Posted by: rockafella | June 3, 2011

XenServer and XenDesktop IntelliCache reducing your XenDestop Storage Costs by 50-80%

NEW: Intellicache TCO calculator. It compares the relative costs of deploying XenServer or VMware underneath XenDesktop.: http://www.citrix.com/xenserver/intellicache

XenServer 5.6 SP2 and XenDesktop 5 SP1 now fully support IntelliCache. 

Intellicache makes hosted Virtual Desktop Infrastructure deployments more cost-effective by enabling you to use a combination of shared storage and local storage.  IntelliCache is one feature of XenServer that caches XenDesktop Master, temporary and non-persistent files for desktop workloads on the local disk of the host server.

InteliCache Animation

InteliCache Demonstration 

Citrix TV Demo of IntelliCache and MCShttp://www.citrix.com/tv/#videos/4038


IntelliCache Key Points

  • IntelliCache allows you to take advantage of Storage tiering methods to transfer XD workloads to less expensive shared storage. (i.e. Remove the disks from the LUNs/RAID, Use SATA disks instead of FC Disks, Reduce to slow RAID speeds RAID 5, Move the XenDesktops to slower SATA or NAS storage)
  • You can also use local disk storage for your XenDesktop instead of Shared Storage
  • Ensure that the local XenServer storage can support a high number of IOPS, as this is where the majority of reads/writes are happening.  It’s recommend that you use SSD’s or multiple 15K SAS disks  (RAID’d w/ battery backed write cache) in order to handle the IOPS that are no longer going to the shared storage but rather to the local cache
  • XenDesktop labs has run a series of tests with 1000 MCS pooled desktops on NFS, both without IntelliCache and with IntelliCache turned on. Without IC, the total setup generated about 16,000 IOs per second at its peak. With IC turned on, there was an initial peak of about 3,000 IO’s where the caches on the hypervisor hosts are getting primed, but that soon dropped off to 35-40 IO’s per second sustained.  If you turn that into requirements for your shared storage, you can see that you can get away with a much lower class storage array than previously, as your storage is going to be sitting practically idle after the first VM booted. For dedicated desktops, we are still running more tests, but preliminary results show that the IO load will be cut in half, and the bulk of the IO’s to the shared storage will be writes (just like PVS, not surprisingly).
  • The first phase of IntelliCache supports MCS which is NFS/CFS only. Block-based options will come in future quarters as IntelliCache will eventually support PVS. Requirements also change based on shared vs. persistent/assigned Vms as persistent writes still need to go back to central storage.
  • There is are individual temporary write cache files and a single shared read cache file created for every shared desktop VM. The same is true for persistent image (dedicated) desktops, however the write cache files are duplicated on shared storage and are persistent. A single MCS shared image can support as many desktops as the XD controller is able to support. Current testing has demonstrated roughly 5,500 desktops on a single controller
  • Reduce storage costs and IOPS by 50-80%. The savings is more like 40% when using dedicated image desktops.
  • Without IntelliCache: Approximately 20 spindles of NetApp NFS storage is needed to support 107 XD5 desktops running on a single host.  Let’s say that the cost for this is approximately $25,000-$30,000.

  • With IntelliCache: Depending on whether or not you use local SAS or SSDs for storage of read & write cache, you will have between 49 (w/ SAS) and 114 (w/SSDs) desktop running on the same size host. The difference now, however, is that the shared NFS storage needed to support these desktops drops to only a few GB and the amount of IOPS drops by 99%+, meaning that expensive enterprise-class storage isn’t required. A customer could effectively use a very low cost storage appliance for that purpose (e.g. http://www.drobostore.com/store/drobo/en_US/home) which might only cost $2,000-$3,000 by comparison.

  • As you can see, the cost of storage (and the potential for storage cost savings) when it comes to virtual desktops is all about the IOPS needed. By moving the IOPS demand from shared to local storage, the costs drop significantly.
  • The following screen shot is from a XenServer master class video demonstrating XenDesktops with and without IntelliCache..  When IntelliCache is enabled the IOPS flatline. Here is the link to the video  https://www1.gotomeeting.com/register/796232337

  • The Following two Diagrams shows where to enable IntelliCache.. The second diagram shows enabling IntelliCache at the XenServer Install


Reference:

Calculate and Measure your existing Windows XP or 7 Workloads for different types of users:

Type of Work      Average IOPS

Mobile                           x

Remote                         x

Office-Based               x

Contractor                 x

Task-based                x

NB: You can use IOMeter to measure the IOPS for these workloads for more detailed information on measure IOPS use this guide  Windows 7 IOPS Deep Dive

About these ads

Responses

  1. [...] How does it work? The folks over at Citrix blogs have already done an excellent job explaining how Intellicache works so we’ll try not to repeat too much here.  However, at a fairly basic level, the offloading of IO is achieved by caching blocks of data accessed from shared storage by virtual desktops onto host local storage.  So if Intellicache is enabled and a Windows 7 VM boots from a particular XenServer host, it will cache the roughly ~200MB accessed by the Operating System during the boot process on the hosts local storage.  Subsequent VMs that boot up on that host will then access these blocks from local storage instead of the SAN.  In addition, if you are using non-persistent images, your writes will occur exclusively on local storage as well.  Persistent (aka “Dedicated”) images will write to local and shared storage.  I think this image from the Citrix blog sums it up nicely: You might also be wondering about storage space and what happens when you run out of room on your local storage.  With both read and write caches happening on local storage, this is bound to happen from time to time.  Luckily, Intellicache has taken this into account and will seamlessly fail back to shared storage in the event the local storage runs out of space.   For more on “how it works”, see the link above or read more here. [...]

  2. […] You might also be wondering about storage space and what happens when you run out of room on your local storage.  With both read and write caches happening on local storage, this is bound to happen from time to time.  Luckily, Intellicache has taken this into account and will seamlessly fail back to shared storage in the event the local storage runs out of space.   For more on “how it works”, see the link above or read more here. […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Categories

Follow

Get every new post delivered to your Inbox.

Join 26 other followers