example: Add a tag to a resource

The following command adds the tag Stack=production to the specified image, or overwrites an
existing tag for the AMI where the tag key is Stack. If the command succeeds, no output is returned.

aws ec2 create-tags \

--resources ami-78a54011 \
--tags Key=Stack,Value=production

Example Example: Add tags to multiple resources

This example adds (or overwrites) two tags for an AMI and an instance. One of the tags contains just a
key (webserver), with no value (we set the value to an empty string). The other tag consists of a key
(stack) and value (Production). If the command succeeds, no output is returned.

aws ec2 create-tags \

--resources ami-1a2b3c4d i-1234567890abcdef0 \
--tags Key=webserver,Value= Key=stack,Value=Production

Example Example: Add tags with special characters

This example adds the tag [Group]=test to an instance. The square brackets ([ and ]) are special
characters, which must be escaped.

If you are using Linux or OS X, to escape the special characters, enclose the element with the special
character with double quotes ("), and then enclose the entire key and value structure with single quotes
(').

aws ec2 create-tags \

--resources i-1234567890abcdef0 \
--tags 'Key="[Group]",Value=test'

If you are using Windows, to escape the special characters, enclose the element that has special
characters with double quotes ("), and then precede each double quote character with a backslash (\) as
follows:

aws ec2 create-tags ^

--resources i-1234567890abcdef0 ^
--tags Key=\"[Group]\",Value=test

If you are using Windows PowerShell, to escape the special characters, enclose the value that has special
characters with double quotes ("), precede each double quote character with a backslash (\), and then
enclose the entire key and value structure with single quotes (') as follows:

aws ec2 create-tags `

1321
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Add tags to a resource using CloudFormation

--resources i-1234567890abcdef0 `
--tags 'Key=\"[Group]\",Value=test'

Describe tagged resources

The following examples show you how to use filters with the describe-instances to view instances with
specific tags. All EC2 describe commands use this syntax to filter by tag across a single resource type.
Alternatively, you can use the describe-tags command to filter by tag across EC2 resource types.

Example Example: Describe instances with the specified tag key

The following command describes the instances with a Stack tag, regardless of the value of the tag.

aws ec2 describe-instances \

--filters Name=tag-key,Values=Stack

Example Example: Describe instances with the specified tag

The following command describes the instances with the tag Stack=production.

aws ec2 describe-instances \

--filters Name=tag:Stack,Values=production

Example Example: Describe instances with the specified tag value

The following command describes the instances with a tag with the value production, regardless of the
tag key.

aws ec2 describe-instances \

--filters Name=tag-value,Values=production

Example Example: Describe all EC2 resources with the specified tag
The following command describes all EC2 resources with the tag Stack=Test.

aws ec2 describe-tags \

--filters Name=key,Values=Stack Name=value,Values=Test

Add tags to a resource using CloudFormation

With Amazon EC2 resource types, you specify tags using either a Tags or TagSpecifications
property.

The following examples add the tag Stack=Production to AWS::EC2::Instance using its Tags property.

Example Example: Tags in YAML

Tags:
- Key: "Stack"
Value: "Production"

Example Example: Tags in JSON

"Tags": [
{

1322
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Service quotas

"Key": "Stack",
"Value": "Production"
}
]

The following examples add the tag Stack=Production to AWS::EC2::LaunchTemplate

LaunchTemplateData using its TagSpecifications property.

Example Example: TagSpecifications in YAML

TagSpecifications:
- ResourceType: "instance"
Tags:
- Key: "Stack"
Value: "Production"

Example Example: TagSpecifications in JSON

"TagSpecifications": [
{
"ResourceType": "instance",
"Tags": [
{
"Key": "Stack",
"Value": "Production"
}
]
}
]

Amazon EC2 service quotas

Amazon EC2 provides different resources that you can use. These resources include images, instances,
volumes, and snapshots. When you create your AWS account, we set default quotas (also referred to as
limits) on these resources on a per-Region basis. For example, there is a maximum number of instances
that you can launch in a Region. So if you were to launch an instance in the US West (Oregon) Region, for
example, the request must not cause your usage to exceed your maximum number of instances in that
Region.

The Amazon EC2 console provides limit information for the resources managed by the Amazon EC2 and
Amazon VPC consoles. You can request an increase for many of these limits. Use the limit information
that we provide to manage your AWS infrastructure. Plan to request any limit increases in advance of the
time that you'll need them.

For more information, see Amazon EC2 endpoints and quotas in the Amazon Web Services General
Reference. For information about Amazon EBS quotas, see Amazon EBS quotas (p. 1268).

View your current limits

Use the Limits page in the Amazon EC2 console to view the current limits for resources provided by
Amazon EC2 and Amazon VPC, on a per-Region basis.

To view your current limits

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select a Region.

1323
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
View your current limits

3. From the navigation pane, choose Limits.

4. Locate the resource in the list. You can use the search fields to filter the list by resource name or
resource group. The Current limit column displays the current maximum for the resource for your
account.

1324
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Request an increase

Request an increase
Use the Limits page in the Amazon EC2 console to request an increase in your Amazon EC2 or Amazon
VPC resources, on a per-Region basis.

Alternatively, request an increase using Service Quotas. For more information, see Requesting a quota
increase in the Service Quotas User Guide.

To request an increase using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. From the navigation bar, select a Region.
3. From the navigation pane, choose Limits.
4. Select the resource in the list, and choose Request limit increase.
5. Complete the required fields on the limit increase form and choose Submit. We'll respond to you
using the contact method that you specified.

Limits on email sent using port 25

Amazon EC2 restricts traffic on port 25 of all instances by default. You can request that this restriction be
removed. For more information, see How do I remove the restriction on port 25 from my EC2 instance? in
the AWS Knowledge Center.

Amazon EC2 usage reports

AWS provides a free reporting tool called AWS Cost Explorer that enables you to analyze the cost and
usage of your EC2 instances and the usage of your Reserved Instances. You can view data up to the last
13 months, and forecast how much you are likely to spend for the next three months. You can use Cost
Explorer to see patterns in how much you spend on AWS resources over time, identify areas that need
further inquiry, and see trends that you can use to understand your costs. You also can specify time
ranges for the data, and view time data by day or by month.

Here's an example of some of the questions that you can answer when using Cost Explorer:

• How much am I spending on instances of each instance type?

• How many instance hours are being used by a particular department?
• How is my instance usage distributed across Availability Zones?
• How is my instance usage distributed across AWS accounts?
• How well am I using my Reserved Instances?
• Are my Reserved Instances helping me save money?

For more information about working with reports in Cost Explorer, including saving reports, see
Analyzing your costs with Cost Explorer.

1325
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Troubleshoot launch issues

Troubleshoot EC2 instances

The following documentation can help you troubleshoot problems that you might have with your
instance.

Contents
• Troubleshoot instance launch issues (p. 1326)
• Troubleshoot connecting to your instance (p. 1329)
• Troubleshoot stopping your instance (p. 1339)
• Troubleshoot instance termination (shutting down) (p. 1342)
• Troubleshoot instances with failed status checks (p. 1342)
• Troubleshoot an unreachable instance (p. 1363)
• Boot from the wrong volume (p. 1366)
• Use EC2Rescue for Linux (p. 1367)
• Send a diagnostic interrupt (for advanced users) (p. 1377)

For additional help with Windows instances, see Troubleshooting Windows Instances in the Amazon EC2
User Guide for Windows Instances.

Troubleshoot instance launch issues

The following issues prevent you from launching an instance.

Launch Issues
• Instance limit exceeded (p. 1326)
• Insufficient instance capacity (p. 1327)
• The requested configuration is currently not supported. Please check the documentation for
supported configurations. (p. 1327)
• Instance terminates immediately (p. 1328)

Instance limit exceeded

Description
You get the InstanceLimitExceeded error when you try to launch a new instance or restart a stopped
instance.

Cause
If you get an InstanceLimitExceeded error when you try to launch a new instance or restart a
stopped instance, you have reached the limit on the number of instances that you can launch in a Region.
When you create your AWS account, we set default limits on the number of instances you can run on a
per-Region basis.

1326
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Insufficient instance capacity

Solution
You can request an instance limit increase on a per-region basis. For more information, see Amazon EC2
service quotas (p. 1323).

Insufficient instance capacity

Description
You get the InsufficientInstanceCapacity error when you try to launch a new instance or restart
a stopped instance.

Cause
If you get this error when you try to launch an instance or restart a stopped instance, AWS does not
currently have enough available On-Demand capacity to fulfill your request.

Solution
To resolve the issue, try the following:

• Wait a few minutes and then submit your request again; capacity can shift frequently.
• Submit a new request with a reduced number of instances. For example, if you're making a single
request to launch 15 instances, try making 3 requests for 5 instances, or 15 requests for 1 instance
instead.
• If you're launching an instance, submit a new request without specifying an Availability Zone.
• If you're launching an instance, submit a new request using a different instance type (which you can
resize at a later stage). For more information, see Change the instance type (p. 311).
• If you are launching instances into a cluster placement group, you can get an insufficient capacity
error. For more information, see Placement group rules and limitations (p. 930).

The requested configuration is currently not

supported. Please check the documentation for
supported configurations.
Description
You get the Unsupported error when you try to launch a new instance because the instance
configuration is not supported.

Cause
The error message provides additional details. For example, an instance type or instance purchasing
option might not be supported in the specified Region or Availability Zone.

Solution
Try a different instance configuration. To search for an instance type that meets your requirements, see
Find an Amazon EC2 instance type (p. 310).

1327
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance terminates immediately

Instance terminates immediately

Description
Your instance goes from the pending state to the terminated state.

Cause
The following are a few reasons why an instance might immediately terminate:

• You've exceeded your EBS volume limits. For more information, see Instance volume limits (p. 1291).
• An EBS snapshot is corrupted.
• The root EBS volume is encrypted and you do not have permissions to access the CMK for decryption.
• A snapshot specified in the block device mapping for the AMI is encrypted and you do not have
permissions to access the CMK for decryption or you do not have access to the CMK to encrypt the
restored volumes.
• The instance store-backed AMI that you used to launch the instance is missing a required part (an
image.part.xx file).

For more information, get the termination reason using one of the following methods.

To get the termination reason using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and select the instance.
3. On the first tab, find the reason next to State transition reason.

To get the termination reason using the AWS Command Line Interface

1. Use the describe-instances command and specify the instance ID.

aws ec2 describe-instances --instance-id instance_id

2. Review the JSON response returned by the command and note the values in the StateReason
response element.

The following code block shows an example of a StateReason response element.

"StateReason": {
"Message": "Client.VolumeLimitExceeded: Volume limit exceeded",
"Code": "Server.InternalError"
},

To get the termination reason using AWS CloudTrail

For more information, see Viewing events with CloudTrail event history in the AWS CloudTrail User Guide.

Solution
Depending on the termination reason, take one of the following actions:

1328
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Connect to your instance

• Client.VolumeLimitExceeded: Volume limit exceeded — Delete unused volumes. You can

submit a request to increase your volume limit.
• Client.InternalError: Client error on launch — Ensure that you have the permissions
required to access the CMKs used to decrypt and encrypt volumes. For more information, see Using key
policies in AWS KMS in the AWS Key Management Service Developer Guide.

Troubleshoot connecting to your instance

The following information can help you troubleshoot issues with connecting to your instance. For
additional help with Windows instances, see Troubleshooting Windows Instances in the Amazon EC2 User
Guide for Windows Instances.

Connection problems and errors

• Common causes for connection issues (p. 1329)
• Error connecting to your instance: Connection timed out (p. 1330)
• Error: unable to load key ... Expecting: ANY PRIVATE KEY (p. 1333)
• Error: User key not recognized by server (p. 1333)
• Error: Permission denied or connection closed by [instance] port 22 (p. 1334)
• Error: Unprotected private key file (p. 1336)
• Error: Private key must begin with "-----BEGIN RSA PRIVATE KEY-----" and end with "-----END RSA
PRIVATE KEY-----" (p. 1337)
• Error: Server refused our key or No supported authentication methods available (p. 1337)
• Cannot ping instance (p. 1337)
• Error: Server unexpectedly closed network connection (p. 1338)
• Error: Host key validation failed for EC2 Instance Connect (p. 1338)

Common causes for connection issues

We recommend that you begin troubleshooting by checking some common causes for issues connecting
to your instance.

Verify the user name for your instance

You can connect to your instance using the user name for your user account or the default user
name for the AMI that you used to launch your instance.
• Get the user name for your user account.

For more information about how to create a user account, see Manage user accounts on your
Amazon Linux instance (p. 656).
• Get the default user name for the AMI that you used to launch your instance:
• For Amazon Linux 2 or the Amazon Linux AMI, the user name is ec2-user.
• For a CentOS AMI, the user name is centos.
• For a Debian AMI, the user name is admin.
• For a Fedora AMI, the user name is ec2-user or fedora.
• For a RHEL AMI, the user name is ec2-user or root.
• For a SUSE AMI, the user name is ec2-user or root.
• For an Ubuntu AMI, the user name is ubuntu.

1329
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Error connecting to your instance: Connection timed out

• Otherwise, if ec2-user and root don't work, check with the AMI provider.
Verify that your security group rules allow traffic

Make sure your security group rules allow inbound traffic from your public IPv4 address on
the proper port. For steps to verify, see Error connecting to your instance: Connection timed
out (p. 1330)
Verify that your instance is ready

After you launch an instance, it can take a few minutes for the instance to be ready so that you can
connect to it. Check your instance to make sure it is running and has passed its status checks.

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and then select your instance.
3. Verify the following:

a. In the Instance state column, verify that your instance is in the running state.
b. In the Status check column, verify that your instance has passed the two status checks.

Verify the general prerequisites for connecting to your instance

For more information, see General prerequisites for connecting to your instance (p. 598).

Error connecting to your instance: Connection timed

out
If you try to connect to your instance and get an error message Network error: Connection timed
out or Error connecting to [instance], reason: -> Connection timed out: connect,
try the following:

Check your security group rules.

You need a security group rule that allows inbound traffic from your public IPv4 address on the proper
port.

New console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and then select your instance.
3. On the Security tab at the bottom of the console page, under Inbound rules, check the list of
rules that are in effect for the selected instance.

• For Linux instances: Verify that there is a rule that allows traffic from your computer to port
22 (SSH).
• For Windows instances: Verify that there is a rule that allows traffic from your computer to
port 3389 (RDP).
4. Each time you restart your instance, a new IP address (and host name) will be assigned. If your
security group has a rule that allows inbound traffic from a single IP address, this address
might not be static if your computer is on a corporate network or if you are connecting through
an internet service provider (ISP). Instead, specify the range of IP addresses used by client
computers. If your security group does not have a rule that allows inbound traffic as described
in the previous step, add a rule to your security group. For more information, see Authorize
inbound traffic for your Linux instances (p. 1041).

1330
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Error connecting to your instance: Connection timed out

For more information about security group rules, see Security group rules in the Amazon VPC
User Guide.

Old console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and then select your instance.
3. In the Description tab at the bottom of the console page, next to Security groups, select view
inbound rules to display the list of rules that are in effect for the selected instance.
4. For Linux instances: When you select view inbound rules, a window will appear that displays
the port(s) to which traffic is allowed. Verify that there is a rule that allows traffic from your
computer to port 22 (SSH).

For Windows instances: When you select view inbound rules, a window will appear that displays
the port(s) to which traffic is allowed. Verify that there is a rule that allows traffic from your
computer to port 3389 (RDP).

Each time you restart your instance, a new IP address (and host name) will be assigned. If your
security group has a rule that allows inbound traffic from a single IP address, this address
may not be static if your computer is on a corporate network or if you are connecting through
an internet service provider (ISP). Instead, specify the range of IP addresses used by client
computers. If your security group does not have a rule that allows inbound traffic as described
in the previous step, add a rule to your security group. For more information, see Authorize
inbound traffic for your Linux instances (p. 1041).

For more information about security group rules, see Security group rules in the Amazon VPC
User Guide.

Check the route table for the subnet.

You need a route that sends all traffic destined outside the VPC to the internet gateway for the VPC.

New console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and then select your instance.
3. On the Networking tab, make note of the values for VPC ID and Subnet ID.
4. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.
5. In the navigation pane, choose Internet Gateways. Verify that there is an internet gateway
attached to your VPC. Otherwise, choose Create internet gateway, enter a name for the
internet gateway, and choose Create internet gateway. Then, for the internet gateway you
created, choose Actions, Attach to VPC, select your VPC, and then choose Attach internet
gateway to attach it to your VPC.
6. In the navigation pane, choose Subnets, and then select your subnet.
7. On the Route table tab, verify that there is a route with 0.0.0.0/0 as the destination and the
internet gateway for your VPC as the target. If you're connecting to your instance using its IPv6
address, verify that there is a route for all IPv6 traffic (::/0) that points to the internet gateway.
Otherwise, do the following:

a. Choose the ID of the route table (rtb-xxxxxxxx) to navigate to the route table.
b. On the Routes tab, choose Edit routes. Choose Add route, use 0.0.0.0/0 as the
destination and the internet gateway as the target. For IPv6, choose Add route, use ::/0 as
the destination and the internet gateway as the target.

1331
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Error connecting to your instance: Connection timed out

c. Choose Save routes.

Old console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and then select your instance.
3. In the Description tab, write down the values of VPC ID and Subnet ID.
4. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.
5. In the navigation pane, choose Internet Gateways. Verify that there is an internet gateway
attached to your VPC. Otherwise, choose Create Internet Gateway to create an internet
gateway. Select the internet gateway, and then choose Attach to VPC and follow the directions
to attach it to your VPC.
6. In the navigation pane, choose Subnets, and then select your subnet.
7. On the Route Table tab, verify that there is a route with 0.0.0.0/0 as the destination and the
internet gateway for your VPC as the target. If you're connecting to your instance using its IPv6
address, verify that there is a route for all IPv6 traffic (::/0) that points to the internet gateway.
Otherwise, do the following:

a. Choose the ID of the route table (rtb-xxxxxxxx) to navigate to the route table.
b. On the Routes tab, choose Edit routes. Choose Add route, use 0.0.0.0/0 as the
destination and the internet gateway as the target. For IPv6, choose Add route, use ::/0 as
the destination and the internet gateway as the target.
c. Choose Save routes.

Check the network access control list (ACL) for the subnet.

The network ACLs must allow inbound and outbound traffic from your local IP address on the proper
port. The default network ACL allows all inbound and outbound traffic.

1. Open the Amazon VPC console at https://console.aws.amazon.com/vpc/.

2. In the navigation pane, choose Subnets and select your subnet.
3. On the Description tab, find Network ACL, and choose its ID (acl-xxxxxxxx).
4. Select the network ACL. For Inbound Rules, verify that the rules allow traffic from your computer.
Otherwise, delete or modify the rule that is blocking traffic from your computer.
5. For Outbound Rules, verify that the rules allow traffic to your computer. Otherwise, delete or
modify the rule that is blocking traffic to your computer.

If your computer is on a corporate network

Ask your network administrator whether the internal firewall allows inbound and outbound traffic from
your computer on port 22 (for Linux instances) or port 3389 (for Windows instances).

If you have a firewall on your computer, verify that it allows inbound and outbound traffic from your
computer on port 22 (for Linux instances) or port 3389 (for Windows instances).

Check that your instance has a public IPv4 address.

If not, you can associate an Elastic IP address with your instance. For more information, see Elastic IP
addresses (p. 831).

Check the CPU load on your instance; the server may be overloaded.

1332
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Error: unable to load key ... Expecting: ANY PRIVATE KEY

AWS automatically provides data such as Amazon CloudWatch metrics and instance status, which you
can use to see how much CPU load is on your instance and, if necessary, adjust how your loads are
handled. For more information, see Monitor your instances using CloudWatch (p. 761).

