Hi all, I’m Tristan Scott, a program manager on the Remote Desktop Virtualization team with some exciting news for decreasing the storage cost of Virtual Desktop Infrastructure (VDI) using Windows Server 2012 R2.
The File Server team just posted two great blogs on the value of live data deduplication for VDI (Extending Data Deduplication to New Workloads in Windows Server 2012 R2) and how to deploy it (Deploying Data Deduplication for VDI Storage in Windows Server 2012 R2). As you can see, storing your personal virtual desktop collections on a Windows Server 2012 R2 file server just got A LOT cheaper, and in some cases, faster as well! Coordinating downtime to turn off your virtual machines and deduplicating your virtual hard disks (VHDs) is a thing of the past, as now “live” VDI files can be deduplicated, meaning your VHDs can be deduplicated even while they are running.
To preempt a few follow-up questions you might have:
- The personal virtual desktop collection is the mainline scenario here, but if you have a session collection running as virtual machines, you should get similar benefits (both often have upwards of 90% duplicative content across the VHDs).
- User profile disks are now another great candidate for deduplication (often upwards of 50% duplicative content across the VHDs).
- Yes, Cluster Shared Volumes (CSV) caching will continue to work on your file servers configured with data deduplication. At any given point in time your VDI file servers will contain some content that has been deduplicated and some that has not (i.e. new or modified content); the system cache will handle caching of files that have been deduplicated, and the CSV cache will handle everything else.
So if you haven’t done so already, now is the time to deploy a scale out file server cluster with Data Deduplication enabled by using the Windows Server 2012 R2 Preview.