PROVISIONING WITH EMC VPLEX

INTEGRATED ARRAY SERVICES

Planning, Implementation, and Best Practices

ABSTRACT
This paper is intended to give guidance to the planning, implementation, and best
practices for using VPLEX Integrated Array Services (VIAS) and SMI-S Provider
functionality in your VPLEX environment.

EMC WHITE PAPER June 2017

1
Copyright © 2017 Dell Inc. or its subsidiaries. All Rights Reserved.
Dell Inc. believes the information in this publication is accurate as of its
publication date. The information is subject to change without notice.
The information in this publication is provided “as is.” Dell Inc. makes no
representations or warranties of any kind with respect to the information in this
publication, and specifically disclaims implied warranties of merchantability or
fitness for a particular purpose.
Use, copying, and distribution of any Dell EMC software described in this
publication requires an applicable software license.

Provisioning with EMC VPLEX Integrated Array Services

Planning, Implementation, and Best Practices

Part Number H13049

2
TABLE OF CONTENTS

EXECUTIVE SUMMARY5
Business Case5
Solution Overview5
Audience5

VPLEX INTEGRATED STORAGE PROVISIONING6

VIAS REQUIREMENTS7

VIAS RECOMMENDATIONS7

POOLS AND VIAS INTERACTION8

INSTALLING THE SMI-S PROVIDER9

SMI-S Configuration Suggestions9
SMI-S Post-Installation Steps For VIAS9
SMI-S Firewall Considerations For VIAS12

SETTING ECOM SECURITY13

REGISTERING THE ARRAY MANAGEMENT PROVIDER (AMP)14

Unregistering an AMP17
Registering Redundant AMPs17
Updating Provider Credentials17
Listing SMI-S Pools17

VIAS PROVISIONING17
General Provisioning LIMITS17
Distributed Storage Views18
Creating Virtual Volumes18
Provisioning From Pools18
Provisioning From Storage Volumes22
Provisioning with the VPLEX CLI26
Provisioning requests Jobs28
Monitoring Provisioning Jobs28
Using Hierarchy Maps29
Unclaiming Storage Volumes Created with VIAS30

APPENDIX:30
Business Communication Layer Service (BOLE) 30
Log Files for VIAS30

3
 VPLEX 5.3.0.00.00.10 Limitations31

Figure 1 - Architectural Overview6

Figure 2 - SMI-S Providers7
Table 1- Array and Provider Requirements .................................................... 7
Figure 3 - Downloading SMI-S Provider................................................................ 9
Figure 4 – Using TestSMIProvider to Add VNX Systems ........................................ 10
Figure 5 – Discovering VNX Systems ................................................................. 11
Figure 6 - TestSMIProvider Display View ............................................................ 11
Figure 7 - Advanced Firewall Settings ................................................................ 12
Figure 8 - Creating Inbound TCP Firewall Rules ................................................... 12
Figure 10 - ECOM Administration Page............................................................... 13
Figure 9 - ECOM Login Screen .......................................................................... 13
Figure 11 - SSL Client Authentication ................................................................ 14
Figure 12 - Registering the AMP ........................................................................ 14
Figure 13 - Registering AMP with SMI-S or REST ................................................. 15
Figure 13 - Register AMP with SMI-S Wizard ...................................................... 15
Figure 15 - AMP Naming Conventions ................................................................ 15
Figure 15 - Provision from Pool Wizard .............................................................. 17
Figure 16 - Step One, Select a Consistency Group ............................................... 18
Figure 17 - Step Two, Select Protection Options .................................................. 18
Figure 18 - Step Three, Selecting Storage .......................................................... 19
Figure 19 – Step Four, Expose Storage to Host ................................................... 20
Figure 20 - Step Five, Review and Finish ............................................................ 20
Figure 21 - Step Six, Review the Results ............................................................ 20
Figure 22 - Step One, Select a Consistency Group ............................................... 22
Figure 23 - Step Two, Selecting protection Options ............................................. 22
Figure 24 - Step Three, Select Storage .............................................................. 23
Figure 25 - Step Four, Exposing Storage to Host ................................................. 23
Figure 26 - Step Five, Review and Finish ............................................................ 24
Figure 27 - Step Six, Reviewing the Results ....................................................... 24
Figure 28 - Provisioning Single Legged Local Volumes27
Figure 29 - Provisioning Two Legged RAID1 Volumes27
Figure 30 - Distributed RAID128
Figure 31 - Distributed Device Map ................................................................... 29
Figure 32 - Un-Claiming Wizard for VIAS ........................................................... 29
Figure 33 - The BOLE Command Set ................................................................. 30

4
EXECUTIVE SUMMARY
Business Case
The EMC® VPLEX™ family removes physical barriers within, across, and between data centers. VPLEX Local provides simplified
management and non-disruptive data mobility across heterogeneous arrays. VPLEX Metro provides data access and mobility
between two VPLEX clusters within synchronous distances. With a unique scale-up architecture, VPLEX’s advanced data caching
and distributed cache coherency provide workload resiliency, balancing, plus enables both local and remote data access with
predictable service levels.

Solution Overview
This white paper is designed to give technology decision-makers a deeper understanding of the VPLEX Integrated Array Services
in terms of the overall solution, planning, design, and features & functionality. The key technical considerations for implementing
VPLEX Integrated Array Services for provisioning in VPLEX and how to provision storage from Storage Pools provided from VMAX,
VNX, Unity, and XtremIO arrays. VIAS provisioning will create new virtual volumes on demand, with the required service levels,
and do this provisioning across multiple arrays in five easy steps.

