Memory virtualisation

CSE 598B: Self-* Systems
Memory Resource Management in
VMware ESX Server
by Carl A. Waldspurger
Presented by: Arjun R. Nath
(slide material adapted from C. Waldspurger, and M. Behar)

2
Summary of this Presentation
What is VMware ESX server ?
Virtualization
Memory management techniques employed by
ESX server
– Ballooning
– Memory sharing
– Reclaiming idle memory
Other stuff – similar products, etc.

3
ESX server overview
 Thin kernel designed to run VMs
 Multiplexes hardware resources - virtualizes the Intel IA-32
architecture
 Manages system hardware for high-performance I/O
 Runs unmodified commodity operating systems

4
Virtualization
 Virtualization enables the running of multiple
operating systems on a single machine
 Each Virtual Machine (VM) is isolated and
protected from each other - Illusion of dedicated
physical machine
 Allows an abstraction of server workloads
 Motivation:
– Take advantage of idle machine time
– Easy to maintain and upgrade VMs

6
Memory Virtualization
 Guest OS needs to see a zero-based memory
space
 Terms:
– Machine address -> Host hardware memory space
– “Physical” address -> Virtual machine memory
space

7
Memory Virtualization
 Translation from MPN (machine
page numbers) to PPN (physical
page numbers) is done thru a
pmap data structure for each VM
 Shadow page tables are
maintained for virtual-to-
machine translations
– Allows for fast direct VM to Host
address translations
 Easy remapping of PPN-to-MPN
possible transparent to VM

9
Memory Reclamation
 Each VM gets a configurable max size of physical
memory
 ESX must handle overcommitted memory per VM
– ESX must choose which VM to revoke memory
from

10
Memory Reclamation
 Traditional: add transparent swap layer
– Requires meta-level page replacement decisions
– Best data to guide decisions known only by guest
OS
– Guest and meta-level policies may clash
 Alternative: implicit cooperation
– Coax guest into doing page replacement

11
Ballooning – a neat trick!
 ESX must do the memory
reclamation with no
information from VM OS
 ESX uses Ballooning to
achieve this
– A balloon module or driver is
loaded into VM OS
– The balloon works on pinned
physical pages in the VM
– “Inflating” the balloon reclaims
memory
– “Deflating” the balloon
releases the allocated pages

12
Ballooning – a neat trick!
ESX server can “coax” a guest OS into releasing some memory
Example of how Ballooning can be employed

13
Ballooning - performance
Throughput of a Linux VM running dbench with 40 clients. The black
bars plot the performance when the VM is configured with main memory
sizes ranging from 128 MB to 256 MB. The gray bars plot the
performance of the same VM configured with 256 MB, ballooned down
to the specified size.

14
Ballooning - limitations
 Ballooning is not available all the time: OS boot time, driver
explicitly disabled
 Ballooning does not respond fast enough for certain situations
 Guest OS might have limitations to upper bound on balloon size
ESX Server preferentially uses ballooning to reclaim memory.
However, when ballooning is not possible or insufficient, the
system falls back to a paging mechanism. Memory is reclaimed
by paging out to an ESX Server swap area on disk, without any
guest involvement.

16
Sharing Memory - Page Sharing
 ESX Server can exploit the redundancy of data and
instructions across several VMs
– Multiple instances of the same guest OS share many of the
same applications and data
– Sharing across VMs can reduce total memory usage
– Sharing can also increase the level of over-commitment
available for the VMs
Running multiple OSs in VMs on the same machine may
result in multiple copies of the same code and data
being used in the separate VMs. For example, several
VMs are running the same guest OS and have the same
apps or components loaded.

