VMware Solutions Discussions
VMware Solutions Discussions
Hi all,
Can anyone share their experiences / recommendations / documentation regarding a recovery plan in VMware SRM to include a NetApp vFiler fail-over?
To make it clear:
The application in question spans some VMs hosted on NFS datastores with iSCSI RDMs (so 'regular' stuff), plus flat files in a CIFS file share (already inside a vFiler, which presumably makes fail-over easier). So we need to fail-over VMs and vFiler (with CIFS share) at the same time to make it working.
It does sound doable, as pretty much anything can be scripted as a custom step within SRM Recovery Plan, but it would be really nice to know if someone implemented it (successfully ) already before.
Thanks & regards,
Radek
The sra works the same with a vFiler as a physical controller. I wish we had vFiler dr integration with the plugin but don't. So just snapmirror between filers or vfilers.
The cifs share adds a wrinkle but you can script the break and resync. Haven't had to do that yet but would be the same if a vFiler or not.
Hi Scott,
Many thanks for your response.
Yes, I am mainly concerned about failing-over the CIFS share (hosted in a vFiler). The main issue I see with including vFiler fail-over as a custom step within SRM Recovery Plan, is the fact that executing the plan in test mode won't work - we can't simply clone a vFiler, can we?
Presumably a separate plan would be required to do testing: something along the lines of cloning a volume from the DR vFiler & presenting it to another vFiler, connected to a test network?
Regards,
Radek
You can clone a volume from a vFiler into another vFiler...that would work but the IPs wouldn't be the same and wouldn't be the bubble network automatically...could ifconfig to make it any way you want though.
[bump]
Anyone else?
Any actual implementations?
I got quite useful tip on VMware Communities how to automatically run two different scripts depending whether the SRM Recovery Plan is run in Test, or DR mode:
http://communities.vmware.com/message/1876457#1876457
This means we need just a single plan, triggering two different scripts for vFiler testing & vFiler fail-over respectively.
By pure accident I just stumbled across this blog post from Vaughn Stewart:
Pretty much exactly what I was after - why it didn't come up in my search results?