Audience
This white paper is intended for technology architects, storage administrators, and system administrators who are responsible
for architecting, creating, managing IT environments that utilize EMC VPLEX technologies. The white paper assumes the reader is
familiar with VPLEX, VMAX, VNX, Unity and XtremIO technologies.

5
VPLEX INTEGRATED STORAGE PROVISIONING
The VPLEX Integrated Array Service (VIAS) feature enables VPLEX to provision storage for EMC VMAX, VNX, Unity and XtremIO
storage arrays directly from the VPLEX CLI, UI, and REST API. VPLEX uses Array Management Providers (AMPs) to streamline
provisioning, and allows you to provision a VPLEX virtual volume from a pool on the array.
The VIAS feature uses the Storage Management Initiative -Specification (SMI-S) provider to communicate with VMAX and VNX
arrays while using the REST provider to communicate with XtremIO and Unity arrays to enable storage provisioning. An AMP
must be installed and configured on the same storage area network (SAN) as the VPLEX, and both the VPLEX and the provider
must be configured for the same arrays. After the provider has been configured, it must be registered as an AMP with VPLEX.
This is accomplished using either the VPLEX CLI or Unisphere for VPLEX GUI. After the registration has completed, the managed
arrays are automatically detected and visible to VPLEX and provisioning of virtual volumes can be accomplished using disk
groups or storage pools visible from each array.

Provisioning in VPLEX has been streamlined and enhanced to provide a more efficient way to provision storage for EMC arrays.
Simplified provisioning uses a wizard to create new virtual volumes on demand, with the service levels needed, in five steps:
1. Select consistency groups and add storage

2. Select VPLEX protection services to apply to the production volume (none, local RAID-1, and distributed mirroring),
number of volumes and capacity per volume
3. Select the underlying storage array pools from which to draw storage

4. Expose storage to hosts – See Figure 1 below

5. Review and finish
Simplified provisioning enhances productivity by dramatically reducing the number of operational steps required to provision
integrated storage.

Figure 1 - Architectural Overview

Integrated storage refers to storage created through the VPLEX Integrated Services feature. This feature requires the use of
AMPs to leverage storage pools on the array. If your EMC array configuration includes storage pools and the array is supported
for use with VPLEX, you can integrate the array with VPLEX and provision storage from pools on the array through VPLEX.

In Figure 2, the SMI-S Providers are configured to see the same array VPLEX has already been connected to. Once this
configuration is done a user can register a SMI-S provider using AMP on the VPLEX. This is done by running the CLI or GUI

6
wizard. You will provide the SMI-S Provider’s IP Address and credentials in order to complete registration. At this point the
managed arrays with its disk groups and pools for provisioning will be visible in the VPLEX UI.

Figure 2 - SMI-S Providers

VIAS REQUIREMENTS

Array Requirements Provider Requirements

Array Code Family Target Code Solutions Enabler SMI-S Provider
VMAX 5784-5876 5876.272.177 7.62 5876.272.17
5977 5977.596.583 8.02 n/a
VNX 7.1 7.1.79.8 7.62 5876.272.17
8.1 8.1.6.101 7.62 5876.272.17
5.32 5.32.000.5.218 7.62 5876.272.17
5.33 5.33.006.5.102 7.62 5876.272.17
Xtremio 3.0 3.0 n/a n/a
4.0 4.0 n/a n/a
Unity 4.0 4.0.x n/a n/a
Table 1- Array and Provider Requirements

 Support for VMAX and VNX is available starting with Geosynchrony 5.3
 Support for Xtremio 4.0 is available starting with GeoSynchrony 6.0.1
 Support for Unity is available starting with GeoSynchrony 6.0.1

For the 5.5 and 6.0 GeoSynchrony releases, the max number of provisioning jobs that may be queued is 100. This does not
include cancelled or completed jobs. The maximum number of virtual volumes per request is 10. After one request is complete,
the next request can be initiated. During a provisioning request, it might take several minutes for the job ID to be returned. In
the GUI, the provisioning wizard will display an in progress icon until the job ID is ready. The same is applicable for the CLI, it
might take several minutes for the job ID to be displayed. This behavior has been greatly enhanced since the 5.3 and 5.4
versions of GeoSynchrony.

VIAS RECOMMENDATIONS
 As a best practice, do not use the RESTful interface to poll the pools on a managed array. This can be a very expensive
resource consuming operation. EMC recommends that you limit the number of times you query pools from any of the
interfaces. (CLI, GUI, and RESTful API)
 Avoid having redundant array providers managing the same arrays. It is possible to add many array providers.
Additional providers that are not used to actively manage an array should be removed.
 Listing pools is expensive on the SMI-S provider. While a provisioning request is in progress, try to limit the number of
times you list pools. This request can be synchronous and can cause delays in the response time for listing pools.