• If your load is variable, you can automatically scale your instances up or down using Auto Scaling and
Elastic Load Balancing.
• If your load is steadily growing, you can move to a larger instance type. For more information, see
Change the instance type (p. 311).

To connect to your instance using an IPv6 address, check the following:

• Your subnet must be associated with a route table that has a route for IPv6 traffic (::/0) to an
internet gateway.
• Your security group rules must allow inbound traffic from your local IPv6 address on the proper port
(22 for Linux and 3389 for Windows).
• Your network ACL rules must allow inbound and outbound IPv6 traffic.
• If you launched your instance from an older AMI, it might not be configured for DHCPv6 (IPv6
addresses are not automatically recognized on the network interface). For more information, see
Configure IPv6 on Your Instances in the Amazon VPC User Guide.
• Your local computer must have an IPv6 address, and must be configured to use IPv6.

Error: unable to load key ... Expecting: ANY PRIVATE

KEY
If you try to connect to your instance and get the error message, unable to load key ...
Expecting: ANY PRIVATE KEY, the file in which the private key is stored is incorrectly configured.
If the private key file ends in .pem, it might still be incorrectly configured. A possible cause for an
incorrectly configured private key file is a missing certificate.

If the private key file is incorrectly configured, follow these steps to resolve the error

1. Create a new key pair. For more information, see Option 1: Create a key pair using Amazon
EC2 (p. 1045).
2. Add the new key pair to your instance. For more information, see Connect to your Linux instance if
you lose your private key (p. 1053).
3. Connect to your instance using the new key pair.

Error: User key not recognized by server

If you use SSH to connect to your instance

• Use ssh -vvv to get triple verbose debugging information while connecting:

ssh -vvv -i path/my-key-pair.pem my-instance-user-

[email protected]

The following sample output demonstrates what you might see if you were trying to connect to your
instance with a key that was not recognized by the server:

open/ANT/myusername/.ssh/known_hosts).

1333
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Error: Permission denied or connection
closed by [instance] port 22
debug2: bits set: 504/1024
debug1: ssh_rsa_verify: signature correct
debug2: kex_derive_keys
debug2: set_newkeys: mode 1
debug1: SSH2_MSG_NEWKEYS sent
debug1: expecting SSH2_MSG_NEWKEYS
debug2: set_newkeys: mode 0
debug1: SSH2_MSG_NEWKEYS received
debug1: Roaming not allowed by server
debug1: SSH2_MSG_SERVICE_REQUEST sent
debug2: service_accept: ssh-userauth
debug1: SSH2_MSG_SERVICE_ACCEPT received
debug2: key: boguspem.pem ((nil))
debug1: Authentications that can continue: publickey
debug3: start over, passed a different list publickey
debug3: preferred gssapi-keyex,gssapi-with-mic,publickey,keyboard-interactive,password
debug3: authmethod_lookup publickey
debug3: remaining preferred: keyboard-interactive,password
debug3: authmethod_is_enabled publickey
debug1: Next authentication method: publickey
debug1: Trying private key: boguspem.pem
debug1: read PEM private key done: type RSA
debug3: sign_and_send_pubkey: RSA 9c:4c:bc:0c:d0:5c:c7:92:6c:8e:9b:16:e4:43:d8:b2
debug2: we sent a publickey packet, wait for reply
debug1: Authentications that can continue: publickey
debug2: we did not send a packet, disable method
debug1: No more authentication methods to try.
Permission denied (publickey).

If you use PuTTY to connect to your instance

• Verify that your private key (.pem) file has been converted to the format recognized by PuTTY (.ppk).
For more information about converting your private key, see Connect to your Linux instance from
Windows using PuTTY (p. 614).
Note
In PuTTYgen, load your private key file and select Save Private Key rather than Generate.
• Verify that you are connecting with the appropriate user name for your AMI. Enter the user name in
the Host name box in the PuTTY Configuration window.
• For Amazon Linux 2 or the Amazon Linux AMI, the user name is ec2-user.
• For a CentOS AMI, the user name is centos.
• For a Debian AMI, the user name is admin.
• For a Fedora AMI, the user name is ec2-user or fedora.
• For a RHEL AMI, the user name is ec2-user or root.
• For a SUSE AMI, the user name is ec2-user or root.
• For an Ubuntu AMI, the user name is ubuntu.
• Otherwise, if ec2-user and root don't work, check with the AMI provider.
• Verify that you have an inbound security group rule to allow inbound traffic to the appropriate port.
For more information, see Authorizing Network Access to Your Instances (p. 1041).

Error: Permission denied or connection closed by

[instance] port 22
If you connect to your instance using SSH and get any of the following errors, Host key not found
in [directory], Permission denied (publickey), Authentication failed, permission

1334
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Error: Permission denied or connection
closed by [instance] port 22
denied, or Connection closed by [instance] port 22, verify that you are connecting with the
appropriate user name for your AMI and that you have specified the proper private key (.pem) file for
your instance.

The appropriate user names are as follows:

• For Amazon Linux 2 or the Amazon Linux AMI, the user name is ec2-user.
• For a CentOS AMI, the user name is centos.
• For a Debian AMI, the user name is admin.
• For a Fedora AMI, the user name is ec2-user or fedora.
• For a RHEL AMI, the user name is ec2-user or root.
• For a SUSE AMI, the user name is ec2-user or root.
• For an Ubuntu AMI, the user name is ubuntu.
• Otherwise, if ec2-user and root don't work, check with the AMI provider.

For example, to use an SSH client to connect to an Amazon Linux instance, use the following command:

ssh -i /path/my-key-pair.pem [email protected]

Confirm that you are using the private key file that corresponds to the key pair that you selected when
you launched the instance.

New console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and then select your instance.
3. On the Details tab, under Instance details, verify the value of Key pair name.
4. If you did not specify a key pair when you launched the instance, you can terminate the instance
and launch a new instance, ensuring that you specify a key pair. If this is an instance that you
have been using but you no longer have the .pem file for your key pair, you can replace the key
pair with a new one. For more information, see Connect to your Linux instance if you lose your
private key (p. 1053).

Old console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and then select your instance.
3. In the Description tab, verify the value of Key pair name.
4. If you did not specify a key pair when you launched the instance, you can terminate the instance
and launch a new instance, ensuring that you specify a key pair. If this is an instance that you
have been using but you no longer have the .pem file for your key pair, you can replace the key
pair with a new one. For more information, see Connect to your Linux instance if you lose your
private key (p. 1053).

If you generated your own key pair, ensure that your key generator is set up to create RSA keys. DSA keys
are not accepted.

If you get a Permission denied (publickey) error and none of the above applies (for example,
you were able to connect previously), the permissions on the home directory of your instance may have
been changed. Permissions for /home/my-instance-user-name/.ssh/authorized_keys must be
limited to the owner only.

1335
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Error: Unprotected private key file

To verify the permissions on your instance

1. Stop your instance and detach the root volume. For more information, see Stop and start your
instance (p. 623) and Detach an Amazon EBS volume from a Linux instance (p. 1123).
2. Launch a temporary instance in the same Availability Zone as your current instance (use a similar or
the same AMI as you used for your current instance), and attach the root volume to the temporary
instance. For more information, see Attach an Amazon EBS volume to an instance (p. 1106).
3. Connect to the temporary instance, create a mount point, and mount the volume that you attached.
For more information, see Make an Amazon EBS volume available for use on Linux (p. 1111).
4. From the temporary instance, check the permissions of the /home/my-instance-user-name/
directory of the attached volume. If necessary, adjust the permissions as follows:

[ec2-user ~]$ chmod 600 mount_point/home/my-instance-user-name/.ssh/authorized_keys

[ec2-user ~]$ chmod 700 mount_point/home/my-instance-user-name/.ssh

[ec2-user ~]$ chmod 700 mount_point/home/my-instance-user-name

5. Unmount the volume, detach it from the temporary instance, and re-attach it to the original
instance. Ensure that you specify the correct device name for the root volume; for example, /dev/
xvda.
6. Start your instance. If you no longer require the temporary instance, you can terminate it.

Error: Unprotected private key file

Your private key file must be protected from read and write operations from any other users. If your
private key can be read or written to by anyone but you, then SSH ignores your key and you see the
following warning message below.

@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@ WARNING: UNPROTECTED PRIVATE KEY FILE! @
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
Permissions 0777 for '.ssh/my_private_key.pem' are too open.
It is required that your private key files are NOT accessible by others.
This private key will be ignored.
bad permissions: ignore key: .ssh/my_private_key.pem
Permission denied (publickey).

If you see a similar message when you try to log in to your instance, examine the first line of the error
message to verify that you are using the correct public key for your instance. The above example uses the
private key .ssh/my_private_key.pem with file permissions of 0777, which allow anyone to read or
write to this file. This permission level is very insecure, and so SSH ignores this key. To fix the error, run
the following command, substituting the path for your private key file.

[ec2-user ~]$ chmod 0400 .ssh/my_private_key.pem

1336
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Error: Private key must begin with "-----BEGIN RSA PRIVATE
KEY-----" and end with "-----END RSA PRIVATE KEY-----"

Error: Private key must begin with "-----BEGIN RSA

PRIVATE KEY-----" and end with "-----END RSA
PRIVATE KEY-----"
If you use a third-party tool, such as ssh-keygen, to create an RSA key pair, it generates the private key in
the OpenSSH key format. When you connect to your instance, if you use the private key in the OpenSSH
format to decrypt the password, you'll get the error Private key must begin with "-----BEGIN
RSA PRIVATE KEY-----" and end with "-----END RSA PRIVATE KEY-----".

To resolve the error, the private key must be in the PEM format. Use the following command to create
the private key in the PEM format:

ssh-keygen -m PEM

Error: Server refused our key or No supported

authentication methods available
If you use PuTTY to connect to your instance and get either of the following errors, Error: Server refused
our key or Error: No supported authentication methods available, verify that you are connecting with
the appropriate user name for your AMI. Type the user name in User name in the PuTTY Configuration
window.

The appropriate user names are as follows:

• For Amazon Linux 2 or the Amazon Linux AMI, the user name is ec2-user.
• For a CentOS AMI, the user name is centos.
• For a Debian AMI, the user name is admin.
• For a Fedora AMI, the user name is ec2-user or fedora.
• For a RHEL AMI, the user name is ec2-user or root.
• For a SUSE AMI, the user name is ec2-user or root.
• For an Ubuntu AMI, the user name is ubuntu.
• Otherwise, if ec2-user and root don't work, check with the AMI provider.

You should also verify that your private key (.pem) file has been correctly converted to the format
recognized by PuTTY (.ppk). For more information about converting your private key, see Connect to
your Linux instance from Windows using PuTTY (p. 614).

Cannot ping instance

The ping command is a type of ICMP traffic — if you are unable to ping your instance, ensure that your
inbound security group rules allow ICMP traffic for the Echo Request message from all sources, or from
the computer or instance from which you are issuing the command.

If you are unable to issue a ping command from your instance, ensure that your outbound security
group rules allow ICMP traffic for the Echo Request message to all destinations, or to the host that you
are attempting to ping.

Ping commands can also be blocked by a firewall or time out due to network latency or hardware issues.
You should consult your local network or system administrator for help with further troubleshooting.

1337
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Error: Server unexpectedly closed network connection

Error: Server unexpectedly closed network

connection
If you are connecting to your instance with PuTTY and you receive the error "Server unexpectedly closed
network connection," verify that you have enabled keepalives on the Connection page of the PuTTY
Configuration to avoid being disconnected. Some servers disconnect clients when they do not receive
any data within a specified period of time. Set the Seconds between keepalives to 59 seconds.

If you still experience issues after enabling keepalives, try to disable Nagle's algorithm on the Connection
page of the PuTTY Configuration.

Error: Host key validation failed for EC2 Instance

Connect
If you rotate your instance host keys, the new host keys are not automatically uploaded to the AWS
trusted host keys database. This causes host key validation to fail when you try to connect to your
instance using the EC2 Instance Connect browser-based client, and you're unable to connect to your
instance.

To resolve the error, you must run the eic_harvest_hostkeys script on your instance, which uploads
your new host key to EC2 Instance Connect. The script is located at /opt/aws/bin/ on Amazon Linux 2
instances, and at /usr/share/ec2-instance-connect/ on Ubuntu instances.

Amazon Linux 2

To resolve the host key validation failed error on an Amazon Linux 2 instance

1. Connect to your instance using SSH.

You can connect by using the EC2 Instance Connect CLI or by using the SSH key pair that was
assigned to your instance when you launched it and the default user name of the AMI that you
used to launch your instance. For Amazon Linux 2, the default user name is ec2-user.

For example, if your instance was launched using Amazon Linux 2, your instance's public
DNS name is ec2-a-b-c-d.us-west-2.compute.amazonaws.com, and the key pair is
my_ec2_private_key.pem, use the following command to SSH into your instance:

$ ssh -i my_ec2_private_key.pem [email protected]

west-2.compute.amazonaws.com

For more information about connecting to your instance, see Connect to your Linux instance
using SSH (p. 601).
2. Navigate to the following folder.

[ec2-user ~]$ cd /opt/aws/bin/

3. Run the following command on your instance.

[ec2-user ~]$ ./eic_harvest_hostkeys

Note that a successful call results in no output.

You can now use the EC2 Instance Connect browser-based client to connect to your instance.

1338
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Stop your instance

Ubuntu

To resolve the host key validation failed error on an Ubuntu instance

1. Connect to your instance using SSH.

You can connect by using the EC2 Instance Connect CLI or by using the SSH key pair that was
assigned to your instance when you launched it and the default user name of the AMI that you
used to launch your instance. For Ubuntu, the default user name is ubuntu.

For example, if your instance was launched using Ubuntu, your instance's public DNS
name is ec2-a-b-c-d.us-west-2.compute.amazonaws.com, and the key pair is
my_ec2_private_key.pem, use the following command to SSH into your instance:

$ ssh -i my_ec2_private_key.pem [email protected]

For more information about connecting to your instance, see Connect to your Linux instance
using SSH (p. 601).
2. Navigate to the following folder.

[ec2-user ~]$ cd /usr/share/ec2-instance-connect/

3. Run the following command on your instance.

[ec2-user ~]$ ./eic_harvest_hostkeys

Note that a successful call results in no output.

You can now use the EC2 Instance Connect browser-based client to connect to your instance.

Troubleshoot stopping your instance

If you have stopped your Amazon EBS-backed instance and it appears stuck in the stopping state, there
may be an issue with the underlying host computer.

There is no cost for instance usage while an instance is in the stopping state or in any other state
except running. You are only charged for instance usage when an instance is in the running state.

Force stop the instance

Force the instance to stop using either the console or the AWS CLI.

New console

To force stop the instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and select the stuck instance.
3. Choose Instance state, Force stop instance, Stop.

Old console

To force stop the instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

1339
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create a replacement instance

2. In the navigation pane, choose Instances and select the stuck instance.
3. Choose Instance State, Stop, Yes, Forcefully Stop.

AWS CLI

To force stop the instance using the AWS CLI

Use the stop-instances command and the --force option as follows:

aws ec2 stop-instances --instance-ids i-0123ab456c789d01e --force

If, after 10 minutes, the instance has not stopped, post a request for help in the Amazon EC2 forum.
To help expedite a resolution, include the instance ID, and describe the steps that you've already taken.
Alternatively, if you have a support plan, create a technical support case in the Support Center.

Create a replacement instance

To attempt to resolve the problem while you are waiting for assistance from the Amazon EC2 forum or
the Support Center, create a replacement instance. Create an AMI of the stuck instance, and launch a
new instance using the new AMI.

New console

To create a replacement instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and select the stuck instance.
3. Choose Actions, Image and templates, Create image.
4. On the Create image page, do the following:

a. Enter a name and description for the AMI.

b. Choose No reboot.
c. Choose Create image.

For more information, see Create a Linux AMI from an instance (p. 126).
5. Launch a new instance from the AMI and verify that the new instance is working.
6. Select the stuck instance, and choose Actions, Instance state, Terminate instance. If the
instance also gets stuck terminating, Amazon EC2 automatically forces it to terminate within a
few hours.

Old console

To create a replacement instance using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances and select the stuck instance.
3. Choose Actions, Image, Create Image.
4. In the Create Image dialog box, fill in the following fields, and then choose Create Image:

a. Specify a name and description for the AMI.

1340
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Create a replacement instance

b. Choose No reboot.

For more information, see Create a Linux AMI from an instance (p. 126).
5. Launch a new instance from the AMI and verify that the new instance is working.
6. Select the stuck instance, and choose Actions, Instance State, Terminate. If the instance also
gets stuck terminating, Amazon EC2 automatically forces it to terminate within a few hours.

AWS CLI

To create a replacement instance using the CLI

1. Create an AMI from the stuck instance using the create-image (AWS CLI) command and the --
no-reboot option as follows:.

aws ec2 create-image --instance-id i-0123ab456c789d01e --name "AMI" --

description "AMI for replacement instance" --no-reboot

2. Launch a new instance from the AMI using the run-instances (AWS CLI) command as follows:

aws ec2 run-instances --image-id ami-1a2b3c4d --count 1 --instance-type c3.large --

key-name MyKeyPair --security-groups MySecurityGroup

3. Verify that the new instance is working.

4. Terminate the stuck instance using the terminate-instances (AWS CLI) command as follows:

aws ec2 terminate-instances --instance-ids i-1234567890abcdef0

If you are unable to create an AMI from the instance as described in the previous procedure, you can set
up a replacement instance as follows:

(Alternate) To create a replacement instance using the console

1. Select the instance and choose Description, Block devices. Select each volume and make note of its
volume ID. Be sure to note which volume is the root volume.
2. In the navigation pane, choose Volumes. Select each volume for the instance, and choose Actions,
Create Snapshot.
3. In the navigation pane, choose Snapshots. Select the snapshot that you just created, and choose
Actions, Create Volume.
4. Launch an instance with the same operating system as the stuck instance. Note the volume ID and
device name of its root volume.
5. In the navigation pane, choose Instances, select the instance that you just launched, and choose
Instance state, Stop instance.
6. In the navigation pane, choose Volumes, select the root volume of the stopped instance, and choose
Actions, Detach Volume.
7. Select the root volume that you created from the stuck instance, choose Actions, Attach Volume,
and attach it to the new instance as its root volume (using the device name that you made note of).
Attach any additional non-root volumes to the instance.
8. In the navigation pane, choose Instances and select the replacement instance. Choose Instance
state, Start instance. Verify that the instance is working.
9. Select the stuck instance, choose Instance state, Terminate instance. If the instance also gets stuck
terminating, Amazon EC2 automatically forces it to terminate within a few hours.

1341
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Terminate your instance

Troubleshoot instance termination (shutting down)

You are not billed for any instance usage while an instance is not in the running state. In other words,
when you terminate an instance, you stop incurring charges for that instance as soon as its state changes
to shutting-down.

Delayed instance termination

If your instance remains in the shutting-down state longer than a few minutes, it might be delayed
due to shutdown scripts being run by the instance.

Another possible cause is a problem with the underlying host computer. If your instance remains in the
shutting-down state for several hours, Amazon EC2 treats it as a stuck instance and forcibly terminates
it.

If it appears that your instance is stuck terminating and it has been longer than several hours, post a
request for help to the Amazon EC2 forum. To help expedite a resolution, include the instance ID and
describe the steps that you've already taken. Alternatively, if you have a support plan, create a technical
support case in the Support Center.

