Network and Storage Protocols

XenServer 5.5+ on NFS experiencing frozen or corrupted VMs during cluster failover


Looking for a XenServer guru to help answer some questions from a partner:

One of our customer’s is currently running XenServer 5.5+, Ontap 7.3, and using NFS SR.
The last FAS System cluster failover caused several VMs to freeze and some were corrupted (requiring a rebuild of these VMs).
The VMs consist of Apache and MySQL servers.  The VMs OS reside on VHDs and the VMs Apache and Database reside on NFS mounts within their OS.
Are there any other recommendations for setting up the environment other than what’s already noted in TR-3732?

I’ve read that XenServer HA should be disabled whenever maintenance is being performed on the network or the SAN.  Is this true? 

Can you recommend any other best practice guidelines?  Also, what logs should they be capturing?




Looking to add info to this thread. Anyone have any information regarding configuring a XenServer for a FAS during takeover? Any specific settings best-practices?