HyperV Server 2012 R2 – Shared VHDX (TechEd Europe part 3)

NOTE: while I’m still keeping the current posts live as they still seem to help, currently my focus has changed and new activity moved to the new site iternia.be

When building Windows Clusters, one of the least flexible requirements has always been the centralised storage. iSCSI disks f.e. were needed as a quorom / witness resource and to put your application data on. In modern scenario’s with multi-tenant environments that is however not something a storage admin gets happy or excited about. LUNs have to be masked, storage firewalls have to be used (to avoid a client machine from using other ports than only the allowed iSCSI ports f.e.) or even CHAP-authentication had to be implemented.

In our own hosting environment there’s a storage firewall cluster in place with its own frontend and backend VLANs and physically dedicated ethernet cabling (to make sure storage traffic would never be able to impact frontend applicative traffic and client request performance). A costly investment…

But finally there’s the option to use shared virtual disks for clustering. “Old news”, some will argue. It was indeed already possible for test/dev environments, but never supported on production systems by Microsoft… until now! Shared VHDX-files can now be used as cluster resources. This has certainly a lot of advantages:

• Flexibility: if ever needed you can move your disks to other storage without the intervention of expensive storage specialists.
• Security: your infrastructure layer and your frontend layer don’t have to have a backdoor anymore, allowing a machine (accessible to clients) to connect to storage nodes / controllers.

Another feature leveraging the maturity of HyperV. I must say the progress has been high the past year…

There’s a lot more coming to HyperV 2012 R2 so stay tuned for updates. Thanks!

(this is part 3 of several posts from Microsoft TechEd)