7
  Know the SMI-S provider health. Before executing a provisioning request, you might want to check the health of the
provider. This is best done by looking for the ECOM thread or service. There have been some cases where the SMI-S
provider take up a percentage of the CPU (25%) when there is no interaction. In such cases, the provider is having an
issue.
 If there are multiple masking views, VPLEX will place the provisioned LUNs in the masking view with the least number of
devices. In other words, provisioning is deterministic in terms of the masking view with the least number of devices.
 The minimum capacity that may be provisioned is 150MB.
 The automatic provisioning job deletion only deletes completed and failed jobs. The jobs are automatically deleted after
seven days. In-progress and canceled jobs must still be deleted manually.
 Local jobs will be executed even if the remote jobs fail on the remote cluster.
 In a scenario where one cluster fails, the distributed RAID1 provisioning requests will remain queued until the other
cluster is reachable.
 There are different timers that are executed at each level of a provisioning request. When a cluster fails or an AMP
becomes unreachable, a timeout occurs. In the event of a timeout, provisioning requests will fail with the message:
“futures timeout”
 The provisioning job status will persist through any SMS failures or reboots.
 If a provisioning job is in progress, and there is a VPN or SMS failure, the job listing on the remote cluster will not be
visible. If the jobs are executed on the remote cluster, then the provisioning job listings will remain empty until the VPN
is re-established or the remote cluster is again reachable.
 For VMAX, if a LUN is provisioned from a pool that has a specific policy, then the storage group in the masking view
must also have the same policy. During provisioning, when a LUN was created from a pool with a specific policy that the
masking view configured to VPLEX did not have, the LUN was created. However, the LUN could not be added to the
masking view and was not exposed to VPLEX.

 Before provisioning from storage pools, you must first register the AMP that manages the array. Your VPLEX system can
include AMPs that manage some or all of the arrays in your environment. An array must be integrated with VPLEX in
order to provision storage from pools on the array. Note that you can also provision from storage volumes as well as on
an integrated array.
Note: VIAS uses the Storage Management Initiative - Specification (SMI-S) provider to communicate with the VMX and VNX
arrays. Please validate your versioning requirements with the EMC Simple Support Matrix for VPLEX.

POOLS AND VIAS INTERACTION

It is important to remember for the end user of VIAS to ensure they have some insight into the pools from which they will be
provisioning. The VPLEX VIAS functionality will show all the non-restricted pools on the array. VIAS has implemented capabilities
that try to reduce the possibility of showing invalid pools but there are scenarios that can arise with unsupported array
configurations that may prevent provisioning. A user should know how the VPLEX is configured to the array and the types of
pools that should be used. The following is an example of different types of pools:
 Primordial Storage Pool: A primordial Storage Pool is a type of Storage Pool that contains unformatted, unprepared,
or unassigned capacity. Storage capacity is drawn from the primordial Storage Pool to create concrete Storage Pools. A
primordial Storage Pool is not shown after registering with VPLEX AMP because provisioning is not permitted from these
types of pools.
 Concrete Storage Pool: A concrete Storage Pool is comprised of VMAX Disk Groups or VNX RAID Groups. They may
aggregate the capacity of one or many these Disk and/or RAID groups. Concrete Pools are shown after VPLEX AMP
registration.
 Reserved Disk Pools: These are Storage Pools reserved by the storage array for hot sparing.
 Reserved LUN Pools: These are Storage Pools reserved by the storage array for Snaps and Clone technologies.
Provisioning requests from VPLEX may fail with generic ECOM errors that are due to limitations and restrictions on the array. If
the provisioning request fails with an error please check the pool configuration and how VPLEX is configured. The generic ECOM
error messages are displayed when provisioning fails and are usually due to array configuration issues and need to be
investigated on the provider and the log files. The log files are located here: “C:\Program Files\EMC\ECIM\ECOM\log”
8
INSTALLING THE SMI-S PROVIDER
The SMI-S provider is a key component of the VIAS service. It is used by VPLEX to communicate with VNX and VMAX arrays to
enable storage provisioning. An SMI-S provider needs to be installed and configured correctly on the SAN that is going to
interact with the arrays that VPLEX is also connected to. The assumption is that the SMI-S setup and configuration has already
been done by an administrator that has knowledge on how to configure the SMI-S provider.

Figure 3 - Downloading SMI-S Provider

SMI-S CONFIGURATION SUGGESTIONS
The following are some best practices for the installation and configuration of the SMI-S providers:
 Validate your VNX, VNX2, VMAX and VMAX3 requirements with the ESSM prior to installation.
 Each generation of array may require different versions of SMI-S.
 The gatekeepers should NOT be shared between hosts, instead they should be unique per host.
 Ensure that the SMI-S Provider server does not lose connectivity to the Symmetrix system(s) under management.
 Make sure that the SMI-S provider is reachable from the VPLEX management server with the given IP address and
port number. Firewalls and other network security systems need to be configured to allow access from VPLEX to the
SMIS-S provider and vice versa.
 Before registering a provider to manage an array it is a good practice to make sure that both VPLEX and the SMI-S
provider can access the managed array(s).

SMI-S POST-INSTALLATION STEPS FOR VIAS

This section explains the post-installation tasks that you must complete.
- VMAX Arrays: When using the SMI-S Provider to manage Symmetrix arrays, it is recommended that you configure six
gatekeepers for each Symmetrix array accessed by the provider. Only set up these gatekeepers for the host on which the
SMI-S Provider is running. When started, the SMI-S Provider automatically discovers all Symmetrix storage arrays
connected to the host on which the Array Provider is running. No other action is required, such as a running the symcfg
discover command.
- SYMAPI_DB.BIN Database Sharing: When the SMI-S Provider is installed on the same host as the ECC Symmetrix
agent and/or the Unisphere for VMAX 1.0, you may see the following memory allocation errors in the syampi log file:
EMC:SMBASE __iload_db_osl pdsDbRecRead() failed : OSL:CONN_INFO ([PDS/DB] (Unable to allocate memory)
EMC:SMBASE emcSymDBLoad Error encountered while reading from DB file
[C:\Program Files\EMC\SYMAPI\db\symapi_db.bin] (SYMAPI_C_MEMORY_ALLOC_ERROR)