Terminated instance still displayed

After you terminate an instance, it remains visible for a short while before being deleted. The state
shows as terminated. If the entry is not deleted after several hours, contact Support.

Instances automatically launched or terminated

Generally, the following behaviors mean that you've used Amazon EC2 Auto Scaling, EC2 Fleet, or Spot
Fleet to scale your computing resources automatically based on criteria that you've defined:

• You terminate an instance and a new instance launches automatically.

• You launch an instance and one of your instances terminates automatically.
• You stop an instance and it terminates and a new instance launches automatically.

To stop automatic scaling, see the Amazon EC2 Auto Scaling User Guide, Launch instances using an EC2
Fleet (p. 548), or Create a Spot Fleet request (p. 407).

Troubleshoot instances with failed status checks

The following information can help you troubleshoot issues if your instance fails a status check. First
determine whether your applications are exhibiting any problems. If you verify that the instance is not
running your applications as expected, review the status check information and the system logs.

Contents
• Review status check information (p. 1343)
• Retrieve the system logs (p. 1344)
• Troubleshoot system log errors for Linux-based instances (p. 1344)
• Out of memory: kill process (p. 1345)
• ERROR: mmu_update failed (Memory management update failed) (p. 1346)

1342
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Review status check information

• I/O error (block device failure) (p. 1346)

• I/O ERROR: neither local nor remote disk (Broken distributed block device) (p. 1347)
• request_module: runaway loop modprobe (Looping legacy kernel modprobe on older Linux
versions) (p. 1348)
• "FATAL: kernel too old" and "fsck: No such file or directory while trying to open /dev" (Kernel and AMI
mismatch) (p. 1349)
• "FATAL: Could not load /lib/modules" or "BusyBox" (Missing kernel modules) (p. 1350)
• ERROR Invalid kernel (EC2 incompatible kernel) (p. 1351)
• fsck: No such file or directory while trying to open... (File system not found) (p. 1352)
• General error mounting filesystems (failed mount) (p. 1353)
• VFS: Unable to mount root fs on unknown-block (Root filesystem mismatch) (p. 1355)
• Error: Unable to determine major/minor number of root device... (Root file system/device
mismatch) (p. 1356)
• XENBUS: Device with no driver... (p. 1357)
• ... days without being checked, check forced (File system check required) (p. 1358)
• fsck died with exit status... (Missing device) (p. 1358)
• GRUB prompt (grubdom>) (p. 1359)
• Bringing up interface eth0: Device eth0 has different MAC address than expected, ignoring. (Hard-
coded MAC address) (p. 1361)
• Unable to load SELinux Policy. Machine is in enforcing mode. Halting now. (SELinux
misconfiguration) (p. 1362)
• XENBUS: Timeout connecting to devices (Xenbus timeout) (p. 1363)

Review status check information

To investigate impaired instances using the Amazon EC2 console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and then select your instance.
3. In the details pane, choose Status Checks to see the individual results for all System Status Checks
and Instance Status Checks.

If a system status check has failed, you can try one of the following options:

• Create an instance recovery alarm. For more information, see Create alarms that stop, terminate,
reboot, or recover an instance (p. 784).
• If you changed the instance type to an instance built on the Nitro System (p. 209), status checks fail
if you migrated from an instance that does not have the required ENA and NVMe drivers. For more
information, see Compatibility for resizing instances (p. 311).
• For an instance using an Amazon EBS-backed AMI, stop and restart the instance.
• For an instance using an instance-store backed AMI, terminate the instance and launch a replacement.
• Wait for Amazon EC2 to resolve the issue.
• Post your issue to the Amazon EC2 forum.
• If your instance is in an Auto Scaling group, the Amazon EC2 Auto Scaling service automatically
launches a replacement instance. For more information, see Health Checks for Auto Scaling Instances
in the Amazon EC2 Auto Scaling User Guide.
• Retrieve the system log and look for errors.

1343
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Retrieve the system logs

Retrieve the system logs

If an instance status check fails, you can reboot the instance and retrieve the system logs. The logs may
reveal an error that can help you troubleshoot the issue. Rebooting clears unnecessary information from
the logs.

To reboot an instance and retrieve the system log

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the navigation pane, choose Instances, and select your instance.
3. Choose Instance state, Reboot instance. It might take a few minutes for your instance to reboot.
4. Verify that the problem still exists; in some cases, rebooting may resolve the problem.
5. When the instance is in the running state, choose Actions, Monitor and troubleshoot, Get system
log.
6. Review the log that appears on the screen, and use the list of known system log error statements
below to troubleshoot your issue.
7. If your experience differs from our check results, or if you are having an issue with your instance that
our checks did not detect, choose Submit feedback on the Status Checks tab to help us improve our
detection tests.
8. If your issue is not resolved, you can post your issue to the Amazon EC2 forum.

Troubleshoot system log errors for Linux-based

instances
For Linux-based instances that have failed an instance status check, such as the instance reachability
check, verify that you followed the steps above to retrieve the system log. The following list contains
some common system log errors and suggested actions you can take to resolve the issue for each error.

Memory Errors

• Out of memory: kill process (p. 1345)

• ERROR: mmu_update failed (Memory management update failed) (p. 1346)

Device Errors

• I/O error (block device failure) (p. 1346)

• I/O ERROR: neither local nor remote disk (Broken distributed block device) (p. 1347)

Kernel Errors

• request_module: runaway loop modprobe (Looping legacy kernel modprobe on older Linux
versions) (p. 1348)
• "FATAL: kernel too old" and "fsck: No such file or directory while trying to open /dev" (Kernel and AMI
mismatch) (p. 1349)
• "FATAL: Could not load /lib/modules" or "BusyBox" (Missing kernel modules) (p. 1350)
• ERROR Invalid kernel (EC2 incompatible kernel) (p. 1351)

File System Errors

1344
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Out of memory: kill process

• fsck: No such file or directory while trying to open... (File system not found) (p. 1352)
• General error mounting filesystems (failed mount) (p. 1353)
• VFS: Unable to mount root fs on unknown-block (Root filesystem mismatch) (p. 1355)
• Error: Unable to determine major/minor number of root device... (Root file system/device
mismatch) (p. 1356)
• XENBUS: Device with no driver... (p. 1357)
• ... days without being checked, check forced (File system check required) (p. 1358)
• fsck died with exit status... (Missing device) (p. 1358)

Operating System Errors

• GRUB prompt (grubdom>) (p. 1359)

• Bringing up interface eth0: Device eth0 has different MAC address than expected, ignoring. (Hard-
coded MAC address) (p. 1361)
• Unable to load SELinux Policy. Machine is in enforcing mode. Halting now. (SELinux
misconfiguration) (p. 1362)
• XENBUS: Timeout connecting to devices (Xenbus timeout) (p. 1363)

Out of memory: kill process

An out-of-memory error is indicated by a system log entry similar to the one shown below.

[115879.769795] Out of memory: kill process 20273 (httpd) score 1285879

or a child
[115879.769795] Killed process 1917 (php-cgi) vsz:467184kB, anon-
rss:101196kB, file-rss:204kB

Potential cause
Exhausted memory

Suggested actions

For this instance type Do this

Amazon EBS-backed Do one of the following:

• Stop the instance, and modify the instance

to use a different instance type, and start
the instance again. For example, a larger or a
memory-optimized instance type.
• Reboot the instance to return it to a non-
impaired status. The problem will probably
occur again unless you change the instance
type.

Instance store-backed Do one of the following:

• Terminate the instance and launch a new

instance, specifying a different instance type.
For example, a larger or a memory-optimized
instance type.

1345
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ERROR: mmu_update failed (Memory
management update failed)
For this instance type Do this
• Reboot the instance to return it to an
unimpaired status. The problem will probably
occur again unless you change the instance
type.

ERROR: mmu_update failed (Memory management

update failed)
Memory management update failures are indicated by a system log entry similar to the following:

...
Press `ESC' to enter the menu... 0 [H[J Booting 'Amazon Linux 2011.09
(2.6.35.14-95.38.amzn1.i686)'

root (hd0)

Filesystem type is ext2fs, using whole disk

kernel /boot/vmlinuz-2.6.35.14-95.38.amzn1.i686 root=LABEL=/ console=hvc0 LANG=

en_US.UTF-8 KEYTABLE=us

initrd /boot/initramfs-2.6.35.14-95.38.amzn1.i686.img

ERROR: mmu_update failed with rc=-22

Potential cause
Issue with Amazon Linux

Suggested action
Post your issue to the Developer Forums or contact AWS Support.

I/O error (block device failure)

An input/output error is indicated by a system log entry similar to the following example:

[9943662.053217] end_request: I/O error, dev sde, sector 52428288

[9943664.191262] end_request: I/O error, dev sde, sector 52428168
[9943664.191285] Buffer I/O error on device md0, logical block 209713024
[9943664.191297] Buffer I/O error on device md0, logical block 209713025
[9943664.191304] Buffer I/O error on device md0, logical block 209713026
[9943664.191310] Buffer I/O error on device md0, logical block 209713027
[9943664.191317] Buffer I/O error on device md0, logical block 209713028
[9943664.191324] Buffer I/O error on device md0, logical block 209713029
[9943664.191332] Buffer I/O error on device md0, logical block 209713030
[9943664.191339] Buffer I/O error on device md0, logical block 209713031
[9943664.191581] end_request: I/O error, dev sde, sector 52428280
[9943664.191590] Buffer I/O error on device md0, logical block 209713136
[9943664.191597] Buffer I/O error on device md0, logical block 209713137
[9943664.191767] end_request: I/O error, dev sde, sector 52428288
[9943664.191970] end_request: I/O error, dev sde, sector 52428288
[9943664.192143] end_request: I/O error, dev sde, sector 52428288

1346
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
I/O ERROR: neither local nor remote
disk (Broken distributed block device)
[9943664.192949] end_request: I/O error, dev sde, sector 52428288
[9943664.193112] end_request: I/O error, dev sde, sector 52428288
[9943664.193266] end_request: I/O error, dev sde, sector 52428288
...

Potential causes

Instance type Potential cause

Amazon EBS-backed A failed Amazon EBS volume

Instance store-backed A failed physical drive

Suggested actions

For this instance type Do this

Amazon EBS-backed Use the following procedure:

1. Stop the instance.

2. Detach the volume.
3. Attempt to recover the volume.
Note
It's good practice to snapshot your
Amazon EBS volumes often. This
dramatically decreases the risk of data
loss as a result of failure.
4. Re-attach the volume to the instance.
5. Start the instance.

Instance store-backed Terminate the instance and launch a new instance.

Note
Data cannot be recovered. Recover from
backups.
Note
It's a good practice to use either Amazon
S3 or Amazon EBS for backups. Instance
store volumes are directly tied to single
host and single disk failures.

I/O ERROR: neither local nor remote disk (Broken

distributed block device)
An input/output error on the device is indicated by a system log entry similar to the following example:

...
block drbd1: Local IO failed in request_timer_fn. Detaching...

Aborting journal on device drbd1-8.

1347
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
request_module: runaway loop modprobe (Looping
legacy kernel modprobe on older Linux versions)
block drbd1: IO ERROR: neither local nor remote disk

Buffer I/O error on device drbd1, logical block 557056

lost page write due to I/O error on drbd1

JBD2: I/O error detected when updating journal superblock for drbd1-8.

Potential causes

Instance type Potential cause

Amazon EBS-backed A failed Amazon EBS volume

Instance store-backed A failed physical drive

Suggested action
Terminate the instance and launch a new instance.

For an Amazon EBS-backed instance you can recover data from a recent snapshot by creating an image
from it. Any data added after the snapshot cannot be recovered.

request_module: runaway loop modprobe (Looping

legacy kernel modprobe on older Linux versions)
This condition is indicated by a system log similar to the one shown below. Using an unstable or old
Linux kernel (for example, 2.6.16-xenU) can cause an interminable loop condition at startup.

Linux version 2.6.16-xenU ([email protected]) (gcc version 4.0.1

20050727 (Red Hat 4.0.1-5)) #1 SMP Mon May 28 03:41:49 SAST 2007

BIOS-provided physical RAM map:

Xen: 0000000000000000 - 0000000026700000 (usable)

0MB HIGHMEM available.

...

request_module: runaway loop modprobe binfmt-464c

request_module: runaway loop modprobe binfmt-464c

request_module: runaway loop modprobe binfmt-464c

request_module: runaway loop modprobe binfmt-464c

request_module: runaway loop modprobe binfmt-464c

Suggested actions

For this instance type Do this

Amazon EBS-backed Use a newer kernel, either GRUB-based or static,

using one of the following options:

1348
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
"FATAL: kernel too old" and "fsck: No such file or directory
while trying to open /dev" (Kernel and AMI mismatch)
For this instance type Do this
Option 1: Terminate the instance and launch
a new instance, specifying the –kernel and –
ramdisk parameters.

Option 2:

1. Stop the instance.

2. Modify the kernel and ramdisk attributes to use
a newer kernel.
3. Start the instance.

Instance store-backed Terminate the instance and launch a new

instance, specifying the –kernel and –ramdisk
parameters.

"FATAL: kernel too old" and "fsck: No such file or

directory while trying to open /dev" (Kernel and AMI
mismatch)
This condition is indicated by a system log similar to the one shown below.

Linux version 2.6.16.33-xenU ([email protected])

(gcc version 4.1.1 20070105 (Red Hat 4.1.1-52)) #2 SMP Wed Aug 15 17:27:36 SAST 2007
...
FATAL: kernel too old
Kernel panic - not syncing: Attempted to kill init!

Potential causes
Incompatible kernel and userland

Suggested actions

For this instance type Do this

Amazon EBS-backed Use the following procedure:

1. Stop the instance.

2. Modify the configuration to use a newer kernel.
3. Start the instance.

Instance store-backed Use the following procedure:

1. Create an AMI that uses a newer kernel.

2. Terminate the instance.
3. Start a new instance from the AMI you created.

1349
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
"FATAL: Could not load /lib/modules"
or "BusyBox" (Missing kernel modules)

"FATAL: Could not load /lib/modules" or

"BusyBox" (Missing kernel modules)
This condition is indicated by a system log similar to the one shown below.

[ 0.370415] Freeing unused kernel memory: 1716k freed

Loading, please wait...
WARNING: Couldn't open directory /lib/modules/2.6.34-4-virtual: No such file or directory
FATAL: Could not open /lib/modules/2.6.34-4-virtual/modules.dep.temp for writing: No such
file or directory
FATAL: Could not load /lib/modules/2.6.34-4-virtual/modules.dep: No such file or directory
Couldn't get a file descriptor referring to the console
Begin: Loading essential drivers... ...
FATAL: Could not load /lib/modules/2.6.34-4-virtual/modules.dep: No such file or directory
FATAL: Could not load /lib/modules/2.6.34-4-virtual/modules.dep: No such file or directory
Done.
Begin: Running /scripts/init-premount ...
Done.
Begin: Mounting root file system... ...
Begin: Running /scripts/local-top ...
Done.
Begin: Waiting for root file system... ...
Done.
Gave up waiting for root device. Common problems:
- Boot args (cat /proc/cmdline)
- Check rootdelay= (did the system wait long enough?)
- Check root= (did the system wait for the right device?)
- Missing modules (cat /proc/modules; ls /dev)
FATAL: Could not load /lib/modules/2.6.34-4-virtual/modules.dep: No such file or directory
FATAL: Could not load /lib/modules/2.6.34-4-virtual/modules.dep: No such file or directory
ALERT! /dev/sda1 does not exist. Dropping to a shell!

BusyBox v1.13.3 (Ubuntu 1:1.13.3-1ubuntu5) built-in shell (ash)

Enter 'help' for a list of built-in commands.

(initramfs)

Potential causes
One or more of the following conditions can cause this problem:

• Missing ramdisk
• Missing correct modules from ramdisk
• Amazon EBS root volume not correctly attached as /dev/sda1

Suggested actions

For this instance type Do this

Amazon EBS-backed Use the following procedure:

1. Select corrected ramdisk for the Amazon EBS

volume.
2. Stop the instance.
3. Detach the volume and repair it.

1350
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
ERROR Invalid kernel (EC2 incompatible kernel)

For this instance type Do this

4. Attach the volume to the instance.
5. Start the instance.
6. Modify the AMI to use the corrected ramdisk.

Instance store-backed Use the following procedure:

1. Terminate the instance and launch a new

instance with the correct ramdisk.
2. Create a new AMI with the correct ramdisk.

ERROR Invalid kernel (EC2 incompatible kernel)

This condition is indicated by a system log similar to the one shown below.

...
root (hd0)

Filesystem type is ext2fs, using whole disk

kernel /vmlinuz root=/dev/sda1 ro

initrd /initrd.img

ERROR Invalid kernel: elf_xen_note_check: ERROR: Will only load images

built for the generic loader or Linux images
xc_dom_parse_image returned -1

Error 9: Unknown boot failure

Booting 'Fallback'

root (hd0)

Filesystem type is ext2fs, using whole disk

kernel /vmlinuz.old root=/dev/sda1 ro

Error 15: File not found

Potential causes
One or both of the following conditions can cause this problem:

• Supplied kernel is not supported by GRUB

• Fallback kernel does not exist

Suggested actions

For this instance type Do this

Amazon EBS-backed Use the following procedure:

1. Stop the instance.

1351
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
fsck: No such file or directory while
trying to open... (File system not found)
For this instance type Do this
2. Replace with working kernel.
3. Install a fallback kernel.
4. Modify the AMI by correcting the kernel.

Instance store-backed Use the following procedure:

1. Terminate the instance and launch a new

instance with the correct kernel.
2. Create an AMI with the correct kernel.
3. (Optional) Seek technical assistance for data
recovery using AWS Support.

fsck: No such file or directory while trying to open...

(File system not found)
This condition is indicated by a system log similar to the one shown below.

Welcome to Fedora
Press 'I' to enter interactive startup.
Setting clock : Wed Oct 26 05:52:05 EDT 2011 [ OK ]

Starting udev: [ OK ]

Setting hostname localhost: [ OK ]

No devices found
Setting up Logical Volume Management: File descriptor 7 left open
No volume groups found
[ OK ]

Checking filesystems
Checking all file systems.
[/sbin/fsck.ext3 (1) -- /] fsck.ext3 -a /dev/sda1
/dev/sda1: clean, 82081/1310720 files, 2141116/2621440 blocks
[/sbin/fsck.ext3 (1) -- /mnt/dbbackups] fsck.ext3 -a /dev/sdh
fsck.ext3: No such file or directory while trying to open /dev/sdh

/dev/sdh:
The superblock could not be read or does not describe a correct ext2
filesystem. If the device is valid and it really contains an ext2
filesystem (and not swap or ufs or something else), then the superblock
is corrupt, and you might try running e2fsck with an alternate superblock:
e2fsck -b 8193 <device>

[FAILED]

*** An error occurred during the file system check.

*** Dropping you to a shell; the system will reboot
*** when you leave the shell.
Give root password for maintenance
(or type Control-D to continue):

Potential causes
• A bug exists in ramdisk filesystem definitions /etc/fstab

1352
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General error mounting filesystems (failed mount)

• Misconfigured filesystem definitions in /etc/fstab

• Missing/failed drive

Suggested actions

For this instance type Do this

Amazon EBS-backed Use the following procedure:

1. Stop the instance, detach the root volume,

repair/modify /etc/fstab the volume, attach
the volume to the instance, and start the
instance.
2. Fix ramdisk to include modified /etc/fstab (if
applicable).
3. Modify the AMI to use a newer ramdisk.

The sixth field in the fstab defines availability

requirements of the mount – a nonzero value
implies that an fsck will be done on that volume
and must succeed. Using this field can be
problematic in Amazon EC2 because a failure
typically results in an interactive console prompt
that is not currently available in Amazon EC2.
Use care with this feature and read the Linux man
page for fstab.