17
Page Sharing
 ESX uses page content to implement sharing
 ESX does not need to modify guest OS to work
 ESX uses hashing to reduce scan comparison
complexity
– A hash value is used to summarize page content
– A hint entry is used to optimize not yet shared
pages
– Hash table content have a COW (copy-on-write) to
make a private copy when they are written too

18
Page Sharing: Scan Candidate PPN

19
Page Sharing: Successful Match

20
Page Sharing - performance
•Best-case. workload.
•Identical Linux VMs.
•SPEC95 benchmarks.
•Lots of potential sharing.
•Metrics
•Total guest PPNs.
•Shared PPNs →67%.
•Saved MPNs →60%.
•Effective sharing
•Negligible overhead

21
This graph plots the metrics shown earlier as a percentage
of aggregate VM memory. For large numbers of VMs,
sharing approaches 67% and nearly 60% of all VM memory
is reclaimed.

22
Real-World Page Sharing metrics from production deployments of ESX Server.
(A) 10 Win NT VMs serving users at a Fortune 50 company, running a variety of
DBs (Oracle, SQL Server), web (IIS,Websphere), development (Java, VB), and
other applications.
(B) 9 Linux VMs serving a large user community for a nonprofit organization,
executing a mix of web (Apache), mail (Majordomo, Postfix, POP/IMAP,
MailArmor), and other servers.
(C) 5 Linux VMs providing web proxy (Squid), mail (Postfix, RAV), and remote
access (ssh) services toVMware employees.

24
Proportional allocation
 ESX allows proportional memory allocation for
VMs
– With maintained memory performance
– With VM isolation
– Admin configurable

25
Proportional allocation
 Resource rights are distributed to clients through
shares
– Clients with more shares get more resources
relative to the total resources in the system
– In overloaded situations client allocation degrades
gracefully
– Proportional-share can be unfair, ESX uses an
“idle memory tax” to overcome this

26
Idle memory tax
 When memory is scarce, clients with idle pages will be penalized
compared to more active ones
 The tax rate specifies the max number of idle pages that can be
reallocated to active clients
– When a idle paging client starts increasing its activity the pages can
be reallocated back to full share
– Idle page cost: k = 1/(1 - tax_rate) with tax_rate: 0 < tax_rate < 1
 ESX statically samples pages in each VM to estimate active
memory usage
 ESX has a default tax rate of .75
 ESX by default samples 100 pages every 30 seconds

27
Idle memory tax
Experiment:
2 VMs, 256 MB, same shares.
VM1: Windows boot+idle.
VM2:Linux boot+dbench.
Solid: usage, Dotted:active.
Change tax rate 0%  75%
After: high tax.
Redistribute VM1→VM2.
VM1 reduced to min size.
VM2 throughput improves 30%

28
Dynamic allocation
 ESX uses thresholds to dynamically allocate
memory to VMs
– ESX has 4 levels from high, soft, hard and low
– The default levels are 6%, 4%, 2% and 1%
– ESX can block a VM when levels are at low
– Rapid state fluctuations are prevented by
changing back to higher level only after higher
threshold is significantly exceeded

29
I/O page remapping
 IA-32 supports PAE to address up to 64GB of
memory over a 36bit address space
 ESX can remap “hot” pages in high “physical”
memory addresses to lower machine addresses

30
Conclusion
 Key features
– Flexible dynamic partitioning
– Efficient support for overcommitted workloads
 Novel mechanisms
– Ballooning leverages guest OS algorithms
– Content-based page sharing
– Proportional-sharing with idle memory tax

31
Similar Products
 VM (IBM), very early, roots in System/360, ’64
–’65
 Bochs, open source emulator.
 Xen, open source VMM, requires changes to
guest OS.
 SIMICS, full system simulator
 VirtualPC (Microsoft)

32
Current status of ESX Server
C. Waldspurger’s Paper - 2002, Today 2005
 Supports enterprise workloads in multi-processor
virtual machines.
 Resource controls for virtual machine CPU,
memory, disk I/O, and network I/O usage.
Supports SLA type guarantees
 Has “VMotion”: Migrate a running VM to a
different physical server connected to the same
storage area network without service interruption

33
That’s all folks,
Thank You.

Memory virtualisation

Recommended

More Related Content

What's hot (20)

Similar to Memory virtualisation (20)

More from Kumar (20)

Recently uploaded (20)

Memory virtualisation

Editor's Notes