The factors determining these memory allocation errors are governed by the amount of physical memory on the host as well
as the number and size of the array configurations. Because it is difficult to predict how much memory is required for this
type of installation scenario, perform the following steps to prevent the above errors from occurring:
 Instruct SMI-S Provider to use its own symapi database by editing the following file:
C:\Program Files\EMC\ECIM\ECOM\providers\oslsprovider.conf

9
  Change the following line in oslsprovider.conf:
#OSLSProvider/com.emc.cmp.osls.se.array.StorApi.database.filename =
to:
OSLSProvider/com.emc.cmp.osls.se.array.StorApi.database.filename = c:/program
files/emc/symapi/db/symapi_smi_db.bin
 Stop ECOM, the ECC Symmetrix agents, Unisphere for VMAX 1.0, and the Solutions Enabler daemons.
 Remove the existing symapi_db.bin file, then restart ECOM, the ECC Symmetrix agents, Unisphere for VMAX 1.0, and
the Solutions Enabler daemons.

- VNX Arrays: SMI-S Provider discovers CLARiiON and VNX storage arrays using one of the following discovery
methods:
In-band discovery: This is conceptually the same as the Symmetrix SMI-S Provider-based discovery. If the correct
information is in place, SMI-S Provider can automatically discover a CLARiiON array or VNX storage system during its
startup. For SMI-S Provider to automatically discover locally-attached CLARiiON or VNX arrays, you must add
authorization information for its storage processors. A locally-attached storage array is one in which at least one
CLARiiON or VNX LUN is visible to the host on which the SMI-S Provider is running. Use the SYMCLI symcfg command,
as shown in the following syntax example, to add the username and password for each storage processor:
symcfg authorization add -host <HostName> -username <UserName> -password <PassWord>

Out-of-band discovery: You can add a VNX array to the SMI-S Provider without a VNX LUN being present on the host
on which SMI-S Provider is installed. Typically, this method is preferred over in-band discovery. SMI-S Provider has a
programmatic interface that provides management applications integrated with the provider the ability to discover VNX
storage arrays out of band. This discovery method only requires an IP connection to the storage array.
- Modify PATH Environment Variable:
set PATH=%PATH%;”C:\Program Files\EMC\SYMCLI\bin;C:\Program Files\EMC\ECIM\ECOM\bin”

Figure 4 – Using TestSMIProvider to Add VNX Systems

10
Figure 5 – Discovering VNX Systems

After connecting to the SMI-S Provider, you may issue the “dv” (display view) command to list all of the basic provider
information and each storage array that has been attached or remotely registered as seen below.

Figure 6 - TestSMIProvider Display View

11
SMI-S FIREWALL CONSIDERATIONS FOR VIAS
This section explains the required firewall ports that must be opened for SMI-S to communicate with VPLEX VIAS. If a firewall
exists between the SMI-S Provider installation and a VNX array, port 443 must be opened in both directions in the firewall for
management communications to occur with the array. By default, the ECOM server listens on ports 5988 (for HTTP) and 5989
(for HTTPS) and must be able to communicate both directions as well as port 443. This may be done by opening Server

Figure 7 - Advanced Firewall Settings

12
Figure 8 - Creating Inbound TCP Firewall Rules
Manager in Windows and selecting “Advanced Firewall Settings.” From there create both an inbound and outbound TCP/UDP
rules for ports 5988 and 5989 as seen in Figure-7.

SETTING ECOM SECURITY

The EMC provider security can be configured through a webpage. Go to https://<sms-i_provider>:5989/ecomconfig and login
with the default account (admin) and password (#1Password).

Figure 9 - ECOM Login Screen

Figure 10 - ECOM Administration Page

13
Next, proceed to change the password, add an additional user or make any other changes to the security as desired. Note the
user name and password since you will need this when you register the provider for use with the Storage Service.

There is one more change that we will need to make. Click on the Dynamic Settings link from the ECOM Administration Page and
locate the setting for SSLClientAuthentication. Select “None” and check the “Persist” box, then “Apply” these settings. This
avoids a potential problem with SSL negotiations without lowering the security level. You will need to restart the ECOM service if
you modify parameters on this page. (“net stop ecom” –and- “net start ecom”)

ARRAY MANAGEMENT PROVIDER (AMP)

VPLEX's Integrated Array Services feature uses array management providers (AMPs) to expose functionality on supported arrays.
Specifically, AMPs allow VPLEX to provision storage from predefined storage pools on the arrays. To use the Integrated Array
Services feature in VPLEX, you must register the AMP with VPLEX and then use the CLI or the Provision from Pools wizard in the
GUI to provision virtual volumes from pools on the managed arrays. For an AMP to manage an array, the array must be visible to
the AMP and to the VPLEX cluster on which you are registering the AMP. An AMP can manage multiple arrays and is registered to
the specific cluster.
VPLEX supports two types of AMPs:
 VPLEX RESTful Interface (REST) — Manages XtremIO and Unity arrays.

 SMI-S — manages all other arrays.

There is no restriction on the type of SMI-S AMP that can be registered. If the username, password, IP address and port are
accepted during the registration process, then the AMP can be registered with VPLEX. Once the AMP is successfully registered, it
appears on the Array Management Provider view of the cluster. Select an AMP to see the arrays that it manages, as well as the
storage pools and storage groups available from each of the managed arrays.

