Last week saw VMware release the latest version of VMware View, VMware Horizon View 5.2. While there are a whole bunch of enhancements in this release, I wanted to focus on a few enhancements that are specifically storage related. I’ve covered some of these in the past from a vSphere perspective, but now we have a new release of View which can take advantage of these features.
1. Cluster Scalability Improvements – In vSphere 5.1, VMware introduced a new locking mechanism to VMFS which increases the number of hosts which can share a read-only file lock from 8 to 32. Specifically, this means that linked clones referencing the same base disk can now be placed on up to 32 hosts. VMware Horizon View 5.2 leverages this new locking enhancement so that a cluster can now consist of 32 hosts, and it doesn’t matter whether the underlying storage is NFS or VMFS, or if the pool type is linked clone or not.
2. Space Efficient Sparse Virtual Disks – This eagerly anticipated new VMDK format is now readily available in View 5.2. In fact, this is now the default format for linked clones. The SE Sparse Disk was introduced in vSphere 5.1, but we had to wait for View 5.2 to become available for these new VMDKs to be leveraged. Basically SE Sparse Disks give View users two features:
- A linked clone which has a 4KB block size akin to standard VMDKs, as opposed to older linked clones which had a grain size of 512 bytes. This smaller grain size did not perform optimally on some arrays due to alignment issues and partial write issues.
- An automated way to shrink the desktop size. This has been a bugbear for some time, as desktops based on linked clones would grow with use and consume disk space, and considerable administrative overhead was incurred to reduce their size.
For additional information on SE Sparse Disks, there is an earlier article that I wrote about them located here. There is also a very good article posted on the vSphere blog here covering the SE Sparse Disk. One final note – it seems that to use SE Sparse Disks with VMware Horizon View 5.2, you must use vSphere 5.1 patch 1. Also, all VMs wishing to use the SE Sparse Disk format must be HWv9 (which is the vSphere 5.1 virtual machine hardware version).
3. View Composer Array Integration – This feature allows the creation of linked clones to be offloaded to the storage array. Unfortunately, this feature (which you can learn more about here) is still in Tech Preview in View 5.2, so is therefore unsupported by VMware. However, if you still want to leverage VCAI (since the feature is available via a number of our storage array partners), you need to be aware that you cannot use SE Sparse Disks.
4. View Storage Accelerator & Tiering – The View Storage Accelerator (aka Content Based Read Cache – CBRC), was introduced in View 5.1 and is used to dramatically improve the read performance of desktops. It is implemented by taking an area of host memory for cache, and then creating ‘digest’ files for each virtual machine disk. This feature is most useful for shared disks that are read frequently, such as View Composer OS disks. View Tiering technology was created to facilitate the placement of replica disk on SSD (Solid State Disks), which in turn meant that there was I/O performance improvements. Prior to 5.2, the thought was that it was not necessary to have both of these technologies co-exist, so customers had to choose which solution to use, if any. With the release of VMware Horizon View 5.2, both technologies can now co-exist. My colleague Andre Leibovici, does a really good job of explaining why we decided to support both features in his blog post here. Finally, you might ask about CBRC and SE Sparse Disks interoperability. Yes, these are also supported together.
It is great to finally see some of these features introduced in vSphere start to make impacts on our other product lines. I’m particularly delighted to see SE Sparse Disks make it into View, as I think this will solve a lot of the linked clone related issues we have seen in the past. Let’s hope we see this format in even more product areas moving forward.