Best Practice for General User File Server HA/Failover

Question

Hi All,
Looking for some general advice or documentation on recommend approaches to file storage.  If you were in our position how would you approach adding more rubustness into our setup?
We currently run a single 2012 R2 VM with around 6TB of user files and data.  We deduplicate the volume and use quota's.
We need a solution that provides better redundancy that a single VM.  If that VM goes offline how do we maintain user access to the files.
We use DFS to publish file shares to users and machines.
Solutions I have researched with potential draw backs:
Create a guest VM cluster and use a Continuosly Available File Share (not SOFS)
 - This would leave us without support for de-duplication. (we get around 50% savings atm and space is tight)
Create a second VM and add it as secondary DFS folder targets, configure replication between the two servers
 -  Is this the prefered enterprise approach to share avialability?  How will hosting user shares (documents etc...) cope in a replication environment.
Note: we have run a physical clustered file server in the past with great results except for the ~5 mins downtime when failover occurs.
Any thoughts on where I should be focusing my efforts?
Thanks

Answer 1

Hi All,
Looking for some general advice or documentation on recommend approaches to file storage.  If you were in our position how would you approach adding more rubustness into our setup?
We currently run a single 2012 R2 VM with around 6TB of user files and data.  We deduplicate the volume and use quota's.
We need a solution that provides better redundancy that a single VM.  If that VM goes offline how do we maintain user access to the files.
We use DFS to publish file shares to users and machines.
Solutions I have researched with potential draw backs:
Create a guest VM cluster and use a Continuosly Available File Share (not SOFS)
 - This would leave us without support for de-duplication. (we get around 50% savings atm and space is tight)
Create a second VM and add it as secondary DFS folder targets, configure replication between the two servers
 -  Is this the prefered enterprise approach to share avialability?  How will hosting user shares (documents etc...) cope in a replication environment.
Note: we have run a physical clustered file server in the past with great results except for the ~5 mins downtime when failover occurs.
Any thoughts on where I should be focusing my efforts?
Thanks
If you care about performance and real failover transparency then guest VM cluster is a way to go (compared to DFS of course). I don't get your point about "no deduplication". You can still use dedupe inside your VM just will have sure you "shrink" the VHDX
from time to time to give away space to host file system. See:
Using Guest Clustering for High Availability
http://technet.microsoft.com/en-us/library/dn440540.aspx
Super-fast
Failovers with VM Guest Clustering in Windows Server 2012 Hyper-V
http://blogs.technet.com/b/keithmayer/archive/2013/03/21/virtual-machine-guest-clustering-with-windows-server-2012-become-a-virtualization-expert-in-20-days-part-14-of-20.aspx
can't
shrink vhdx file after applying deduplication
http://social.technet.microsoft.com/Forums/windowsserver/en-US/533aac39-b08d-4a67-b3d4-e2a90167081b/cant-shrink-vhdx-file-after-applying-deduplication?forum=winserver8gen
Hope this helped :)
StarWind VSAN [Virtual SAN] clusters Hyper-V without SAS, Fibre Channel, SMB 3.0 or iSCSI, uses Ethernet to mirror internally mounted SATA disks between hosts.