ADDITIONAL RESOURCES
 For a list of arrays supported for use with VPLEX's Integrated Array Services, refer to the EMC VPLEX and
GeoSynchrony Simple Support Matrix document, available on EMC Online Support.
 To correctly install and configure a SMI-S Provider, refer to the EMC SMI-S Provider Release Notes, available on
EMC Online Support.
 To configure and use the REST provider, refer to the VPLEX Element Manager API Guide.
To use VIAS in VPLEX, you must register an array management provider, and then use the GUI provisioning wizard or the CLI to
provision virtual volumes. In our example, we will use the VPLEX GUI to show the following tasks needed to provision storage:
 Register the array management provider.
 Register the initiators that access VPLEX storage.
 Create a storage view that include virtual volumes, initiators, and VPLEX ports
o Select the storage array and storage volumes to create virtual volumes.

Figure 11 - SSL Client Authentication

14
Figure 12 - Registering the AMP

15
After the array management provider is registered, VPLEX correlates the arrays to identify which arrays it is connected to and
which arrays that VPLEX can also manage. There is no restriction on the number of SMI-S providers that can be registered. If the
credentials passed during registration are valid and the provider is IP reachable the AMP can be successfully registered. The
username, password, IP, and port all must be valid in order to register the amp. It is best to validate that the credential are
correct using the ECOM Explorer before registering a provider. This way you can ensure the SMI-S configuration is correct. If the
registration fails the AMP is not persisted and password is not saved.
 Passwords are stored in the lockbox on each management server.
 Passwords are not stored in any log file or in capture logs.

 A provider can be registered from either management server.

Register AMP
Figure 13 - Registering with
AMP SMI-S
with Wizard
SMI-S or REST
 If multiple registered AMPs manage the same storage array, VPLEX will display the managed array only once under the
AMP chosen by VPLEX to do provisioning operations. It is recommended (but not required) that an array be managed by
only one registered AMP.

A managed array is an array that is both visible by SMI-S and registered with VPLEX. If an array is visible by the SMI-S provider
and not registered with VPLEX, it will not be considered managed. You do not need to un-register and re-register your SMI-S
provider to force updates when adding new arrays. If the configuration is setup correctly, the GUI will poll every 3 minutes and
automatically discover when new arrays are added.
Each managed array has a list of pools associated with it. In order to provision, you must know the pool you want to provision
from. Once AMP has been registered, the SMI-S provider will list all Disk Groups and Pools from your VMAX and VNX arrays so
it’s a good idea to create a naming convention that will make it easy to identify your tiered storage pools which are intended for
VPLEX provisioning.

Figure 15 - AMP Naming Conventions

16
UNREGISTERING AN AMP
Before unregistering an AMP, ensure that all arrays associated with the provider are no longer in use and are unregistered. After
the AMP is unregistered, VPLEX removes all references. The only restriction is that the AMP cannot be unregistered if there is a
provisioning or de-provisioning action taking place.

REGISTERING REDUNDANT AMPS

You cannot register the same AMP on a cluster. It is possible to register an AMP that is managing the same arrays as an AMP
already registered. This means you would essentially have two AMPs that are managing the same arrays. VIAS will only list
managed arrays for one of the two AMPs. The managed arrays will show up under the AMP chosen by VPLEX to do provisioning.
The other AMP won’t display the same managed arrays.

UPDATING PROVIDER CREDENTIALS

There is no specific action to update the array management provider’s credentials if they change on the back end array. The user
will have to unregister and the register the AMP again in VPLEX.

LISTING SMI-S POOLS

Not all the pools from the array may be listed-- only pools that are unrestricted and are capable of provisioning LUNs. This
means that the AMP will only try and list the pools that are capable of provisioning. There are some pools on the arrays that are
not possible to provision a LUN from and these pools we try to prevent from being displayed in the list. For example primordial
pools. Primordial storage refers to unallocated storage capacity on a storage device. Storage capacity can be allocated from
primordial pools to create storage pools (TDATs/SAV). This means that primordial and restricted pools are disk/device sources
for allocation of storage pools and will not be listed because they cannot have LUNs directly provisioned from them.

VIAS PROVISIONING
Provisioning is available through both the CLI and the GUI. Please see http:support.emc.com for the most current VPLEX
Administration Guide and VPLEX CLI Guides.

GENERAL PROVISIONING LIMITS

When provisioning is performed there are a few details that are worth noting.
 Maximum number of Virtual Volumes that can be provisioned on a single request is 10.
 The minimum capacity that is capable of being provisioned is 150 MB.
 An array rediscovery is needed to make the LUNs visible to VPLEX. The rediscovery is a targeted rediscovery on the
array and only done once on each array after all LUNS have been created.
 One can have multiple storage groups/masking views on the underlying arrays. Placement will be done automatically,
determined by the lowest volume count at the time the provisioning request is processed. A subsequent provisioning
request may result in a different storage group/masking view being used.
 Cascaded storage groups for VMAX are not supported at this time.
 For VMAX, if a LUN is provisioned from a pool that has a specific policy then the storage group in the masking view
needs to also have the same policy.
 In order to enforce HA a different array for each cluster is enforced.
 When provisioning a virtual volume greater than 240 GB on VMAX, ensure that the auto Meta configuration flag is set
on the array. Please use the VMAX admin guide to determine the best value for the environment that the VMAX is in.
 For Unity, only one volume may be provisioned at a time in versions X/Y/Z. In GeoSynchrony 6.0 Service Pack 1, this
restriction is removed.
 If a VIAS provisioning job fails on a Unity array with this error, it is likely that the array LUN limit has been reached:
Failure provisioning the storage volume(s), cause: An error occurred during the 'Provisioning a
Volume from a Storage Pool' operation in the 'CreateOrModifyElementFromStoragePool' method. Return
code =4 The rollback was successful. No LUNs provisioned on array, no LUNs exposed to VPLEX.