Instance store-backed Use the following procedure:

1. Terminate the instance and launch a new

instance.
2. Detach any errant Amazon EBS volumes and
the reboot instance.
3. (Optional) Seek technical assistance for data
recovery using AWS Support.

General error mounting filesystems (failed mount)

This condition is indicated by a system log similar to the one shown below.

Loading xenblk.ko module

xen-vbd: registered block device major 8

Loading ehci-hcd.ko module

Loading ohci-hcd.ko module
Loading uhci-hcd.ko module
USB Universal Host Controller Interface driver v3.0

Loading mbcache.ko module

Loading jbd.ko module
Loading ext3.ko module
Creating root device.
Mounting root filesystem.
kjournald starting. Commit interval 5 seconds

1353
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
General error mounting filesystems (failed mount)

EXT3-fs: mounted filesystem with ordered data mode.

Setting up other filesystems.

Setting up new root fs
no fstab.sys, mounting internal defaults
Switching to new root and running init.
unmounting old /dev
unmounting old /proc
unmounting old /sys
mountall:/proc: unable to mount: Device or resource busy
mountall:/proc/self/mountinfo: No such file or directory
mountall: root filesystem isn't mounted
init: mountall main process (221) terminated with status 1

General error mounting filesystems.

A maintenance shell will now be started.
CONTROL-D will terminate this shell and re-try.
Press enter for maintenance
(or type Control-D to continue):

Potential causes

Instance type Potential cause

Amazon EBS-backed
• Detached or failed Amazon EBS volume.
• Corrupted filesystem.
• Mismatched ramdisk and AMI combination
(such as Debian ramdisk with a SUSE AMI).

Instance store-backed
• A failed drive.
• A corrupted file system.
• A mismatched ramdisk and combination (for
example, a Debian ramdisk with a SUSE AMI).

Suggested actions

For this instance type Do this

Amazon EBS-backed Use the following procedure:

1. Stop the instance.

2. Detach the root volume.
3. Attach the root volume to a known working
instance.
4. Run filesystem check (fsck –a /dev/...).
5. Fix any errors.
6. Detach the volume from the known working
instance.
7. Attach the volume to the stopped instance.
8. Start the instance.
9. Recheck the instance status.

1354
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
VFS: Unable to mount root fs on unknown-
block (Root filesystem mismatch)
For this instance type Do this

Instance store-backed Try one of the following:

• Start a new instance.

• (Optional) Seek technical assistance for data
recovery using AWS Support.

VFS: Unable to mount root fs on unknown-block

(Root filesystem mismatch)
This condition is indicated by a system log similar to the one shown below.

Linux version 2.6.16-xenU ([email protected]) (gcc version 4.0.1

20050727 (Red Hat 4.0.1-5)) #1 SMP Mon May 28 03:41:49 SAST 2007
...
Kernel command line: root=/dev/sda1 ro 4
...
Registering block device major 8
...
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(8,1)

Potential causes

Instance type Potential cause

Amazon EBS-backed • Device not attached correctly.

• Root device not attached at correct device
point.
• Filesystem not in expected format.
• Use of legacy kernel (such as 2.6.16-XenU).
• A recent kernel update on your instance (faulty
update, or an update bug)

Instance store-backed Hardware device failure.

Suggested actions

For this instance type Do this

Amazon EBS-backed Do one of the following:

• Stop and then restart the instance.

• Modify root volume to attach at the correct
device point, possible /dev/sda1 instead of /
dev/sda.
• Stop and modify to use modern kernel.
• Refer to the documentation for your Linux
distribution to check for known update bugs.
Change or reinstall the kernel.

1355
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Error: Unable to determine major/minor number
of root device... (Root file system/device mismatch)
For this instance type Do this

Instance store-backed Terminate the instance and launch a new instance

using a modern kernel.

Error: Unable to determine major/minor number of

root device... (Root file system/device mismatch)
This condition is indicated by a system log similar to the one shown below.

...
XENBUS: Device with no driver: device/vif/0
XENBUS: Device with no driver: device/vbd/2048
drivers/rtc/hctosys.c: unable to open rtc device (rtc0)
Initializing network drop monitor service
Freeing unused kernel memory: 508k freed
:: Starting udevd...
done.
:: Running Hook [udev]
:: Triggering uevents...<30>udevd[65]: starting version 173
done.
Waiting 10 seconds for device /dev/xvda1 ...
Root device '/dev/xvda1' doesn't exist. Attempting to create it.
ERROR: Unable to determine major/minor number of root device '/dev/xvda1'.
You are being dropped to a recovery shell
Type 'exit' to try and continue booting
sh: can't access tty; job control turned off
[ramfs /]#

Potential causes
• Missing or incorrectly configured virtual block device driver
• Device enumeration clash (sda versus xvda or sda instead of sda1)
• Incorrect choice of instance kernel

Suggested actions

For this instance type Do this

Amazon EBS-backed Use the following procedure:

1. Stop the instance.

2. Detach the volume.
3. Fix the device mapping problem.
4. Start the instance.
5. Modify the AMI to address device mapping
issues.

Instance store-backed Use the following procedure:

1. Create a new AMI with the appropriate fix (map

block device correctly).

1356
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
XENBUS: Device with no driver...

For this instance type Do this

2. Terminate the instance and launch a new
instance from the AMI you created.

XENBUS: Device with no driver...

This condition is indicated by a system log similar to the one shown below.

XENBUS: Device with no driver: device/vbd/2048

drivers/rtc/hctosys.c: unable to open rtc device (rtc0)
Initalizing network drop monitor service
Freeing unused kernel memory: 508k freed
:: Starting udevd...
done.
:: Running Hook [udev]
:: Triggering uevents...<30>udevd[65]: starting version 173
done.
Waiting 10 seconds for device /dev/xvda1 ...
Root device '/dev/xvda1' doesn't exist. Attempting to create it.
ERROR: Unable to determine major/minor number of root device '/dev/xvda1'.
You are being dropped to a recovery shell
Type 'exit' to try and continue booting
sh: can't access tty; job control turned off
[ramfs /]#

Potential causes
• Missing or incorrectly configured virtual block device driver
• Device enumeration clash (sda versus xvda)
• Incorrect choice of instance kernel

Suggested actions

For this instance type Do this

Amazon EBS-backed Use the following procedure:

1. Stop the instance.

2. Detach the volume.
3. Fix the device mapping problem.
4. Start the instance.
5. Modify the AMI to address device mapping
issues.

Instance store-backed Use the following procedure:

1. Create an AMI with the appropriate fix (map

block device correctly).
2. Terminate the instance and launch a new
instance using the AMI you created.

1357
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
... days without being checked, check
forced (File system check required)

... days without being checked, check forced (File

system check required)
This condition is indicated by a system log similar to the one shown below.

...
Checking filesystems
Checking all file systems.
[/sbin/fsck.ext3 (1) -- /] fsck.ext3 -a /dev/sda1
/dev/sda1 has gone 361 days without being checked, check forced

Potential causes
Filesystem check time passed; a filesystem check is being forced.

Suggested actions
• Wait until the filesystem check completes. A filesystem check can take a long time depending on the
size of the root filesystem.
• Modify your filesystems to remove the filesystem check (fsck) enforcement using tune2fs or tools
appropriate for your filesystem.

fsck died with exit status... (Missing device)

This condition is indicated by a system log similar to the one shown below.

Cleaning up ifupdown....
Loading kernel modules...done.
...
Activating lvm and md swap...done.
Checking file systems...fsck from util-linux-ng 2.16.2
/sbin/fsck.xfs: /dev/sdh does not exist
fsck died with exit status 8
[31mfailed (code 8).[39;49m

Potential causes
• Ramdisk looking for missing drive
• Filesystem consistency check forced
• Drive failed or detached

Suggested actions

For this instance type Do this

Amazon EBS-backed Try one or more of the following to resolve the

issue:

• Stop the instance, attach the volume to an

existing running instance.

1358
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
GRUB prompt (grubdom>)

For this instance type Do this

• Manually run consistency checks.
• Fix ramdisk to include relevant utilities.
• Modify filesystem tuning parameters to remove
consistency requirements (not recommended).

Instance store-backed Try one or more of the following to resolve the

issue:

• Rebundle ramdisk with correct tooling.

• Modify file system tuning parameters to remove
consistency requirements (not recommended).
• Terminate the instance and launch a new
instance.
• (Optional) Seek technical assistance for data
recovery using AWS Support.

GRUB prompt (grubdom>)

This condition is indicated by a system log similar to the one shown below.

GNU GRUB version 0.97 (629760K lower / 0K upper memory)

[ Minimal BASH-like line editing is supported. For

the first word, TAB lists possible command

completions. Anywhere else TAB lists the possible

completions of a device/filename. ]

grubdom>

Potential causes

Instance type Potential causes

Amazon EBS-backed • Missing GRUB configuration file.

• Incorrect GRUB image used, expecting GRUB
configuration file at a different location.
• Unsupported filesystem used to store
your GRUB configuration file (for example,
converting your root file system to a type that is
not supported by an earlier version of GRUB).

Instance store-backed • Missing GRUB configuration file.

• Incorrect GRUB image used, expecting GRUB
configuration file at a different location.
• Unsupported filesystem used to store
your GRUB configuration file (for example,
converting your root file system to a type that is
not supported by an earlier version of GRUB).

1359
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
GRUB prompt (grubdom>)

Suggested actions

For this instance type Do this

Amazon EBS-backed Option 1: Modify the AMI and relaunch the

instance:

1. Modify the source AMI to create a GRUB

configuration file at the standard location (/
boot/grub/menu.lst).
2. Verify that your version of GRUB supports the
underlying file system type and upgrade GRUB
if necessary.
3. Pick the appropriate GRUB image, (hd0-1st
drive or hd00 – 1st drive, 1st partition).
4. Terminate the instance and launch a new one
using the AMI that you created.

Option 2: Fix the existing instance:

1. Stop the instance.

2. Detach the root filesystem.
3. Attach the root filesystem to a known working
instance.
4. Mount filesystem.
5. Create a GRUB configuration file.
6. Verify that your version of GRUB supports the
underlying file system type and upgrade GRUB
if necessary.
7. Detach filesystem.
8. Attach to the original instance.
9. Modify kernel attribute to use the appropriate
GRUB image (1st disk or 1st partition on 1st
disk).
10.Start the instance.

Instance store-backed Option 1: Modify the AMI and relaunch the

instance:

1. Create the new AMI with a GRUB configuration

file at the standard location (/boot/grub/
menu.lst).
2. Pick the appropriate GRUB image, (hd0-1st
drive or hd00 – 1st drive, 1st partition).
3. Verify that your version of GRUB supports the
underlying file system type and upgrade GRUB
if necessary.
4. Terminate the instance and launch a new
instance using the AMI you created.

Option 2: Terminate the instance and launch a

new instance, specifying the correct kernel.

1360
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Bringing up interface eth0: Device eth0
has different MAC address than expected,
For this instance type ignoring. (Hard-coded MAC Doaddress)
this
Note
To recover data from the existing
instance, contact AWS Support.

Bringing up interface eth0: Device eth0 has different

MAC address than expected, ignoring. (Hard-coded
MAC address)
This condition is indicated by a system log similar to the one shown below.

...
Bringing up loopback interface: [ OK ]

Bringing up interface eth0: Device eth0 has different MAC address than expected, ignoring.
[FAILED]

Starting auditd: [ OK ]

Potential causes
There is a hardcoded interface MAC in the AMI configuration

Suggested actions

For this instance type Do this

Amazon EBS-backed Do one of the following:

• Modify the AMI to remove the hardcoding and

relaunch the instance.
• Modify the instance to remove the hardcoded
MAC address.

OR

Use the following procedure:

1. Stop the instance.

2. Detach the root volume.
3. Attach the volume to another instance and
modify the volume to remove the hardcoded
MAC address.
4. Attach the volume to the original instance.
5. Start the instance.

Instance store-backed Do one of the following:

• Modify the instance to remove the hardcoded

MAC address.

1361
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Unable to load SELinux Policy. Machine is in enforcing
mode. Halting now. (SELinux misconfiguration)
For this instance type Do this
• Terminate the instance and launch a new
instance.

Unable to load SELinux Policy. Machine is

in enforcing mode. Halting now. (SELinux
misconfiguration)
This condition is indicated by a system log similar to the one shown below.

audit(1313445102.626:2): enforcing=1 old_enforcing=0 auid=4294967295

Unable to load SELinux Policy. Machine is in enforcing mode. Halting now.
Kernel panic - not syncing: Attempted to kill init!

Potential causes
SELinux has been enabled in error:

• Supplied kernel is not supported by GRUB

• Fallback kernel does not exist

Suggested actions

For this instance type Do this

Amazon EBS-backed Use the following procedure:

1. Stop the failed instance.

2. Detach the failed instance's root volume.
3. Attach the root volume to another running
Linux instance (later referred to as a recovery
instance).
4. Connect to the recovery instance and mount
the failed instance's root volume.
5. Disable SELinux on the mounted root volume.
This process varies across Linux distributions;
for more information, consult your OS-specific
documentation.
Note
On some systems, you disable SELinux
by setting SELINUX=disabled in the
/mount_point/etc/sysconfig/
selinux file, where mount_point
is the location that you mounted the
volume on your recovery instance.
6. Unmount and detach the root volume from the
recovery instance and reattach it to the original
instance.

1362
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
XENBUS: Timeout connecting to devices (Xenbus timeout)

For this instance type Do this

7. Start the instance.

Instance store-backed Use the following procedure:

1. Terminate the instance and launch a new

instance.
2. (Optional) Seek technical assistance for data
recovery using AWS Support.

XENBUS: Timeout connecting to devices (Xenbus

timeout)
This condition is indicated by a system log similar to the one shown below.

Linux version 2.6.16-xenU ([email protected]) (gcc version 4.0.1

20050727 (Red Hat 4.0.1-5)) #1 SMP Mon May 28 03:41:49 SAST 2007
...
XENBUS: Timeout connecting to devices!
...
Kernel panic - not syncing: No init found. Try passing init= option to kernel.

Potential causes
• The block device is not connected to the instance
• This instance is using an old instance kernel

Suggested actions

For this instance type Do this

Amazon EBS-backed Do one of the following:

• Modify the AMI and instance to use a modern

kernel and relaunch the instance.
• Reboot the instance.

Instance store-backed Do one of the following:

• Terminate the instance.

• Modify the AMI to use a modern kernel, and
launch a new instance using this AMI.

Troubleshoot an unreachable instance

You can use the following methods to troubleshoot an unreachable instance.

Contents

1363
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance reboot

• Instance reboot (p. 1364)

• Instance console output (p. 1364)
• Capture a screenshot of an unreachable instance (p. 1365)
• Instance recovery when a host computer fails (p. 1366)

Instance reboot
The ability to reboot instances that are otherwise unreachable is valuable for both troubleshooting and
general instance management.

Just as you can reset a computer by pressing the reset button, you can reset EC2 instances using the
Amazon EC2 console, CLI, or API. For more information, see Reboot your instance (p. 639).
Warning
For Windows instances, this operation performs a hard reboot that might result in data
corruption.

Instance console output

Console output is a valuable tool for problem diagnosis. It is especially useful for troubleshooting
kernel problems and service configuration issues that could cause an instance to terminate or become
unreachable before its SSH daemon can be started.

For Linux/Unix, the instance console output displays the exact console output that would normally
be displayed on a physical monitor attached to a computer. The console output returns buffered
information that was posted shortly after an instance transition state (start, stop, reboot, and terminate).
The posted output is not continuously updated; only when it is likely to be of the most value.

For Windows instances, the instance console output includes the last three system event log errors.

You can optionally retrieve the latest serial console output at any time during the instance lifecycle. This
option is only supported on Instances built on the Nitro System (p. 209). It is not supported through the
Amazon EC2 console.
Note
Only the most recent 64 KB of posted output is stored, which is available for at least 1 hour
after the last posting.

Only the instance owner can access the console output. You can retrieve the console output for your
instances using the console or the command line.

To get console output using the console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the left navigation pane, choose Instances, and select the instance.
3. Choose Actions, Monitor and troubleshoot, Get system log.

To get console output using the command line

You can use one of the following commands. For more information about these command line interfaces,
see Access Amazon EC2 (p. 3).

• get-console-output (AWS CLI)

1364
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Capture a screenshot of an unreachable instance

• Get-EC2ConsoleOutput (AWS Tools for Windows PowerShell)

For more information about common system log errors, see Troubleshoot system log errors for Linux-
based instances (p. 1344).

Capture a screenshot of an unreachable instance

If you are unable to reach your instance via SSH or RDP, you can capture a screenshot of your instance
and view it as an image. The image can provide visibility as to the status of the instance, and allows
for quicker troubleshooting. You can generate screenshots while the instance is running or after it has
crashed. There is no data transfer cost for this screenshot. The image is generated in JPG format and is
no larger than 100 kb. This feature is not supported when the instance is using an NVIDIA GRID driver, is
on bare metal instances (instances of type *.metal), or is powered by Arm-based Graviton or Graviton 2
processors. This feature is available in the following Regions:

• US East (N. Virginia) Region

• US East (Ohio) Region
• US West (Oregon) Region
• US West (N. California) Region
• Europe (Ireland) Region
• Europe (Frankfurt) Region
• Asia Pacific (Tokyo) Region
• Asia Pacific (Seoul) Region
• Asia Pacific (Singapore) Region
• Asia Pacific (Sydney) Region)
• South America (São Paulo) Region)
• Asia Pacific (Mumbai) Region
• Canada (Central) Region)
• Europe (London) Region
• Europe (Paris) Region

To access the instance console

1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

2. In the left navigation pane, choose Instances.
3. Select the instance to capture.
4. Choose Actions, Monitor and troubleshoot.
5. Choose Get instance screenshot.

Right-click the image to download and save it.

To capture a screenshot using the command line

You can use one of the following commands. The returned content is base64-encoded. For more
information about these command line interfaces, see Access Amazon EC2 (p. 3).

• get-console-screenshot (AWS CLI)

• GetConsoleScreenshot (Amazon EC2 Query API)

1365
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Instance recovery when a host computer fails

Instance recovery when a host computer fails

If there is an unrecoverable issue with the hardware of an underlying host computer, AWS may schedule
an instance stop event. You are notified of such an event ahead of time by email.

To recover an Amazon EBS-backed instance running on a host computer that failed

1. Back up any important data on your instance store volumes to Amazon EBS or Amazon S3.
2. Stop the instance.
3. Start the instance.
4. Restore any important data.

For more information, see Stop and start your instance (p. 623).

To recover an instance store-backed instance running on a host computer that failed

1. Create an AMI from the instance.

2. Upload the image to Amazon S3.
3. Back up important data to Amazon EBS or Amazon S3.
4. Terminate the instance.
5. Launch a new instance from the AMI.
6. Restore any important data to the new instance.

For more information, see Create an instance store-backed Linux AMI (p. 130).

Boot from the wrong volume

In some situations, you may find that a volume other than the volume attached to /dev/xvda or /dev/
sda has become the root volume of your instance. This can happen when you have attached the root
volume of another instance, or a volume created from the snapshot of a root volume, to an instance with
an existing root volume.

