nfs connectivity lost and resume

[rvhi]

Hi,

During a recent switch software upgrade, the network connectivity to NFS storage using Netapp was lost for a few minutes. We stored everything in NFS with Netapp, including vmdk files. When the connectivity came back, some VMs were in shutdown down mode and we had to using command line to power them on. Some VMs had disks in read-only mode and had to restart manually. This greatly concerned me. Of course, we can improve the setup to prevent network outage. But if the network is down, is there a better way to force restarting all vms in certain orders?

Thanks,

Richard

[prasadsv]

Are the VM's really powered OFF.I guess VM's got crashed.You can find zdump of the VM's in it's home directory.VM's getting crashed is expected since the VM''s vmdk was not accessible due to network outage.

What are the values of HeartbeatFrequency,HeartTimeout and HeartBeatMaxFailures  on your hosts.Please refer to page 9 in  doc attached.

If you are still facing the issue of VM's getting crashed with these values, try increasing these values and check.In this case your nfs client will try for to check more number of times before VM's get crashed.