17
DISTRIBUTED STORAGE VIEWS
Distributed storage refers to storage objects (consistency groups, virtual volumes, devices) that are created by using storage
from both clusters. All distributed storage objects can be conveniently accessed from the Distributed menu option in the
Provision Storage hover menu. This menu option opens the Distributed storage section in the navigation pane on the left. From
this section you can view a grouping of all distributed consistency groups, distributed virtual volumes and distributed devices.

CREATING VIRTUAL VOLUMES

You can use the VPLEX GUI or the CLI to provision virtual volumes. However, it is easier to provision using the GUI.
There are a few considerations:
 Only one provisioning request can be executed at a time on a VPLEX cluster.
 The maximum number of virtual volumes that can be provisioned using a single request is 10.
 The minimum capacity that can be provisioned is 150 MB.

PROVISIONING FROM POOLS

Using the VPLEX GUI to Provisioning Storage from the “Provision from Pools” wizard.

Figure 15 - Provision from Pool Wizard

Figure 16 - Step One, Select a Consistency Group
Using the VPLEX GUI to Provisioning Storage from the “Provision from Pools” wizard.

18
Step One: For the volume to be created, select an existing consistency group or create a new one. To add the virtual volume to
an existing consistency group, select the Add storage to existing consistency group radio button and then select the consistency
group in the list. To filter the list of consistency groups by name, type the full or partial name of the consistency group in the
Filter text box. To filter by storage location, click the drop down box to the right of the Filter text box, and then select the
storage location. To create a new consistency group for the virtual volume, select the Create new consistency group radio button

Figure 17 - Step Two, Select Protection Options

Step Two: Select mirroring on the cluster where the virtual volumes will be created (optional), and then enter the quantity, size
and base name for the virtual volumes.
 High Availability Options: Select the check box to enable mirroring at the cluster indicated. The option to mirror at a
cluster is disabled if you have less than two arrays in the cluster.
o For a local consistency group, select the check box to enable mirroring at the local cluster.
o For a distributed consistency group, select the appropriate check box to enable mirroring at one cluster or both
clusters, or leave the check boxes unselected to disable mirroring.
 Number of Volumes: The number of volumes to create. You can create a maximum of 10 volumes at a time. Each
volume will have the same RAID geometry.
 Capacity per Volume: The size of each volume. Select the size in MB, GB or TB. The minimum size is 150MB. Each
volume will have the same capacity.

 Total Capacity: The total capacity of all the virtual volumes to be created.
 Volume Base Name: Type a base name for the virtual volumes. A number will be added to the base name for each
volume created. For example, using the base name Test, volumes will be named Test_1, Test_2, Test_3, and so forth.
Note: If a base name has been previously used with a VIAS Provisioning job, then it cannot be re-used when you desire to provision
additional LUNs to a Storage View or set of Hosts.

19
Figure 18 - Step Three, Selecting Storage
Step Three: Select a storage pool to provide storage for the virtual volumes to be created. Click the drop down box and select
an array, and then select a pool from the list of available pools on the array. The following information is shown for each
available pool:
 Pools: The name of the pool on the array.
 Type: Indicates if the pool is thickly- or thinly-provisioned on the array. If you select a thin pool, a thin rebuild will be
automatically set for the storage volume that is created during the provisioning process.
 Free: The amount of storage available for provisioning from the pool.
 Total: The total amount of storage available from the pool.

Step Four: In Figure-19 we select a storage view for the virtual volume, or select “None” to continue and select a storage view
later. If you select an active storage view, the virtual volume is added to that storage view and automatically becomes visible to
hosts with access to the storage view. In some cases a rescan of the host may be required (depending on hba and multipathing
versions) If the storage view is inactive, go to the Storage Views screen and ensure that the components necessary to activate
the storage view (hosts initiators, vplex fe-ports, storage volumes). Should you select “None”, you must manually add the virtual
volume to a storage view when ready.

Figure 19 – Step Four, Expose Storage to Host

20
Step Five: In Figure-20 we will Review and Finish our selections. This page shows a summary of your selections in the wizard
and a topology map of the volume to be created. Review your selections, and if satisfied, click Finish to submit the provisioning
request. To make changes, click Back to go back to a previous step and make the necessary changes.