This is due to how the initial ramdisk in Linux works. It chooses the volume defined as / in the /etc/
fstab, and in some distributions, this is determined by the label attached to the volume partition.
Specifically, you find that your /etc/fstab looks something like the following:

LABEL=/ / ext4 defaults,noatime 1 1

tmpfs /dev/shm tmpfs defaults 0 0
devpts /dev/pts devpts gid=5,mode=620 0 0
sysfs /sys sysfs defaults 0 0
proc /proc proc defaults 0 0

If you check the label of both volumes, you see that they both contain the / label:

[ec2-user ~]$ sudo e2label /dev/xvda1

/
[ec2-user ~]$ sudo e2label /dev/xvdf1
/

In this example, you could end up having /dev/xvdf1 become the root device that your instance boots
to after the initial ramdisk runs, instead of the /dev/xvda1 volume from which you had intended to

1366
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
EC2Rescue for Linux

boot. To solve this, use the same e2label command to change the label of the attached volume that you
do not want to boot from.

In some cases, specifying a UUID in /etc/fstab can resolve this. However, if both volumes come from
the same snapshot, or the secondary is created from a snapshot of the primary volume, they share a
UUID.

[ec2-user ~]$ sudo blkid

/dev/xvda1: LABEL="/" UUID=73947a77-ddbe-4dc7-bd8f-3fe0bc840778 TYPE="ext4"
PARTLABEL="Linux" PARTUUID=d55925ee-72c8-41e7-b514-7084e28f7334
/dev/xvdf1: LABEL="old/" UUID=73947a77-ddbe-4dc7-bd8f-3fe0bc840778 TYPE="ext4"
PARTLABEL="Linux" PARTUUID=d55925ee-72c8-41e7-b514-7084e28f7334

To change the label of an attached ext4 volume

1. Use the e2label command to change the label of the volume to something other than /.

[ec2-user ~]$ sudo e2label /dev/xvdf1 old/

2. Verify that the volume has the new label.

[ec2-user ~]$ sudo e2label /dev/xvdf1

old/

To change the label of an attached xfs volume

• Use the xfs_admin command to change the label of the volume to something other than /.

[ec2-user ~]$ sudo xfs_admin -L old/ /dev/xvdf1

writing all SBs
new label = "old/"

After changing the volume label as shown, you should be able to reboot the instance and have the
proper volume selected by the initial ramdisk when the instance boots.
Important
If you intend to detach the volume with the new label and return it to another instance to use
as the root volume, you must perform the above procedure again and change the volume label
back to its original value. Otherwise, the other instance does not boot because the ramdisk is
unable to find the volume with the label /.

Use EC2Rescue for Linux

EC2Rescue for Linux is an easy-to-use, open-source tool that can be run on an Amazon EC2 Linux
instance to diagnose and troubleshoot common issues using its library of over 100 modules. A few
generalized use cases for EC2Rescue for Linux include gathering syslog and package manager logs,
collecting resource utilization data, and diagnosing/remediating known problematic kernel parameters
and common OpenSSH issues.

If you are using a Windows instance, see EC2Rescue for Windows Server.

Contents

• Install EC2Rescue for Linux (p. 1368)

1367
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Install EC2Rescue for Linux

• Work with EC2Rescue for Linux (p. 1371)

• Develop EC2Rescue modules (p. 1373)

Install EC2Rescue for Linux

The EC2Rescue for Linux tool can be installed on an Amazon EC2 Linux instance that meets the following
prerequisites.

Prerequisites

• Supported operating systems:

• Amazon Linux 2
• Amazon Linux 2016.09+
• SUSE Linux Enterprise Server 12+
• RHEL 7+
• Ubuntu 16.04+
• Software requirements:
• Python 2.7.9+ or 3.2+

If your system has the required Python version, you can install the standard build. Otherwise, you can
install the bundled build, which includes a minimal copy of Python.

To install the standard build

1. From a working Linux instance, download the EC2Rescue for Linux tool:

curl -O https://s3.amazonaws.com/ec2rescuelinux/ec2rl.tgz

2. (Optional) Before proceeding, you can optionally verify the signature of the EC2Rescue for Linux
installation file. For more information, see (Optional) Verify the signature of EC2Rescue for
Linux (p. 1369).
3. Download the sha256 hash file:

curl -O https://s3.amazonaws.com/ec2rescuelinux/ec2rl.tgz.sha256

4. Verify the integrity of the tarball:

sha256sum -c ec2rl.tgz.sha256

5. Unpack the tarball:

tar -xvf ec2rl.tgz

6. Verify the installation by listing out the help file:

cd ec2rl-<version_number>
./ec2rl help

To install the bundled build

For a link to the download and a list of limitations, see EC2Rescue for Linux on github.

1368
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
(Optional) Verify the signature of EC2Rescue for Linux

(Optional) Verify the signature of EC2Rescue for

Linux
The following is the recommended process of verifying the validity of the EC2Rescue for Linux package
for Linux-based operating systems.

When you download an application from the internet, we recommend that you authenticate the identity
of the software publisher and check that the application has not been altered or corrupted after it was
published. This protects you from installing a version of the application that contains a virus or other
malicious code.

If, after running the steps in this topic, you determine that the software for EC2Rescue for Linux is
altered or corrupted, do not run the installation file. Instead, contact Amazon Web Services.

EC2Rescue for Linux files for Linux-based operating systems are signed using GnuPG, an open-source
implementation of the Pretty Good Privacy (OpenPGP) standard for secure digital signatures. GnuPG
(also known as GPG) provides authentication and integrity checking through a digital signature. AWS
publishes a public key and signatures that you can use to verify the downloaded EC2Rescue for Linux
package. For more information about PGP and GnuPG (GPG), see http://www.gnupg.org.

The first step is to establish trust with the software publisher. Download the public key of the software
publisher, check that the owner of the public key is who they claim to be, and then add the public key to
your keyring. Your keyring is a collection of known public keys. After you establish the authenticity of the
public key, you can use it to verify the signature of the application.

Tasks
• Install the GPG tools (p. 1369)
• Authenticate and import the public key (p. 1370)
• Verify the signature of the package (p. 1370)

Install the GPG tools

If your operating system is Linux or Unix, the GPG tools may already be installed. To test whether the
tools are installed on your system, enter gpg2 at a command prompt. If the GPG tools are installed,
you see a GPG command prompt. If the GPG tools are not installed, you see an error stating that the
command cannot be found. You can install the GnuPG package from a repository.

To install GPG tools on Debian-based Linux

• From a terminal, run the following command:

apt-get install gnupg2

To install GPG tools on Red Hat–based Linux

• From a terminal, run the following command:

yum install gnupg2

1369
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
(Optional) Verify the signature of EC2Rescue for Linux

Authenticate and import the public key

The next step in the process is to authenticate the EC2Rescue for Linux public key and add it as a trusted
key in your GPG keyring.

To authenticate and import the EC2Rescue for Linux public key

1. At a command prompt, use the following command to obtain a copy of our public GPG build key:

curl -O https://s3.amazonaws.com/ec2rescuelinux/ec2rl.key

2. At a command prompt in the directory where you saved ec2rl.key, use the following command to
import the EC2Rescue for Linux public key into your keyring:

gpg2 --import ec2rl.key

The command returns results similar to the following:

gpg: /home/ec2-user/.gnupg/trustdb.gpg: trustdb created

gpg: key 2FAE2A1C: public key "[email protected] <EC2 Rescue for Linux>" imported
gpg: Total number processed: 1
gpg: imported: 1 (RSA: 1)

Verify the signature of the package

After you've installed the GPG tools, authenticated and imported the EC2Rescue for Linux public key, and
verified that the EC2Rescue for Linux public key is trusted, you are ready to verify the signature of the
EC2Rescue for Linux installation script.

To verify the EC2Rescue for Linux installation script signature

1. At a command prompt, run the following command to download the signature file for the
installation script:

curl -O https://s3.amazonaws.com/ec2rescuelinux/ec2rl.tgz.sig

2. Verify the signature by running the following command at a command prompt in the directory
where you saved ec2rl.tgz.sig and the EC2Rescue for Linux installation file. Both files must be
present.

gpg2 --verify ./ec2rl.tgz.sig

The output should look something like the following:

gpg: Signature made Thu 12 Jul 2018 01:57:51 AM UTC using RSA key ID 6991ED45
gpg: Good signature from "[email protected] <EC2 Rescue for Linux>"
gpg: WARNING: This key is not certified with a trusted signature!
gpg: There is no indication that the signature belongs to the owner.
Primary key fingerprint: E528 BCC9 0DBF 5AFA 0F6C C36A F780 4843 2FAE 2A1C
Subkey fingerprint: 966B 0D27 85E9 AEEC 1146 7A9D 8851 1153 6991 ED45

If the output contains the phrase Good signature from "[email protected] <EC2
Rescue for Linux>", it means that the signature has successfully been verified, and you can
proceed to run the EC2Rescue for Linux installation script.

1370
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with EC2Rescue for Linux

If the output includes the phrase BAD signature, check whether you performed the procedure
correctly. If you continue to get this response, contact Amazon Web Services and do not run the
installation file that you downloaded previously.

The following are details about the warnings that you might see:

• WARNING: This key is not certified with a trusted signature! There is no indication that the
signature belongs to the owner. This refers to your personal level of trust in your belief that you
possess an authentic public key for EC2Rescue for Linux. In an ideal world, you would visit an Amazon
Web Services office and receive the key in person. However, more often you download it from a
website. In this case, the website is an Amazon Web Services website.
• gpg2: no ultimately trusted keys found. This means that the specific key is not "ultimately trusted" by
you (or by other people whom you trust).

For more information, see http://www.gnupg.org.

Work with EC2Rescue for Linux

The following are common tasks you can perform to get started using this tool.

Tasks
• Run EC2Rescue for Linux (p. 1371)
• Upload the results (p. 1372)
• Create backups (p. 1372)
• Get help (p. 1372)

Run EC2Rescue for Linux

You can run EC2Rescue for Linux as shown in the following examples.

Example Example: Run all modules

To run all modules, run EC2Rescue for Linux with no options:

./ec2rl run

Some modules require root access. If you are not a root user, use sudo to run these modules as follows:

sudo ./ec2rl run

Example Example: Run a specific module

To run only specific modules, use the --only-modules parameter:

./ec2rl run --only-modules=module_name --arguments

For example, this command runs the dig module to query the amazon.com domain:

./ec2rl run --only-modules=dig --domain=amazon.com

1371
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Work with EC2Rescue for Linux

Example Example: View the results

You can view the results in /var/tmp/ec2rl:

cat /var/tmp/ec2rl/logfile_location

For example, view the log file for the dig module:

cat /var/tmp/ec2rl/2017-05-11T15_39_21.893145/mod_out/run/dig.log

Upload the results

If AWS Support has requested the results or to share the results from an S3 bucket, upload them using
the EC2Rescue for Linux CLI tool. The output of the EC2Rescue for Linux commands should provide the
commands that you need to use.

Example Example: Upload results to AWS Support

./ec2rl upload --upload-directory=/var/tmp/ec2rl/2017-05-11T15_39_21.893145 --support-

url="URLProvidedByAWSSupport"

Example Example: Upload results to an S3 bucket

./ec2rl upload --upload-directory=/var/tmp/ec2rl/2017-05-11T15_39_21.893145 --presigned-

url="YourPresignedS3URL"

For more information about generating pre-signed URLs for Amazon S3, see Uploading Objects Using
Pre-Signed URLs.

Create backups
Create a backup for your instance, one or more volumes, or a specific device ID using the following
commands.

Example Example: Back up an instance using an Amazon Machine Image (AMI)

./ec2rl run --backup=ami

Example Example: Back up all volumes associated with the instance

./ec2rl run --backup=allvolumes

Example Example: Back up a specific volume

./ec2rl run --backup=volumeID

Get help
EC2Rescue for Linux includes a help file that gives you information and syntax for each available
command.

Example Example: Display the general help

./ec2rl help

1372
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Develop EC2Rescue modules

Example Example: List the available modules

./ec2rl list

Example Example: Display the help for a specific module

./ec2rl help module_name

For example, use the following command to show the help file for the dig module:

./ec2rl help dig

Develop EC2Rescue modules

Modules are written in YAML, a data serialization standard. A module's YAML file consists of a single
document, representing the module and its attributes.

Add module attributes

The following table lists the available module attributes.

Attribute Description

name The name of the module. The name should be less

than or equal to 18 characters in length.

version The version number of the module.

title A short, descriptive title for the module.

This value should be less than or equal to 50
characters in length.

helptext The extended description of the module. Each

line should be less than or equal to 75 characters
in length. If the module consumes arguments,
required or optional, include them in the helptext
value.

For example:

helptext: !!str |
Collect output from ps for system
analysis
Consumes --times= for number of times to
repeat
Consumes --period= for time period
between repetition

placement The stage in which the module should be run.

Supported values:

• prediagnostic
• run
• postdiagnostic

1373
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Develop EC2Rescue modules

Attribute Description

language The language that the module code is written in.

Supported values:

• bash
• python

Note
Python code must be compatible with
both Python 2.7.9+ and Python 3.2+.

remediation Indicates whether the module supports

remediation. Supported values are True or
False.

The module defaults to False if this is absent,

making it an optional attribute for those modules
that do not support remediation.

content The entirety of the script code.

constraint The name of the object containing the constraint

values.

domain A descriptor of how the module is grouped or

classified. The set of included modules uses the
following domains:

• application
• net
• os
• performance

class A descriptor of the type of task performed by the

module. The set of included modules uses the
following classes:

• collect (collects output from programs)

• diagnose (pass/fail based on a set of criteria)
• gather (copies files and writes to specific file)

distro The list of Linux distributions that this module

supports. The set of included modules uses the
following distributions:

• alami (Amazon Linux)

• rhel
• ubuntu
• suse

required The required arguments that the module is

consuming from the CLI options.

optional The optional arguments that the module can use.

1374
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Develop EC2Rescue modules

Attribute Description

software The software executables used in the module. This

attribute is intended to specify software that is
not installed by default. The EC2Rescue for Linux
logic ensures that these programs are present and
executable before running the module.

package The source software package for an executable.

This attribute is intended to provide extended
details on the package with the software,
including a URL for downloading or getting
further information.

sudo Indicates whether root access is required to run

the module.

You do not need to implement sudo checks in

the module script. If the value is true, then the
EC2Rescue for Linux logic only runs the module
when the executing user has root access.

perfimpact Indicates whether the module can have significant

performance impact upon the environment in
which it is run. If the value is true and the --
perfimpact=true argument is not present, then
the module is skipped.

parallelexclusive Specifies a program that requires mutual

exclusivity. For example, all modules specifying
"bpf" run in a serial manner.

Add environment variables

The following table lists the available environment variables.

Environment Variable Description

EC2RL_CALLPATH The path to ec2rl.py. This path can be used to

locate the lib directory and use vendored Python
modules.

EC2RL_WORKDIR The main tmp directory for the diagnostic tool.

Default value: /var/tmp/ec2rl.

EC2RL_RUNDIR The directory where all output is stored.

Default value: /var/tmp/ec2rl/

<date&timestamp>.

EC2RL_GATHEREDDIR The root directory for placing gathered module

data.

Default value:/var/tmp/ec2rl/
<date&timestamp>/mod_out/gathered/.

1375
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Develop EC2Rescue modules

Environment Variable Description

EC2RL_NET_DRIVER The driver in use for the first, alphabetically

ordered, non-virtual network interface on the
instance.

Examples:

• xen_netfront
• ixgbevf
• ena

EC2RL_SUDO True if EC2Rescue for Linux is running as root;

otherwise, false.

EC2RL_VIRT_TYPE The virtualization type as provided by the instance

metadata.

Examples:

• default-hvm
• default-paravirtual

EC2RL_INTERFACES An enumerated list of interfaces on the system.

The value is a string containing names, such
as eth0, eth1, etc. This is generated via the
functions.bash and is only available for
modules that have sourced it.

Use YAML syntax

The following should be noted when constructing your module YAML files:

• The triple hyphen (---) denotes the explicit start of a document.

• The !ec2rlcore.module.Module tag tells the YAML parser which constructor to call when creating
the object from the data stream. You can find the constructor inside the module.py file.
• The !!str tag tells the YAML parser to not attempt to determine the type of data, and instead
interpret the content as a string literal.
• The pipe character (|) tells the YAML parser that the value is a literal-style scalar. In this case, the
parser includes all whitespace. This is important for modules because indentation and newline
characters are kept.
• The YAML standard indent is two spaces, which can be seen in the following examples. Ensure that you
maintain standard indentation (for example, four spaces for Python) for your script and then indent
the entire content two spaces inside the module file.

Example modules
Example one (mod.d/ps.yaml):

--- !ec2rlcore.module.Module
# Module document. Translates directly into an almost-complete Module object
name: !!str ps
path: !!str
version: !!str 1.0
title: !!str Collect output from ps for system analysis

1376
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Send a diagnostic interrupt

helptext: !!str |
Collect output from ps for system analysis
Requires --times= for number of times to repeat
Requires --period= for time period between repetition
placement: !!str run
package:
- !!str
language: !!str bash
content: !!str |
#!/bin/bash
error_trap()
{
printf "%0.s=" {1..80}
echo -e "\nERROR: "$BASH_COMMAND" exited with an error on line ${BASH_LINENO[0]}"
exit 0
}
trap error_trap ERR

# read-in shared function

source functions.bash
echo "I will collect ps output from this $EC2RL_DISTRO box for $times times every $period
seconds."
for i in $(seq 1 $times); do
ps auxww
sleep $period
done
constraint:
requires_ec2: !!str False
domain: !!str performance
class: !!str collect
distro: !!str alami ubuntu rhel suse
required: !!str period times
optional: !!str
software: !!str
sudo: !!str False
perfimpact: !!str False
parallelexclusive: !!str

Send a diagnostic interrupt (for advanced users)

Warning
Diagnostic interrupts are intended for use by advanced users. Incorrect usage could negatively
impact your instance. Sending a diagnostic interrupt to an instance could trigger an instance to
crash and reboot, which could lead to the loss of data.

You can send a diagnostic interrupt to an unreachable or unresponsive Linux instance to manually trigger
a kernel panic.

Linux operating systems typically crash and reboot when a kernel panic occurs. The specific behavior
of the operating system depends on its configuration. A kernel panic can also be used to cause the
instance's operating system kernel to perform tasks, such as generating a crash dump file. You can then
use the information in the crash dump file to conduct root cause analysis and debug the instance.

The crash dump data is generated locally by the operating system on the instance itself.

Before sending a diagnostic interrupt to your instance, we recommend that you consult the
documentation for your operating system and then make the necessary configuration changes.

Contents
• Supported instance types (p. 1378)

1377
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Supported instance types

• Prerequisites (p. 1378)

• Send a diagnostic interrupt (p. 1380)

Supported instance types

Diagnostic interrupt is supported on all Nitro-based instance types, except A1. For more information, see
Instances built on the Nitro System (p. 209).

Prerequisites
Before using a diagnostic interrupt, you must configure your instance's operating system. This ensures
that it performs the actions that you need when a kernel panic occurs.

To configure Amazon Linux 2 to generate a crash dump when a kernel panic occurs

1. Connect to your instance.

2. Install kexec and kdump.

[ec2-user ~]$ sudo yum install kexec-tools -y

3. Configure the kernel to reserve an appropriate amount of memory for the secondary kernel. The
amount of memory to reserve depends on the total available memory of your instance. Open
the /etc/default/grub file using your preferred text editor, locate the line that starts with
GRUB_CMDLINE_LINUX_DEFAULT, and then add the crashkernel parameter in the following
format: crashkernel=memory_to_reserve. For example, to reserve 160MB, modify the grub file
as follows:

GRUB_CMDLINE_LINUX_DEFAULT="crashkernel=160M console=tty0 console=ttyS0,115200n8

net.ifnames=0 biosdevname=0 nvme_core.io_timeout=4294967295 rd.emergency=poweroff
rd.shell=0"
GRUB_TIMEOUT=0
GRUB_DISABLE_RECOVERY="true"

4. Save the changes and close the grub file.

5. Rebuild the GRUB2 configuration file.

[ec2-user ~]$ sudo grub2-mkconfig -o /boot/grub2/grub.cfg

6. On instances based on Intel and AMD processors, the send-diagnostic-interrupt command

sends an unknown non-maskable interrupt (NMI) to the instance. You must configure the kernel to
crash when it receives the unknown NMI. Open the /etc/sysctl.conf file using your preferred
text editor and add the following.

kernel.unknown_nmi_panic=1

7. Reboot and reconnect to your instance.

8. Verify that the kernel has been booted with the correct crashkernel parameter.

$ grep crashkernel /proc/cmdline

The following example output indicates successful configuration.

