Hey guys, I'm having a major problem here. Using 5.5U1 and latest Veeam.
Three days ago, something stuck in a Veeam backup process, and after that all the backup jobs created snapshots and then never deleted them.
So yesterday I discovered through SSH that each veeam-protected production VM had like 90 delta files and they weren't showing up in the Snapshot Manager. Also, "Delete all snapshots" failed. And creating new snapshot failed too. After fiddling around with restarting management services on the hosts and adding/removing snapshots I was able to consolidate most of the snapshots successfully.
However, I've still yet to fix some of the issues, namely:
1) One very large VM still has a full chain of -deltas attached to one of its vmdks. vCenter offers to "consolidate", but then consolidation fails with "Unable to allocate memory". Theres plenty of RAM available on the host, like 100GB+, and memory-wise it's a small VM. There's like 0,8x the size of VM free space on the datastore, but since all of the deltas take about 10GB combined, I think this shouldnt be a problem too. There was a similar problem with another big VM, but it was solved by powering the VM off and then consolidating. With this VM powering it down doesn't help.
I still kinda think it's a datastore space issue, however I don't see why so much free space would be required to commit the snapshots. Is there any way maybe to do this one delta file at a time?
2) Creating new snapshots is now broken. If I try to create a new snapshot of any VM on the affected hosts (even if its powered off and not one of the Veeam-protected bunch), it fails with "A general system error occurred:". The delta files are then created, but snapshots don't appear in the snapshot manager. However, if I restart the management services with /etc/init.d/hostd restart and /etc/init.d/vmxa restart, these failed snapshots suddenly appear in the manager, and work like they're supposed to. And can be successfully removed.
If I try to take a snapshot again, it fails with the same error, and the cycle continues.
I tried restarting one ESXi host, but that didn't do anything besides what restarting management services already did. The behaviour is the same if I create/remove snapshots in Web Client through vCenter, or if I do it directly on the host in the clish.
I have a suspicion that this is what caused the Veeam issue in the first place, since the snapshotting seemingly fails, but actually happens behind the scenes. The question is what could be broken and why?
We dont have an active vmware support contract, and I think Veeam guys will say its a vmware issue. I've yet to hear a reply, but my previous experience with them tells me that's what's going to happen.
So you are my only hope 
Thanks in advance.