Figure 20 - Step Five, Review and Finish

Step Six: Reviewing the Results. This page shows the results of the provisioning request. Each provisioning request creates a
provisioning job that runs in the background. To monitor the job, click View Job Status in the lower-right corner of the screen to
go to the Provisioning Jobs page. Otherwise, click Close to close the wizard

Figure 21 - Step Six, Review the Results

21
PROVISIONING FROM STORAGE VOLUMES
The Provision from Storage Volumes wizard allows you to provision a virtual volume directly from a storage volume or preserve
data on an existing storage volume that you want to expose to hosts. The wizard simplifies the provisioning process by
automatically claiming storage (if not already claimed) and creating all of the underlying storage objects (extents and devices),
and then creating a local or distributed virtual volume that is the total capacity of the selected storage volume. When
provisioning from storage volumes, you can create only one virtual volume at a time. Each virtual volume created maps to a
storage volume on the array. You can provision from storage volumes using integrated or non-integrated storage arrays. Refer
to Provisioning Overview for more information on integrated and non-integrated storage.
Creating a new virtual volume is as simple as selecting or creating a consistency group for the volume, selecting mirroring
options (optional), and then selecting an available storage volume from an array on the cluster on which you want to create the
virtual volume. You do not need to claim the storage volume before you begin to provision, and you can select claimed or
unclaimed storage volumes from the selected array. Unclaimed storage volumes will be automatically claimed.

You may launch the Provision from Storage Volumes wizard from any of the below listed locations. Depending on where you
launch the wizard, you are presented with the option to create specific types of consistency groups.
 Provisioning storage hover menu (Local or distributed)
 Distributed Virtual Volumes screen (Distributed only)
 Virtual Volumes screen on a specific cluster (Local only)
 Storage Volumes screen on a specific cluster (Local only)

The Provision from Storage Volumes wizard consists of the following steps:
1. Select an existing consistency group for the volume or create a new consistency group
2. Select mirroring options, synchronization options, and provide a name for the virtual volume
3. Select a storage volume to use to create the virtual volume
4. Expose the virtual volume to hosts by selecting a storage view
5. Review your selections and submit the provisioning request
6. View the results of the operation

Step One: Select a consistency group for the virtual volume. For the volume to be created, select an existing consistency group
or create a new one. To view the topology of a consistency group, select the group in the list. In the map that displays to the
right, the storage icon and blue shading around a cluster indicate that the selected consistency group uses storage from that
cluster.
 To add the virtual volume to an existing consistency group, select the Add storage to existing consistency group radio
button and then select the consistency group in the list.

 To filter the list of consistency groups by name, type the full or partial name of the consistency group in the Filter text
box. To filter by storage location, click the drop down box to the right of the Filter text box, and then select the storage
location.
 To create a new consistency group for the virtual volume, select the Create new consistency group radio button.

22
Figure 22 - Step One, Select a Consistency Group

Step Two: Select protection options. Select high availability options, synchronization options, and enter a name for the volume.
If a cluster has only one array, the ability to mirror is disabled for that cluster.
 Source Cluster: Select the cluster where you want to create the virtual volume.
 High Availability Options: Select a check box to enable mirroring at the cluster indicated. You can select a single
cluster, both clusters, or leave both check boxes unselected to disable mirroring at both clusters. When you select the
check box to enable mirroring at a cluster, Local Mirroring appears above the cluster in the map.
 Synchronization Options: Select the appropriate option to indicate whether or not you want to synchronize data from
the source storage volume to the target storage volume.
 Volume Name: Type a name for the virtual volume

Figure 23 - Step Two, Selecting protection Options

23
Step Three: Select source storage from cluster-1 and cluster-2 volumes to create the virtual volume. Select a single storage
volume to create the virtual volume. The virtual volume will use the entire capacity of the selected storage volume. You can
select claimed or unclaimed storage volumes. To select the source storage on cluster-1, select an array in the drop down box,
and then select a storage volume from the list of available storage volumes on the array. Repeat process for cluster-2.

Figure 24 - Step Three, Select Storage

Step Four: Expose the virtual volume to hosts. Select a storage view for the virtual volume, or select “None” to continue and
select a storage view later. If you select an active storage view, the virtual volume is added to that storage view and
automatically becomes visible to hosts with access to the storage view. In some cases a rescan of the host may be required. If
the storage view is inactive, go to the Storage Views screen and add the components necessary to activate the storage view. For
more information, refer to Creating a storage view. If you select “None”, when you are ready to expose the virtual volume to
hosts, you must manually add the virtual volume to a storage view.

Figure 25 - Step Four, Exposing Storage to Host

24
Step Five: Review your selections and finish. This page shows a summary of your selections in the wizard and a topology map
of the volume to be created. Review your selections, and if satisfied, click Finish to submit the provisioning request. To make
changes, click Back to go back to a previous step and make the necessary changes.

Figure 26 - Step Five, Review and Finish

Step Six: View results. This page shows the results of the provisioning request. The top portion of the page shows the success
or failure of the operation and the details. Select the row in the top portion of the page to see the expanded details in the Details
view at the bottom of the page. You can scroll through the detailed information and cut and paste information if needed. Click
Close to close the wizard when done.

Figure 27 - Step Six, Reviewing the Results

25
PROVISIONING WITH THE VPLEX CLI
The “storage-tool compose” command will create virtual-volumes on top of the specified storage-volumes, building all
intermediate extents, local, and distributed devices as necessary.