BOOT_IMAGE=/boot/vmlinuz-4.14.128-112.105.amzn2.x86_64 root=UUID=a1e1011e-
e38f-408e-878b-fed395b47ad6 ro crashkernel=160M console=tty0 console=ttyS0,115200n8

1378
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Prerequisites

net.ifnames=0 biosdevname=0 nvme_core.io_timeout=4294967295 rd.emergency=poweroff

rd.shell=0

9. Verify that the kdump service is running.

[ec2-user ~]$ systemctl status kdump.service

The following example output shows the result if the kdump service is running.

kdump.service - Crash recovery kernel arming

Loaded: loaded (/usr/lib/systemd/system/kdump.service; enabled; vendor preset:
enabled)
Active: active (exited) since Fri 2019-05-24 23:29:13 UTC; 22s ago
Process: 2503 ExecStart=/usr/bin/kdumpctl start (code=exited, status=0/SUCCESS)
Main PID: 2503 (code=exited, status=0/SUCCESS)

Note
By default, the crash dump file is saved to /var/crash/. To change the location, modify the /
etc/kdump.conf file using your preferred text editor.

To configure Amazon Linux to generate a crash dump when a kernel panic occurs

1. Connect to your instance.

2. Install kexec and kdump.

[ec2-user ~]$ sudo yum install kexec-tools -y

3. Configure the kernel to reserve an appropriate amount of memory for the secondary kernel. The
amount of memory to reserve depends on the total available memory of your instance.

$ sudo grubby --args="crashkernel=memory_to_reserve" --update-kernel=ALL

For example, to reserve 160MB for the crash kernel, use the following command.

$ sudo grubby --args="crashkernel=160M" --update-kernel=ALL

4. On instances based on Intel and AMD processors, the send-diagnostic-interrupt command

sends an unknown non-maskable interrupt (NMI) to the instance. You must configure the kernel to
crash when it receives the unknown NMI. Open the /etc/sysctl.conf file using your preferred
text editor and add the following.

kernel.unknown_nmi_panic=1

5. Reboot and reconnect to your instance.

6. Verify that the kernel has been booted with the correct crashkernel parameter.

$ grep crashkernel /proc/cmdline

The following example output indicates successful configuration.

root=LABEL=/ console=tty1 console=ttyS0 selinux=0 nvme_core.io_timeout=4294967295

LANG=en_US.UTF-8 KEYTABLE=us crashkernel=160M

7. Verify that the kdump service is running.

1379
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
Send a diagnostic interrupt

[ec2-user ~]$ sudo service kdump status

If the service is running, the command returns the Kdump is operational response.

Note
By default, the crash dump file is saved to /var/crash/. To change the location, modify the /
etc/kdump.conf file using your preferred text editor.

To configure SUSE Linux Enterprise, Ubuntu, or Red Hat Enterprise Linux

See the following websites:

• SUSE Linux Enterprise

• Ubuntu
• Red Hat Enterprise Linux (RHEL)

Note
On instances based on Intel and AMD processors, the send-diagnostic-interrupt
command sends an unknown non-maskable interrupt (NMI) to the instance. You must configure
the kernel to crash when it receives the unknown NMI. Add the following to your configuration
file.

kernel.unknown_nmi_panic=1

Send a diagnostic interrupt

After you have completed the necessary configuration changes, you can send a diagnostic interrupt to
your instance using the AWS CLI or Amazon EC2 API.

To send a diagnostic interrupt to your instance (AWS CLI)

Use the send-diagnostic-interrupt command and specify the instance ID.

aws ec2 send-diagnostic-interrupt --instance-id i-1234567890abcdef0

1380
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Document history
The following table describes important additions to the Amazon EC2 documentation starting in 2019.
We also update the documentation frequently to address the feedback that you send us.

update-history-change update-history-description update-history-date

Multi-Attach support for io2 You can now enable Provisioned December 18, 2020
volumes IOPS SSD (io2) volumes for
Amazon EBS Multi-Attach.

C6gn instances (p. 1381) New instances featuring an AWS December 18, 2020
Graviton2 processor based on
64-bit Arm architecture. These
instances can utilize up to 100
Gbps of network bandwidth.

Amazon Data Lifecycle Manager Use Amazon Data Lifecycle December 17, 2020
Manager to automate the
process of sharing snapshots
and copying them across AWS
accounts.

G4ad instances (p. 1381) New instances powered by AMD December 9, 2020
Radeon Pro V520 GPUs and AMD
2nd Generation EPYC processors.

Tag AMIs and snapshots on AMI When you create an AMI, December 4, 2020
creation you can tag the AMI and the
snapshots with the same tags, or
you can tag them with different
tags.

io2 Block Express preview You can opt in to the io2 Block December 1, 2020
Express volumes preview. io2
Block Express volumes provide
sub-millisecond latency, and
support higher IOPS, higher
throughput, and larger capacity
than io2 volumes.

gp3 volumes (p. 1381) A new Amazon EBS General December 1, 2020
Purpose SSD volume type. You
can specify provisioned IOPS and
throughput when you create or
modify the volume.

D3, D3en, M5zn, and R5b New instance types built on the December 1, 2020
instances (p. 1381) Nitro System.

Throughput Optimized HDD and Throughput Optimized HDD November 30, 2020
Cold HDD volume sizes (st1) and Cold HDD (sc1)
volumes can range in size from
125 GiB to 16 TiB.

1381
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Mac1 instances New instances built on Apple November 30, 2020

Mac mini computers that
support running macOS
workloads on Amazon EC2.

Use Amazon EventBridge to Create EventBridge rules that November 20, 2020
monitor Spot Fleet events trigger programmatic actions
in response to Spot Fleet state
changes and errors.

Use Amazon EventBridge to Create EventBridge rules that November 20, 2020
monitor EC2 Fleet events trigger programmatic actions
in response to EC2 Fleet state
changes and errors.

Delete instant fleets Delete an EC2 Fleet of type November 18, 2020
instant and terminate all the
instances in the fleet in a single
API call.

Hibernation support for T3 and Hibernate your newly-launched November 17, 2020
T3a instances running on T3 and T3a
instance types.

Amazon EFS Quick Create You can create and mount an November 9, 2020
Amazon Elastic File System
(Amazon EFS) file system to an
instance at launch using Amazon
EFS Quick Create.

Amazon Data Lifecycle Manager You can use Amazon Data November 9, 2020
Lifecycle Manager to automate
the creation, retention, and
deletion of EBS-backed AMIs.

Instance metadata category: The approximate time, in UTC, November 4, 2020

events/recommendations/ when the EC2 instance rebalance
rebalance recommendation notification is
emitted for the instance.

EC2 instance rebalance A signal that notifies you when a November 4, 2020
recommendation Spot Instance is at elevated risk
of interruption.

Capacity Reservations in Capacity Reservations can November 4, 2020

Wavelength Zones now be created and used in
Wavelength Zones.

Capacity Rebalancing Configure Spot Fleet or EC2 November 4, 2020

Fleet to launch a replacement
Spot Instance when Amazon
EC2 emits a rebalance
recommendation.

P4d instances (p. 1381) New accelerated computing November 2, 2020

instances that provide a high-
performance platform for
machine learning and HPC
workloads.

1382
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Hibernation support for I3, Hibernate your newly-launched October 21, 2020
M5ad, and R5ad instances running on I3, M5ad,
and R5ad instance types.

Spot Instance vCPU limits Spot Instance limits are now October 1, 2020
managed in terms of the number
of vCPUs that your running Spot
Instances are either using or will
use pending the fulfillment of
open requests.

Capacity Reservations in Local Capacity Reservations can now September 30, 2020
Zones be created and used in Local
Zones.

Amazon Data Lifecycle Manager Amazon Data Lifecycle Manager September 17, 2020
policies can be configured with
up to four schedules.

T4g instances (p. 1381) New general purpose instances September 14, 2020
powered by AWS Graviton2
processors, which are based on
64-bit Arm Neoverse cores and
custom silicon designed by AWS
for optimized performance and
cost.

Hibernation support for M5a Hibernate your newly-launched August 28, 2020
and R5a instances running on M5a and
R5a instance types.

Provisioned IOPS SSD (io2) Provisioned IOPS SSD (io2) August 24, 2020
volumes for Amazon EBS volumes are designed to
provide 99.999 percent volume
durability with an AFR no higher
than 0.001 percent.

Instance metadata provides New instance metadata fields August 24, 2020
instance location and placement under the placement category:
information Region, placement group name,
partition number, host ID, and
Availability Zone ID.

C5ad instances (p. 1381) New compute optimized August 13, 2020
instances featuring second-
generation AMD EYPC
processors.

Wavelength Zones A Wavelength Zone is an August 6, 2020

isolated zone in the carrier
location where the Wavelength
infrastructure is deployed.

Capacity Reservation groups You can use AWS Resource July 29, 2020
Groups to create logical
collections of Capacity
Reservations, and then target
instance launches into those
groups.

1383
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

C6gd, M6gd, and R6gd New general purpose instances July 27, 2020
instances (p. 1381) powered by AWS Graviton2
processors, which are based on
64-bit Arm Neoverse cores and
custom silicon designed by AWS
for optimized performance and
cost.

Fast snapshot restore You can enable fast snapshot July 21, 2020
restore for snaphots that are
shared with you.

C6g and R6g instances (p. 1381) New general purpose instances June 10, 2020
powered by AWS Graviton2
processors, which are based on
64-bit Arm Neoverse cores and
custom silicon designed by AWS
for optimized performance and
cost.

Bare metal instances for New instances that provide your June 5, 2020
G4dn (p. 1381) applications with direct access
to the physical resources of the
host server.

C5a instances (p. 1381) New compute optimized June 4, 2020

instances featuring second-
generation AMD EYPC
processors.

Bring your own IPv6 addresses You can bring part or all of your May 21, 2020
IPv6 address range from your
on-premises network to your
AWS account.

M6g instances (p. 1381) New general purpose instances May 11, 2020
powered by AWS Graviton2
processors, which are based on
64-bit Arm Neoverse cores and
custom silicon designed by AWS
for optimized performance and
cost.

Launch instances using a You can specify a AWS Systems May 5, 2020
Systems Manager parameter Manager parameter instead
of an AMI when you launch an
instance.

Customize scheduled event You can customize scheduled May 4, 2020

notifications event notifications to include
tags in the email notification.

Amazon Linux 2 Kernel Live Kernel Live Patching for Amazon April 28, 2020
Patching Linux 2 enables you to apply
security vulnerability and
critical bug patches to a running
Linux kernel, without reboots
or disruptions to running
applications.

1384
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Amazon EBS Multi-Attach You can now attach a single February 14, 2020
Provisioned IOPS SSD (io1)
volume to up to 16 Nitro-based
instances that are in the same
Availability Zone.

Stop and start a Spot Instance Stop your Spot Instances backed January 13, 2020
by Amazon EBS and start them
at will, instead of relying on the
stop interruption behavior.

Resource tagging (p. 1381) You can tag egress-only internet January 10, 2020
gateways, local gateways, local
gateway route tables, local
gateway virtual interfaces, local
gateway virtual interface groups,
local gateway route table VPC
associations, and local gateway
route table virtual interface
group associations.

Connect to your instance using You can start a Session Manager December 18, 2019
Session Manager session with an instance from
the Amazon EC2 console.

Inf1 instances (p. 1381) New instances featuring AWS December 3, 2019
Inferentia, a machine learning
inference chip designed to
deliver high performance at a
low cost.

Dedicated Hosts and host Dedicated Hosts can now be December 2, 2019
resource groups used with host resource groups.

Dedicated Host sharing You can now share your December 2, 2019
Dedicated Hosts across AWS
accounts.

Default credit specification at You can set the default credit November 25, 2019
the account level specification per burstable
performance instance family
at the account level per AWS
Region.

Instance type discovery You can find an instance type November 22, 2019
that meets your needs.

Dedicated Hosts (p. 1381) You can now configure a November 21, 2019
Dedicated Host to support
multiple instance types in an
instance family.

Amazon EBS fast snapshot You can enable fast snapshot November 20, 2019
restores restores on an EBS snapshot
to ensure that EBS volumes
created from the snapshot are
fully-initialized at creation and
instantly deliver all of their
provisioned performance.

1385
 Amazon Elastic Compute Cloud
User Guide for Linux Instances

Instance Metadata Service You can use Instance Metadata November 19, 2019
Version 2 Service Version 2, which is a
session-oriented method for
requesting instance metadata.

Elastic Fabric Adapter (p. 1381) Elastic Fabric Adapters can now November 15, 2019
be used with Intel MPI 2019
Update 6.

Queued purchases of Reserved You can queue the purchase of October 4, 2019
Instances a Reserved Instance up to three
years in advance.

G4dn instances (p. 1381) New instances featuring NVIDIA September 19, 2019
Tesla GPUs.

Diagnostic interrupt You can send a diagnostic August 14, 2019

interrupt to an unreachable or
unresponsive instance to trigger
a kernel panic.

Capacity optimized allocation Using EC2 Fleet or Spot Fleet, August 12, 2019
strategy you can launch Spot Instances
from Spot pools with optimal
capacity for the number of
instances that are launching.

On-Demand Capacity You can now share your Capacity July 29, 2019
Reservation sharing Reservations across AWS
accounts.

Elastic Fabric Adapter (p. 1381) EFA now supports Open MPI July 26, 2019
3.1.4 and Intel MPI 2019 Update
4.

Resource tagging (p. 1381) Launch templates on creation. July 24, 2019

EC2 Instance Connect EC2 Instance Connect is a simple June 27, 2019
and secure way to connect to
your instances using Secure Shell
(SSH).

Host recovery Automatically restart your June 5, 2019

instances on a new host in
the event of an unexpected
hardware failure on a Dedicated
Host.

Amazon EBS multi-volume You can take exact point-in-time, May 29, 2019
snapshots data coordinated, and crash-
consistent snapshots across
multiple EBS volumes attached
to an EC2 instance.

Resource tagging (p. 1381) You can tag Dedicated Host May 27, 2019
Reservations.

1386
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Amazon EBS encryption by After you enable encryption May 23, 2019
default by default in a Region, all new
EBS volumes you create in the
Region are encrypted using the
default CMK for EBS encryption.

Resource tagging (p. 1381) You can tag VPC endpoints, May 13, 2019
endpoint services, and endpoint
service configurations.

Windows to Linux Replatforming Move existing Microsoft SQL May 8, 2019

Assistant for Microsoft SQL Server workloads from a
Server Databases Windows to a Linux operating
system.

I3en instances (p. 1381) New I3en instances can utilize May 8, 2019
up to 100 Gbps of network
bandwidth.

Elastic Fabric Adapter You can attach an Elastic Fabric April 29, 2019
Adapter to your instances to
accelerate High Performance
Computing (HPC) applications.

T3a instances (p. 1381) New instances featuring AMD April 24, 2019
EYPC processors.

M5ad and R5ad New instances featuring AMD March 27, 2019
instances (p. 1381) EYPC processors.

Resource tagging (p. 1381) You can assign custom tags March 14, 2019
to your Dedicated Host
Reservations to categorize them
in different ways.

Bare metal instances for M5, New instances that provide your February 13, 2019
M5d, R5, R5d, and z1d (p. 1381) applications with direct access
to the physical resources of the
host server.

History for previous years

The following table describes important additions to the Amazon EC2 documentation in 2018 and earlier
years.

Feature API Description Release

version date

Partition placement 2016-11-15 Partition placement groups spread instances 20

groups across logical partitions, ensuring that instances December
in one partition do not share underlying 2018
hardware with instances in other partitions.
For more information, see Partition placement
groups (p. 929).

1387
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

p3dn.24xlarge 2016-11-15 New p3dn.24xlarge instances provide 100 Gbps of 7

instances network bandwidth. December
2018

Hibernate EC2 Linux 2016-11-15 You can hibernate a Linux instance if it's enabled 28
instances for hibernation and it meets the hibernation November
prerequisites. For more information, see Hibernate 2018
your Linux instance (p. 627).

Amazon Elastic 2016-11-15 You can attach an Amazon EI accelerator to your 28

Inference Accelerators instances to add GPU-powered acceleration November
to reduce the cost of running deep learning 2018
inference. For more information, see Amazon
Elastic Inference (p. 738).

Instances featuring 2016-11-15 New C5n instances can utilize up to 100 Gbps of 26
100 Gbps of network network bandwidth. November
bandwidth 2018

Instances featuring 2016-11-15 New A1 instances deliver significant cost savings 26

Arm-based Processors and are ideally suited for scale-out and Arm-based November
workloads. 2018

Spot console 2016-11-15 The Spot console recommends a fleet of 20

recommends a fleet of instances based on Spot best practice (instance November
instances diversification) to meet the minimum hardware 2018
specifications (vCPUs, memory, and storage) for
your application need. For more information, see
Create a Spot Fleet request (p. 407).

New EC2 Fleet request 2016-11-15 EC2 Fleet now supports a new request type, 14
type: instant instant, that you can use to synchronously November
provision capacity across instance types and 2018
purchase models. The instant request returns
the launched instances in the API response, and
takes no further action, enabling you to control
if and when instances are launched. For more
information, see EC2 Fleet request types (p. 551).

Instances featuring 2016-11-15 New general purpose (M5a) and memory 6

AMD EYPC processors optimized instances (R5a) offer lower-priced November
options for microservices, small to medium 2018
databases, virtual desktops, development and test
environments, business applications, and more.

Spot savings 2016-11-15 You can view the savings made from using Spot 5
information Instances for a single Spot Fleet or for all Spot November
Instances. For more information, see Savings from 2018
purchasing Spot Instances (p. 381).

Console support for 2016-11-15 When you launch an instance, you can optimize 31
optimizing CPU options the CPU options to suit specific workloads or October
business needs using the Amazon EC2 console. 2018
For more information, see Optimize CPU
options (p. 668).

1388
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Console support for 2016-11-15 You can create a launch template using 30
creating a launch an instance as the basis for a new launch October
template from an template using the Amazon EC2 console. 2018
instance For more information, see Create a launch
template (p. 534).

On-Demand Capacity 2016-11-15 You can reserve capacity for your Amazon EC2 25
Reservations instances in a specific Availability Zone for any October
duration. This allows you to create and manage 2018
capacity reservations independently from the
billing discounts offered by Reserved Instances
(RI). For more information, see On-Demand
Capacity Reservations (p. 501).

Bring Your Own IP 2016-11-15 You can bring part or all of your public IPv4 23
Addresses (BYOIP) address range from your on-premises network to October
your AWS account. After you bring the address 2018
range to AWS, it appears in your account as an
address pool. You can create an Elastic IP address
from your address pool and use it with your AWS
resources. For more information, see Bring your
own IP addresses (BYOIP) in Amazon EC2 (p. 825).

g3s.xlarge instances 2016-11-15 Expands the range of the accelerated-computing 11

G3 instance family with the introduction of October
g3s.xlarge instances. 2018

Dedicated Host tag 2016-11-15 You can tag your Dedicated Hosts on creation, and 08
on create and console you can manage your Dedicated Host tags using October
support the Amazon EC2 console. For more information, 2018
see Allocate Dedicated Hosts (p. 469).

