Host crashed while deleting snapshots

[btobia]

I have an ESXi 3.5 u3 host that I'm having some I/O issues with a local disk controller (PE 2850, Perc4). I've been consolidating snaphots on the VM's and preparing to move them over to SAN storage to alleviate this problem, but I ran into a snag last night. I did a delete all snapshots to a powered off VM, and the host hung half way through, I had to hard reset it. The issue I have now is when I look at the VI client, it thinks that the VM doesn't have any snapshots. However, looking at the actual file system, I still see delta files for each of the virtual disks.

What is the best way to consolidate these snapshots now that VI client doesn't think any exist? Is there a way to force it to re-poll? My other thought was to take a new snapshot of the VM, and then do a delete all. Would this force the client to consolidate even the older snapshots that it doens't see anymore?

[DSTAVERT]

Consolidating large snapshots can take forever so the host may not have hung. I would post my question at http://sanbarrow.com/sickbay

Before you do anything I would copy what you have to the san and work with the copy.

[java_cat33]

I agree - take a copy of the VM folder that contains all the vmdk's etc if possible. But otherwise, yes creating a new snapshot then doing a "delete all" will usually sort out the snaps that you can't see from within snapshot manager - as long as the VM is not in a bad state from the host reset.

Another option you have is to take a clone of the VM, power on the clone and make sure it's ok - then delete the original. When you take the clone it won't have any snapshots. However, you must have vCenter to take a clone.