Syntax:
storage-tool compose
[-n|--name] name

[-g|--geometry] {raid-0|raid-1|raid-c}
[-d|--storage-volumes] storage-volume [, storage-volume...]
[-s|--source-storage-volume] storage-volume

[-h|--help]
[--verbose]
Optional Arguments:
[-n|--name] name - * Specifies the name for the new virtual volume. Must be unique across the system.
[-g|--geometry] {raid-0|raid-1|raid-c} - * Specifies the desired geometry to use for the local
devices at each cluster. Valid values are raid-0, raid-1, or raid-c.
[-d|--storage-volumes] storage-volume [, storage-volume...] - * Specifies a list of storage volumes
from which to build the virtual volume. These may be claimed, but must be unused.
[-s|--source-storage-volume] storage-volume - Specifies the storage volume to use as a source mirror
when creating local and distributed devices.
Note: If specified, --source-storage-volume will be used as a source-mirror when creating local and
distributed RAID 1 devices. This will trigger a rebuild from the source-mirror to all other mirrors
of the RAID 1 device (local and distributed). While the rebuild is in progress the new virtual volume
(and supporting local and/or distributed devices) will be in a degraded state, which is normal. This
option only applies to RAID 1 local or distributed devices. The --source-storage-volume may also
appear in --storage-volumes.
[-h|--help] - Displays command line help.
[--verbose] - Provides more help during command execution. This may not have any effect for some
commands.

26
Figure 28 - Provisioning Single Legged Local Volumes

Figure 29 - Provisioning Two Legged RAID1 Volumes

27
Figure 30 - Distributed RAID1

PROVISIONING REQUESTS JOBS

Provisioning requests will be executed as a job. Since provisioning can take some time a provisioning request is executed in an
asynchronous manner. The command prompt will return direct to the CLI so the user can continue to perform administrative
tasks. The provisioning job and its status will be identified by a job id. The job id will be returned after a provisioning command
is executed. Note about provisioning jobs:

 The provisioning job status will persist through any SMS failures or reboots.
 Queued jobs may be cancelled, but an in-progress job cannot be.
 Local jobs will be executed even if the remote jobs fail on the remote cluster.

 In a scenario where one cluster fails, the distributed RAID1 provisioning requests will remain queued until the other
cluster is reachable.
 Completed or failed jobs are automatically deleted after seven days. In-progress and canceled jobs must be deleted
manually.



MONITORING PROVISIONING JOBS

The Provisioning Jobs Status page allows you to view the status of a provisioning job. The Provision from Pools wizard allows you
to create multiple virtual volumes which can take some time. Therefore, when you make a provisioning request through the
wizard, a provisioning job is automatically created and runs in the background while you continue to perform other tasks. You
cannot perform any actions on an in-progress job except to monitor its progress. Queued jobs may be cancelled. The status page
shows the Job name, status, description, cause of failure when applicable, and the last status update time for each job. A
provisioning job can be in one of the following states:

Queued: The job is awaiting processing.

In Progress: The job is in progress.
Completed: The job has completed.

28
Failed: The job has failed. Depending on which step in the process that the failure occurred, you may need to take action to
manually rollback any components that were created. Refer to when a provisioning job fails for more information. The job status
is automatically updated every 10 seconds. However, you can manually refresh the job status at any time by clicking the refresh
icon in the upper right corner of the screen. When the job completes, a record of the job stays on the Provisioning Jobs Status
page for 48 hours, after which time it is automatically deleted.

A single provisioning job is created for each provisioning request, regardless of the number of volumes being provisioned. When
a provisioning job starts, the job is given a name that includes information about the job such as job number and the date and
time the job started. The following is the format and description of a job name:”Provision_#_dd-mm-yy:time”

Breakdown of Name:
“Provision_” is a prefix added to all job names.
“#” is the job number. This number increments with each provisioning job.

“dd-mm-yy” is the date the job started.

“time” is the time the job started.

USING HIERARCHY MAPS

Storage hierarchy maps allow you to select a storage object whether provisioned by VIAS (or not) and view a
graphical display of all other objects connected to the selected object. This includes underlying objects, as well as top
level objects that use the selected object. You can then drill down further to see the health and properties of each
connected object. When a map is displayed, you can print, export, or search for objects.
Distributed devices and distributed virtual volumes appear outside of both clusters because they are not associated
with a specific cluster, but instead spans both clusters. A virtual volume on top of a distributed device appears in
both clusters and may appear in different storage views depending on the view membership on either cluster.

Figure 31 - Distributed Device Map

29
UNCLAIMING STORAGE VOLUMES CREATED WITH VIAS
When unclaiming storage volumes created with VIAS, you have the additional option to delete the storage volume from the array
and return the allocated storage volume capacity to the storage pool on the array.

Figure 32 - Un-Claiming Wizard for VIAS

 Note that returning allocated storage volume capacity to the storage pool on the array can take some time.

APPENDIX:
BUSINESS COMMUNICATION LAYER SERVICE (BOLE)
The new array aware provisioning and registration commands use a new service that communicates across the clusters. This
service is started by default and intended to be a zero touch service. When the SMS starts up the business communication layer
service (BOLE) will be started automatically.
The BOLE Service commands:

Figure 33 - The BOLE Command Set

LOG FILES FOR VIAS

There are two additional log files that are created. The bole.log and via.log files are located with all the other VPLEX log files in
the “/var/log” location. These log files contain additional information about the VIAS capability.
 BOLE log will have log information regarding the business layer and its interaction.

 VIA log will have specific information about the SMIS provider and direct calls to it from AMP.
Please keep in mind that some errors will happen during provisioning that are due to array restrictions and configurations that
VIAS may not enforce. This means that the native log files on the array may need to be looked at to identify issues.

30
VPLEX 5.3.0.00.00.10 LIMITATIONS


 VMAX, VNX, and XtremIO support only
 DMX not supported

 VNX requires FLARE 32 or higher

 SMI-S / Solutions Enabler 4.6.1.2 (Provider)/ 7.6-1755 (SE)
 Pool listing (GUI/CLI/REST) times can be impacted by ongoing provisioning taskjobs.

31