High memory instances 2016-11-15 These instances are purpose-built to run large 27
in-memory databases. They offer bare metal September
performance with direct access to host hardware. 2018
For more information, see Memory optimized
instances (p. 272).

f1.4xlarge instances 2016-11-15 Expands the range of the accelerated-computing 25

F1 instance family with the introduction of September
f1.4xlarge instances. 2018

Console support for 2016-11-15 Increase or decrease the current capacity of 20

scheduled scaling for the fleet based on the date and time. For more September
Spot Fleet information, see Scale Spot Fleet using scheduled 2018
scaling (p. 442).

T3 instances 2016-11-15 T3 instances are burstable general-purpose 21 August

instance type that provide a baseline level of 2018
CPU performance with the ability to burst CPU
usage at any time for as long as required. For
more information, see Burstable performance
instances (p. 224).

1389
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Allocation strategies for 2016-11-15 You can specify whether On-Demand capacity 26 July
EC2 Fleets is fulfilled by price (lowest price first) or priority 2018
(highest priority first). You can specify the number
of Spot pools across which to allocate your
target Spot capacity. For more information, see
Allocation strategies for Spot Instances (p. 551).

Allocation strategies for 2016-11-15 You can specify whether On-Demand capacity 26 July
Spot Fleets is fulfilled by price (lowest price first) or priority 2018
(highest priority first). You can specify the number
of Spot pools across which to allocate your
target Spot capacity. For more information, see
Allocation strategy for Spot Instances (p. 372).

R5 and R5d instances 2016-11-15 R5 and R5d instances are ideally suited for high- 25 July
performance databases, distributed in-memory 2018
caches, and in-memory analytics. R5d instances
come with NVMe instance store volumes. For
more information, see Memory optimized
instances (p. 272).

z1d instances 2016-11-15 These instances are designed for applications 25 July
that require high per-core performance with 2018
a large amount of memory, such as electronic
design automation (EDA) and relational databases.
These instances come with NVME instance store
volumes. For more information, see Memory
optimized instances (p. 272).

Automate snapshot 2016-11-15 You can use Amazon Data Lifecycle Manager to 12 July
lifecycle automate creation and deletion of snapshots for 2018
your EBS volumes. For more information, see
Amazon Data Lifecycle Manager (p. 1186).

Launch template CPU 2016-11-15 When you create a launch template using the 11 July
options command line tools, you can optimize the CPU 2018
options to suit specific workloads or business
needs. For more information, see Create a launch
template (p. 534).

Tag Dedicated Hosts 2016-11-15 You can tag your Dedicated Hosts. For more 3 July
information, see Tag Dedicated Hosts (p. 480). 2018

i3.metal instances 2016-11-15 i3.metal instances provide your applications 17 May

with direct access to the physical resources of 2018
the host server, such as processors and memory.
For more information, see Storage optimized
instances (p. 283).

Get latest console 2016-11-15 You can retrieve the latest console output for 9 May
output some instance types when you use the get- 2018
console-output AWS CLI command.

1390
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Optimize CPU options 2016-11-15 When you launch an instance, you can optimize 8 May
the CPU options to suit specific workloads or 2018
business needs. For more information, see
Optimize CPU options (p. 668).

EC2 Fleet 2016-11-15 You can use EC2 Fleet to launch a group of 2 May
instances across different EC2 instance types 2018
and Availability Zones, and across On-Demand
Instance, Reserved Instance, and Spot Instance
purchasing models. For more information, see
Launch instances using an EC2 Fleet (p. 548).

On-Demand Instances 2016-11-15 You can include a request for On-Demand 2 May
in Spot Fleets capacity in your Spot Fleet request to ensure 2018
that you always have instance capacity. For more
information, see How Spot Fleet works (p. 371).

Tag EBS snapshots on 2016-11-15 You can apply tags to snapshots during creation. 2 April
creation For more information, see Create Amazon EBS 2018
snapshots (p. 1129).

Change placement 2016-11-15 You can move an instance in or out of a placement 1 March
groups group, or change its placement group. For more 2018
information, see Change the placement group for
an instance (p. 937).

Longer resource IDs 2016-11-15 You can enable the longer ID format for more 9 February
resource types. For more information, see 2018
Resource IDs (p. 1305).

Network performance 2016-11-15 Instances outside of a cluster placement group 24 January

improvements can now benefit from increased bandwidth 2018
when sending or receiving network traffic
between other instances or Amazon S3. For
more information, see Networking and storage
features (p. 209).

Tag Elastic IP addresses 2016-11-15 You can tag your Elastic IP addresses. For 21
more information, see Tage an Elastic IP December
address (p. 834). 2017

Amazon Linux 2 2016-11-15 Amazon Linux 2 is a new version of Amazon 13

Linux. It provides a high performance, stable, and December
secure foundation for your applications. For more 2017
information, see Amazon Linux (p. 178).

Amazon Time Sync 2016-11-15 You can use the Amazon Time Sync Service to 29
Service keep accurate time on your instance. For more November
information, see Set the time for your Linux 2017
instance (p. 664).

T2 Unlimited 2016-11-15 T2 Unlimited instances can burst above the 29

baseline for as long as required. For more November
information, see Burstable performance 2017
instances (p. 224).

1391
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Launch templates 2016-11-15 A launch template can contain all or some of the 29
parameters to launch an instance, so that you November
don't have to specify them every time you launch 2017
an instance. For more information, see Launch an
instance from a launch template (p. 532).

Spread placement 2016-11-15 Spread placement groups are recommended for 29

applications that have a small number of critical November
instances that should be kept separate from 2017
each other. For more information, see Spread
placement groups (p. 930).

H1 instances 2016-11-15 H1 instances are designed for high-performance 28

big data workloads. For more information, see November
Storage optimized instances (p. 283). 2017

M5 instances 2016-11-15 M5 instances are general purpose compute 28

instances. They provide a balance of compute, November
memory, storage, and network resources. 2017

Spot Instance 2016-11-15 The Spot service can hibernate Spot Instances 28
hibernation in the event of an interruption. For more November
information, see Hibernate interrupted Spot 2017
Instances (p. 454).

Spot Fleet target 2016-11-15 You can set up target tracking scaling policies for 17
tracking your Spot Fleet. For more information, see Scale November
Spot Fleet using a target tracking policy (p. 439). 2017

Spot Fleet integrates 2016-11-15 You can attach one or more load balancers to a 10
with Elastic Load Spot Fleet. November
Balancing 2017

X1e instances 2016-11-15 X1e instances are ideally suited for high- 28
performance databases, in-memory databases, November
and other memory-intensive enterprise 2017
applications. For more information, see Memory
optimized instances (p. 272).

C5 instances 2016-11-15 C5 instances are designed for compute-heavy 6

applications. For more information, see Compute November
optimized instances (p. 264). 2017

Merge and split 2016-11-15 You can exchange (merge) two or more 6
Convertible Reserved Convertible Reserved Instances for a new November
Instances Convertible Reserved Instance. You can also use 2017
the modification process to split a Convertible
Reserved Instance into smaller reservations. For
more information, see Exchange Convertible
Reserved Instances (p. 359).

P3 instances 2016-11-15 P3 instances are compute-optimized GPU 25

instances. For more information, see Linux October
accelerated computing instances (p. 290). 2017

1392
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Modify VPC tenancy 2016-11-15 You can change the instance tenancy attribute 16
of a VPC from dedicated to default. For October
more information, see Change the tenancy of a 2017
VPC (p. 500).

Per second billing 2016-11-15 Amazon EC2 charges for Linux-based usage by the 2 October
second, with a one-minute minimum charge. 2017

Stop on interruption 2016-11-15 You can specify whether Amazon EC2 should 18
stop or terminate Spot Instances when they September
are interrupted. For more information, see 2017
Interruption behaviors (p. 453).

Tag NAT gateways 2016-11-15 You can tag your NAT gateway. For more 7
information, see Tag your resources (p. 1313). September
2017

Security group rule 2016-11-15 You can add descriptions to your security group 31 August
descriptions rules. For more information, see Security group 2017
rules (p. 1060).

Recover Elastic IP 2016-11-15 If you release an Elastic IP address for use in 11 August
addresses a VPC, you might be able to recover it. For 2017
more information, see Recover an Elastic IP
address (p. 837).

Tag Spot Fleet instances 2016-11-15 You can configure your Spot Fleet to 24 July
automatically tag the instances that it launches. 2017

G3 instances 2016-11-15 G3 instances provide a cost-effective, high- 13 July

performance platform for graphics applications 2017
using DirectX or OpenGL. G3 instances also
provide NVIDIA GRID Virtual Workstation features,
supporting 4 monitors with resolutions up to
4096x2160. For more information, see Linux
accelerated computing instances (p. 290).

F1 instances 2016-11-15 F1 instances are accelerated computing instances. 19 April

For more information, see Linux accelerated 2017
computing instances (p. 290).

Tag resources during 2016-11-15 You can apply tags to instances and volumes 28 March
creation during creation. For more information, see Tag 2017
your resources (p. 1313). In addition, you can use
tag-based resource-level permissions to control
the tags that are applied. For more information
see, Grant permission to tag resources during
creation (p. 984).

I3 instances 2016-11-15 I3 instances are storage optimized instances. 23

For more information, see Storage optimized February
instances (p. 283). 2017

1393
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Perform modifications 2016-11-15 With most EBS volumes attached to most EC2 13
on attached EBS instances, you can modify volume size, type, and February
volumes IOPS without detaching the volume or stopping 2017
the instance. For more information, see Amazon
EBS Elastic Volumes (p. 1164).

Attach an IAM role 2016-11-15 You can attach, detach, or replace an IAM role for 9 February
an existing instance. For more information, see 2017
IAM roles for Amazon EC2 (p. 1032).

Dedicated Spot 2016-11-15 You can run Spot Instances on single-tenant 19 January
Instances hardware in a virtual private cloud (VPC). For 2017
more information, see Specify a tenancy for your
Spot Instances (p. 385).

IPv6 support 2016-11-15 You can associate an IPv6 CIDR with your VPC and 1
subnets, and assign IPv6 addresses to instances in December
your VPC. For more information, see Amazon EC2 2016
instance IP addressing (p. 809).

R4 instances 2016-09-15 R4 instances are memory optimized instances. R4 30

instances are well-suited for memory-intensive, November
latency-sensitive workloads such as business 2016
intelligence (BI), data mining and analysis, in-
memory databases, distributed web scale in-
memory caching, and applications performance
real-time processing of unstructured big data.
For more information, see Memory optimized
instances (p. 272)

New t2.xlarge and 2016-09-15 T2 instances are designed to provide moderate 30

t2.2xlarge instance base performance and the capability to burst to November
types significantly higher performance as required by 2016
your workload. They are intended for applications
that need responsiveness, high performance
for limited periods of time, and a low cost. For
more information, see Burstable performance
instances (p. 224).

P2 instances 2016-09-15 P2 instances use NVIDIA Tesla K80 GPUs and are 29
designed for general purpose GPU computing September
using the CUDA or OpenCL programming models. 2016
For more information, see Linux accelerated
computing instances (p. 290).

m4.16xlarge 2016-04-01 Expands the range of the general-purpose M4 6

instances family with the introduction of m4.16xlarge September
instances, with 64 vCPUs and 256 GiB of RAM. 2016

Automatic scaling for   You can now set up scaling policies for your Spot 1
Spot Fleet Fleet. For more information, see Automatic scaling September
for Spot Fleet (p. 437). 2016

1394
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Elastic Network Adapter 2016-04-01 You can now use ENA for enhanced networking. 28 June
(ENA) For more information, see Enhanced networking 2016
support (p. 865).

Enhanced support for 2016-04-01 You can now view and modify longer ID settings 23 June
viewing and modifying for other IAM users, IAM roles, or the root user. For 2016
longer IDs more information, see Resource IDs (p. 1305).

Copy encrypted 2016-04-01 You can now copy encrypted EBS snapshots 21 June
Amazon EBS snapshots between AWS accounts. For more information, see 2016
between AWS accounts Copy an Amazon EBS snapshot (p. 1134).

Capture a screenshot of 2015-10-01 You can now obtain additional information when 24 May
an instance console debugging instances that are unreachable. For 2016
more information, see Capture a screenshot of an
unreachable instance (p. 1365).

X1 instances 2015-10-01 Memory-optimized instances designed for 18 May

running in-memory databases, big data 2016
processing engines, and high performance
computing (HPC) applications. For more
information, see Memory optimized
instances (p. 272).

Two new EBS volume 2015-10-01 You can now create Throughput Optimized 19 April
types HDD (st1) and Cold HDD (sc1) volumes. For 2016
more information, see Amazon EBS volume
types (p. 1084).

Added new Added new NetworkPacketsIn and 23 March

NetworkPacketsIn and NetworkPacketsOut metrics for Amazon EC2. For 2016
NetworkPacketsOut more information, see Instance metrics (p. 764).
metrics for Amazon EC2

CloudWatch metrics for You can now get CloudWatch metrics for your 21 March
Spot Fleet Spot Fleet. For more information, see CloudWatch 2016
metrics for Spot Fleet (p. 428).

Scheduled Instances 2015-10-01 Scheduled Reserved Instances (Scheduled 13 January

Instances) enable you to purchase capacity 2016
reservations that recur on a daily, weekly, or
monthly basis, with a specified start time and
duration. For more information, see Scheduled
Reserved Instances (p. 363).

Longer resource IDs 2015-10-01 We're gradually introducing longer length IDs 13 January
for some Amazon EC2 and Amazon EBS resource 2016
types. During the opt-in period, you can enable
the longer ID format for supported resource
types. For more information, see Resource
IDs (p. 1305).

1395
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

ClassicLink DNS support 2015-10-01 You can enable ClassicLink DNS support for 11 January
your VPC so that DNS hostnames that are 2016
addressed between linked EC2-Classic instances
and instances in the VPC resolve to private
IP addresses and not public IP addresses. For
more information, see Enable ClassicLink DNS
support (p. 956).

New t2.nano instance 2015-10-01 T2 instances are designed to provide moderate 15

type base performance and the capability to burst to December
significantly higher performance as required by 2015
your workload. They are intended for applications
that need responsiveness, high performance
for limited periods of time, and a low cost. For
more information, see Burstable performance
instances (p. 224).

Dedicated hosts 2015-10-01 An Amazon EC2 Dedicated host is a physical 23

server with instance capacity dedicated for November
your use. For more information, see Dedicated 2015
Hosts (p. 465).

Spot Instance duration 2015-10-01 You can now specify a duration for your Spot 6 October
Instances. For more information, see Define a 2015
duration for your Spot Instances (p. 384).

Spot Fleet modify 2015-10-01 You can now modify the target capacity of your 29
request Spot Fleet request. For more information, see September
Modify a Spot Fleet request (p. 417). 2015

Spot Fleet diversified 2015-04-15 You can now allocate Spot Instances in multiple 15
allocation strategy Spot pools using a single Spot Fleet request. For September
more information, see Allocation strategy for Spot 2015
Instances (p. 372).

Spot Fleet instance 2015-04-15 You can now define the capacity units that each 31 August
weighting instance type contributes to your application's 2015
performance, and adjust the amount you are
willing to pay for Spot Instances for each Spot
pool accordingly. For more information, see Spot
Fleet instance weighting (p. 376).

New reboot alarm Added the reboot alarm action and new IAM role 23 July
action and new IAM for use with alarm actions. For more information, 2015
role for use with alarm see Create alarms that stop, terminate, reboot, or
actions recover an instance (p. 784).

New t2.large T2 instances are designed to provide moderate 16 June

instance type base performance and the capability to burst to 2015
significantly higher performance as required by
your workload. They are intended for applications
that need responsiveness, high performance
for limited periods of time, and a low cost. For
more information, see Burstable performance
instances (p. 224).

1396
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

M4 instances General-purpose instances that provide a balance 11 June

of compute, memory, and network resources. M4 2015
instances are powered by a custom Intel 2.4 GHz
Intel® Xeon® E5 2676v3 (Haswell) processor with
AVX2.

Spot Fleets 2015-04-15 You can manage a collection, or fleet, of Spot 18 May
Instances instead of managing separate Spot 2015
Instance requests. For more information, see How
Spot Fleet works (p. 371).

Migrate Elastic IP 2015-04-15 You can migrate an Elastic IP address that you've 15 May
addresses to EC2- allocated for use in EC2-Classic to be used in a 2015
Classic VPC. For more information, see Migrate an Elastic
IP Address from EC2-Classic (p. 948).

Importing VMs with 2015-03-01 The VM Import process now supports importing 23 April
multiple disks as AMIs VMs with multiple disks as AMIs. For more 2015
information, see Importing a VM as an Image
Using VM Import/Export in the VM Import/Export
User Guide .

New g2.8xlarge The new g2.8xlarge instance is backed by four 7 April

instance type high-performance NVIDIA GPUs, making it well 2015
suited for GPU compute workloads including large
scale rendering, transcoding, machine learning,
and other server-side workloads that require
massive parallel processing power.

D2 instances Dense-storage instances that are optimized for 24 March

applications requiring sequential access to large 2015
amount of data on direct attached instance
storage. D2 instances are designed to offer best
price/performance in the dense-storage family.
Powered by 2.4 GHz Intel® Xeon® E5 2676v3
(Haswell) processors, D2 instances improve on
HS1 instances by providing additional compute
power, more memory, and Enhanced Networking.
In addition, D2 instances are available in four
instance sizes with 6TB, 12TB, 24TB, and 48TB
storage options.

For more information, see Storage optimized

instances (p. 283).

1397
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Automatic recovery for You can create an Amazon CloudWatch alarm 12 January
EC2 instances that monitors an Amazon EC2 instance and 2015
automatically recovers the instance if it becomes
impaired due to an underlying hardware failure
or a problem that requires AWS involvement to
repair. A recovered instance is identical to the
original instance, including the instance ID, IP
addresses, and all instance metadata.

For more information, see Recover your

instance (p. 648).

C4 instances Next-generation compute-optimized instances 11 January

that provide very high CPU performance at 2015
an economical price. C4 instances are based
on custom 2.9 GHz Intel® Xeon® E5-2666 v3
(Haswell) processors. With additional Turbo boost,
the processor clock speed in C4 instances can
reach as high as 3.5Ghz with 1 or 2 core turbo.
Expanding on the capabilities of C3 compute-
optimized instances, C4 instances offer customers
the highest processor performance among EC2
instances. These instances are ideally suited
for high-traffic web applications, ad serving,
batch processing, video encoding, distributed
analytics, high-energy physics, genome analysis,
and computational fluid dynamics.

For more information, see Compute optimized

instances (p. 264).

ClassicLink 2014-10-01 ClassicLink enables you to link your EC2-Classic 7 January

instance to a VPC in your account. You can 2015
associate VPC security groups with the EC2-Classic
instance, enabling communication between your
EC2-Classic instance and instances in your VPC
using private IP addresses. For more information,
see ClassicLink (p. 950).

Spot Instance   The best way to protect against Spot Instance 5 January
termination notices interruption is to architect your application to be 2015
fault tolerant. In addition, you can take advantage
of Spot Instance termination notices, which
provide a two-minute warning before Amazon
EC2 must terminate your Spot Instance.

For more information, see Spot Instance

interruption notices (p. 457).

DescribeVolumes 2014-09-01 The DescribeVolumes API call now supports the 23

pagination support pagination of results with the MaxResults and October
NextToken parameters. For more information, 2014
see DescribeVolumes in the Amazon EC2 API
Reference.

1398
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

T2 instances 2014-06-15 T2 instances are designed to provide moderate 30 June

base performance and the capability to burst to 2014
significantly higher performance as required by
your workload. They are intended for applications
that need responsiveness, high performance
for limited periods of time, and a low cost. For
more information, see Burstable performance
instances (p. 224).

New EC2 Service Limits   Use the EC2 Service Limits page in the Amazon 19 June
page EC2 console to view the current limits for 2014
resources provided by Amazon EC2 and Amazon
VPC, on a per-region basis.

Amazon EBS General 2014-05-01 General Purpose SSD volumes offer cost- 16 June
Purpose SSD Volumes effective storage that is ideal for a broad range 2014
of workloads. These volumes deliver single-digit
millisecond latencies, the ability to burst to 3,000
IOPS for extended periods of time, and a base
performance of 3 IOPS/GiB. General Purpose SSD
volumes can range in size from 1 GiB to 1 TiB.
For more information, see General Purpose SSD
volumes (gp2) (p. 1087).

Amazon EBS encryption 2014-05-01 Amazon EBS encryption offers seamless 21 May
encryption of EBS data volumes and snapshots, 2014
eliminating the need to build and maintain a
secure key management infrastructure. EBS
encryption enables data at rest security by
encrypting your data using Amazon-managed
keys. The encryption occurs on the servers that
host EC2 instances, providing encryption of data
as it moves between EC2 instances and EBS
storage. For more information, see Amazon EBS
encryption (p. 1176).

R3 instances 2014-02-01 Memory-optimized instances with the best price 9 April

point per GiB of RAM and high performance. 2014
These instances are ideally suited for relational
and NoSQL databases, in-memory analytics
solutions, scientific computing, and other
memory-intensive applications that can benefit
from the high memory per vCPU, high compute
performance, and enhanced networking
capabilities of R3 instances.

For more information about the hardware

specifications for each Amazon EC2 instance type,
see Amazon EC2 Instance Types.

New Amazon Linux AMI   Amazon Linux AMI 2014.03 is released. 27 March
release 2014

1399
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Amazon EC2 Usage   Amazon EC2 Usage Reports is a set of reports 28 January
Reports that shows cost and usage data of your usage of 2014
EC2. For more information, see Amazon EC2 usage
reports (p. 1325).

Additional M3 instances 2013-10-15 The M3 instance sizes m3.medium and m3.large 20 January
are now supported. For more information about 2014
the hardware specifications for each Amazon EC2
instance type, see Amazon EC2 Instance Types.

I2 instances 2013-10-15 These instances provide very high IOPS and 19

support TRIM on Linux instances for better December
successive SSD write performance. I2 instances 2013
also support enhanced networking that delivers
improve inter-instance latencies, lower network
jitter, and significantly higher packet per second
(PPS) performance. For more information, see
Storage optimized instances (p. 283).

Updated M3 instances 2013-10-15 The M3 instance sizes, m3.xlarge and 19

m3.2xlarge now support instance store with December
SSD volumes. 2013

Importing Linux virtual 2013-10-15 The VM Import process now supports the 16
machines importation of Linux instances. For more December
information, see the VM Import/Export User 2013
Guide.

Resource-level 2013-10-15 You can now create policies in AWS Identity and 20
permissions for Access Management to control resource-level November
RunInstances permissions for the Amazon EC2 RunInstances 2013
API action. For more information and example
policies, see Identity and access management for
Amazon EC2 (p. 977).

C3 instances 2013-10-15 Compute-optimized instances that provide very 14

high CPU performance at an economical price. November
C3 instances also support enhanced networking 2013
that delivers improved inter-instance latencies,
lower network jitter, and significantly higher
packet per second (PPS) performance. These
instances are ideally suited for high-traffic web
applications, ad serving, batch processing, video
encoding, distributed analytics, high-energy
physics, genome analysis, and computational fluid
dynamics.

For more information about the hardware

specifications for each Amazon EC2 instance type,
see Amazon EC2 Instance Types.

Launching an instance   You can now launch an instance from the AWS 11
from the AWS Marketplace using the Amazon EC2 launch November
Marketplace wizard. For more information, see Launch an AWS 2013
Marketplace instance (p. 546).

1400
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

G2 instances 2013-10-01 These instances are ideally suited for video 4

creation services, 3D visualizations, streaming November
graphics-intensive applications, and other 2013
server-side workloads requiring massive parallel
processing power. For more information, see Linux
accelerated computing instances (p. 290).

New launch wizard   There is a new and redesigned EC2 launch wizard. 10
For more information, see Launch an instance October
using the Launch Instance Wizard (p. 526). 2013

Modifying Instance 2013-10-01 You can now modify the instance type of Linux 09
Types of Amazon EC2 Reserved Instances within the same family (for October
Reserved Instances example, M1, M2, M3, C1). For more information, 2013
see Modify Reserved Instances (p. 352).

New Amazon Linux AMI   Amazon Linux AMI 2013.09 is released. 30

release September
2013

Modifying Amazon EC2 2013-08-15 You can now modify Reserved Instances in 11
Reserved Instances a Region. For more information, see Modify September
Reserved Instances (p. 352). 2013

Assigning a public IP 2013-07-15 You can now assign a public IP address when 20 August
address you launch an instance in a VPC. For more 2013
information, see Assign a public IPv4 address
during instance launch (p. 814).

Granting resource-level 2013-06-15 Amazon EC2 supports new Amazon Resource 8 July
permissions Names (ARNs) and condition keys. For more 2013
information, see IAM policies for Amazon
EC2 (p. 979).

Incremental Snapshot 2013-02-01 You can now perform incremental snapshot 11 June
Copies copies. For more information, see Copy an 2013
Amazon EBS snapshot (p. 1134).

New Tags page   There is a new Tags page in the Amazon EC2 04 April
console. For more information, see Tag your 2013
Amazon EC2 resources (p. 1311).

New Amazon Linux AMI   Amazon Linux AMI 2013.03 is released. 27 March
release 2013

Additional EBS- 2013-02-01 The following instance types can now be launched 19 March
optimized instance as EBS-optimized instances: c1.xlarge, 2013
types m2.2xlarge, m3.xlarge, and m3.2xlarge.

For more information, see Amazon EBS–optimized

instances (p. 1217).

1401
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Copy an AMI from one 2013-02-01 You can copy an AMI from one Region to another, 11 March
Region to another enabling you to launch consistent instances in 2013
more than one AWS Region quickly and easily.

For more information, see Copy an AMI (p. 166).

Launch instances into a 2013-02-01 Your AWS account is capable of launching 11 March
default VPC instances into either EC2-Classic or a VPC, or only 2013
into a VPC, on a region-by-region basis. If you
can launch instances only into a VPC, we create a
default VPC for you. When you launch an instance,
we launch it into your default VPC, unless you
create a nondefault VPC and specify it when you
launch the instance.

High-memory cluster 2012-12-01 Have large amounts of memory coupled with high 21 January
(cr1.8xlarge) instance CPU and network performance. These instances 2013
type are well suited for in-memory analytics, graph
analysis, and scientific computing applications.

High storage 2012-12-01 High storage instances provide very high storage 20
(hs1.8xlarge) density and high sequential read and write December
instance type performance per instance. They are well-suited 2012
for data warehousing, Hadoop/MapReduce, and
parallel file systems.

EBS snapshot copy 2012-12-01 You can use snapshot copies to create backups 17
of data, to create new Amazon EBS volumes, or December
to create Amazon Machine Images (AMIs). For 2012
more information, see Copy an Amazon EBS
snapshot (p. 1134).

Updated EBS metrics 2012-10-01 Updated the EBS metrics to include two new 20
and status checks for metrics for Provisioned IOPS SSD volumes. For November
Provisioned IOPS SSD more information, see Amazon CloudWatch 2012
volumes metrics for Amazon EBS (p. 1252). Also added
new status checks for Provisioned IOPS SSD
volumes. For more information, see EBS volume
status checks (p. 1117).

Linux Kernels   Updated AKI IDs; reorganized distribution kernels; 13

updated PVOps section. November
2012

M3 instances 2012-10-01 There are new M3 extra-large and M3 double- 31

extra-large instance types. For more information October
about the hardware specifications for each 2012
Amazon EC2 instance type, see Amazon EC2
Instance Types.

Spot Instance request 2012-10-01 Spot Instance request status makes it easy to 14
status determine the state of your Spot requests. October
2012

1402
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

New Amazon Linux AMI   Amazon Linux AMI 2012.09 is released. 11

release October
2012

Amazon EC2 Reserved 2012-08-15 The Reserved Instance Marketplace matches 11

Instance Marketplace sellers who have Amazon EC2 Reserved Instances September
that they no longer need with buyers who are 2012
looking to purchase additional capacity. Reserved
Instances bought and sold through the Reserved
Instance Marketplace work like any other Reserved
Instances, except that they can have less than a
full standard term remaining and can be sold at
different prices.

Provisioned IOPS SSD 2012-07-20 Provisioned IOPS SSD volumes deliver predictable, 31 July
for Amazon EBS high performance for I/O intensive workloads, 2012
such as database applications, that rely
on consistent and fast response times. For
more information, see Amazon EBS volume
types (p. 1084).

High I/O instances for 2012-06-15 High I/O instances provides very high, low latency, 18 July
Amazon EC2 disk I/O performance using SSD-based local 2012
instance storage.

IAM roles on Amazon 2012-06-01 IAM roles for Amazon EC2 provide: 11 June
EC2 instances 2012
• AWS access keys for applications running on
Amazon EC2 instances.
• Automatic rotation of the AWS access keys on
the Amazon EC2 instance.
• Granular permissions for applications running
on Amazon EC2 instances that make requests to
your AWS services.

Spot Instance features   You can now manage your Spot Instances as 7 June
that make it easier follows: 2012
to get started and
handle the potential of • Specify the amount you are willing to pay
interruption. for Spot Instances using Auto Scaling launch
configurations, and set up a schedule for
specifying the amount you are willing to pay
for Spot Instances. For more information, see
Launching Spot Instances in Your Auto Scaling
Group in the Amazon EC2 Auto Scaling User
Guide.
• Get notifications when instances are launched
or terminated.
• Use AWS CloudFormation templates to launch
Spot Instances in a stack with AWS resources.

EC2 instance export and 2012-05-01 Added support for timestamps on instance status 25 May
timestamps for status and system status to indicate the date and time 2012
checks for Amazon EC2 that a status check failed.

1403
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

EC2 instance export, 2012-05-01 Added support for EC2 instance export to Citrix 25 May
and timestamps in Xen, Microsoft Hyper-V, and VMware vSphere. 2012
instance and system
status checks for Added support for timestamps in instance and
Amazon VPC system status checks.

Cluster Compute Eight 2012-04-01 Added support for cc2.8xlarge instances in a 26 April
Extra Large instances VPC. 2012

AWS Marketplace AMIs 2012-04-01 Added support for AWS Marketplace AMIs. 19 April
2012

New Linux AMI release   Amazon Linux AMI 2012.03 is released. 28 March
2012

New AKI version   We've released AKI version 1.03 and AKIs for the 28 March
AWS GovCloud (US) region. 2012

Medium instances, 2011-12-15 Added support for a new instance type and 64-bit 7 March
support for 64-bit on information. Added procedures for using the Java- 2012
all AMIs, and a Java- based SSH client to connect to Linux instances.
based SSH Client

Reserved Instance 2011-12-15 Added a new section discussing how to take 5 March
pricing tiers advantage of the discount pricing that is built into 2012
the Reserved Instance pricing tiers.

Elastic Network 2011-12-01 Added new section about elastic network 21

Interfaces (ENIs) for EC2 interfaces (ENIs) for EC2 instances in a VPC. December
instances in Amazon For more information, see Elastic network 2011
Virtual Private Cloud interfaces (p. 839).

New GRU Region and   Added information about the release of new 14
AKIs AKIs for the SA-East-1 Region. This release December
deprecates the AKI version 1.01. AKI version 1.02 2011
will continue to be backward compatible.

New offering types for 2011-11-01 You can choose from a variety of Reserved 01
Amazon EC2 Reserved Instance offerings that address your projected use December
Instances of the instance. 2011

Amazon EC2 instance 2011-11-01 You can view additional details about the status 16
status of your instances, including scheduled events November
planned by AWS that might have an impact 2011
on your instances. These operational activities
include instance reboots required to apply
software updates or security patches, or instance
retirements required where there are hardware
issues. For more information, see Monitor the
status of your instances (p. 743).

Amazon EC2 Cluster Added support for Cluster Compute Eight Extra 14
Compute Instance Type Large (cc2.8xlarge) to Amazon EC2. November
2011

1404
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

New PDX Region and   Added information about the release of new AKIs 8
AKIs for the new US-West 2 Region. November
2011

Spot Instances in 2011-07-15 Added information about the support for Spot 11
Amazon VPC Instances in Amazon VPC. With this update, users October
can launch Spot Instances a virtual private cloud 2011
(VPC). By launching Spot Instances in a VPC,
users of Spot Instances can enjoy the benefits of
Amazon VPC.

New Linux AMI release Added information about the release of Amazon 26
Linux AMI 2011.09. This update removes the beta September
tag from the Amazon Linux AMI, supports the 2011
ability to lock the repositories to a specific version,
and provides for notification when updates are
available to installed packages including security
updates.

Simplified VM import 2011-07-15 The VM Import process is simplified with the 15

process for users of the enhanced functionality of ImportInstance September
CLI tools and ImportVolume, which now will perform 2011
the upload of the images into Amazon EC2
after creating the import task. In addition, with
the introduction of ResumeImport, users can
restart an incomplete upload at the point the task
stopped.

Support for importing VM Import can now import virtual machine 24 August
in VHD file format image files in VHD format. The VHD file format 2011
is compatible with the Citrix Xen and Microsoft
Hyper-V virtualization platforms. With this
release, VM Import now supports RAW, VHD and
VMDK (VMware ESX-compatible) image formats.
For more information, see the VM Import/Export
User Guide.

Update to the Amazon Added information about the 1.1 version of the 27 June
EC2 VM Import Amazon EC2 VM Import Connector for VMware 2011
Connector for VMware vCenter virtual appliance (Connector). This update
vCenter includes proxy support for Internet access, better
error handling, improved task progress bar
accuracy, and several bug fixes.

Enabling Linux AMI   Added information about the AKI version change 20 June
to run user-provided from 1.01 to 1.02. This version updates the 2011
kernels PVGRUB to address launch failures associated
with t1.micro Linux instances. For more
information, see Enabling Your Own Linux
Kernels (p. 196).

1405
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Spot Instances 2011-05-15 Added information about the Spot Instances 26 May
Availability Zone pricing Availability Zone pricing feature. In this release, 2011
changes we've added new Availability Zone pricing options
as part of the information returned when you
query for Spot Instance requests and Spot
price history. These additions make it easier to
determine the price required to launch a Spot
Instance into a particular Availability Zone.

AWS Identity and Added information about AWS Identity and 26 April
Access Management Access Management (IAM), which enables users to 2011
specify which Amazon EC2 actions a user can use
with Amazon EC2 resources in general. For more
information, see Identity and access management
for Amazon EC2 (p. 977).

Enabling Linux AMI Added information about enabling a Linux AMI to 26 April
to run user-provided use PVGRUB Amazon Kernel Image (AKI) to run a 2011
kernels user-provided kernel. For more information, see
Enabling Your Own Linux Kernels (p. 196).

Dedicated instances Launched within your Amazon Virtual Private 27 March

Cloud (Amazon VPC), Dedicated Instances are 2011
instances that are physically isolated at the host
hardware level. Dedicated Instances let you
take advantage of Amazon VPC and the AWS
cloud, with benefits including on-demand elastic
provisioning and pay only for what you use, while
isolating your Amazon EC2 compute instances
at the hardware level. For more information, see
Dedicated Instances (p. 496).

Reserved Instances Updates to the AWS Management Console make 27 March

updates to the AWS it easier for users to view their Reserved Instances 2011
Management Console and purchase additional Reserved Instances,
including Dedicated Reserved Instances. For more
information, see Reserved Instances (p. 325).

New Amazon Linux The new Amazon Linux reference AMI replaces 15 March
reference AMI the CentOS reference AMI. Removed information 2011
about the CentOS reference AMI, including the
section named Correcting Clock Drift for Cluster
Instances on CentOS 5.4 AMI.

Metadata information 2011-01-01 Added information about metadata to reflect 11 March

changes in the 2011-01-01 release. For more 2011
information, see Instance metadata and
user data (p. 704) and Instance metadata
categories (p. 723).

1406
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Amazon EC2 VM Import Added information about the Amazon EC2 VM 3 March
Connector for VMware Import Connector for VMware vCenter virtual 2011
vCenter appliance (Connector). The Connector is a plug-in
for VMware vCenter that integrates with VMware
vSphere Client and provides a graphical user
interface that you can use to import your VMware
virtual machines to Amazon EC2.

Force volume You can now use the AWS Management Console 23
detachment to force the detachment of an Amazon EBS February
volume from an instance. For more information, 2011
see Detach an Amazon EBS volume from a Linux
instance (p. 1123).

Instance termination You can now use the AWS Management Console 23
protection to prevent an instance from being terminated. February
For more information, see Enable termination 2011
protection (p. 645).

Correcting Clock Drift Added information about how to correct clock 25 January
for Cluster Instances on drift for cluster instances running on Amazon's 2011
CentOS 5.4 AMI CentOS 5.4 AMI.

VM Import 2010-11-15 Added information about VM Import, which 15

allows you to import a virtual machine or volume December
into Amazon EC2. For more information, see the 2010
VM Import/Export User Guide.

Basic monitoring for 2010-08-31 Added information about basic monitoring for 12
instances EC2 instances. December
2010

Filters and Tags 2010-08-31 Added information about listing, filtering, and 19
tagging resources. For more information, see List September
and filter your resources (p. 1306) and Tag your 2010
Amazon EC2 resources (p. 1311).

Idempotent Instance 2010-08-31 Added information about ensuring idempotency 19

Launch when running instances. For more information, September
see Ensuring Idempotency in the Amazon EC2 API 2010
Reference.

Micro instances 2010-06-15 Amazon EC2 offers the t1.micro instance 8

type for certain types of applications. For September
more information, see Burstable performance 2010
instances (p. 224).

AWS Identity and Amazon EC2 now integrates with AWS Identity 2
Access Management for and Access Management (IAM). For more September
Amazon EC2 information, see Identity and access management 2010
for Amazon EC2 (p. 977).

1407
 Amazon Elastic Compute Cloud
User Guide for Linux Instances
History for previous years

Feature API Description Release

version date

Cluster instances 2010-06-15 Amazon EC2 offers cluster compute instances for 12 July
high-performance computing (HPC) applications. 2010
For more information about the hardware
specifications for each Amazon EC2 instance type,
see Amazon EC2 Instance Types.

Amazon VPC IP Address 2010-06-15 Amazon VPC users can now specify the IP address 12 July
Designation to assign an instance launched in a VPC. 2010

Amazon CloudWatch Amazon CloudWatch monitoring is now 14 June

monitoring for Amazon automatically available for Amazon EBS volumes. 2010
EBS Volumes For more information, see Amazon CloudWatch
metrics for Amazon EBS (p. 1252).

High-memory extra 2009-11-30 Amazon EC2 now supports a High-Memory 22

large instances Extra Large (m2.xlarge) instance type. For more February
information about the hardware specifications for 2010
each Amazon EC2 instance type, see Amazon EC2
Instance Types.

1408