cancel
Showing results for 
Search instead for 
Did you mean: 

Error Code E0008516 "Resource specified for snaphoots..." for Hyper-V 2008 R2 using Clustering

ahall
Level 3

I've battling this issue for a very long time. It goes away for a couple of months only to come back again. The issue is backing up my Hyper-V host on a two node HA 2008 R2 clustering configuraiton using clustered shared volumes. 

Current setup: Backup Exec 2010 R2 w/SP1 Hotfix 150096,144101,148347,153090 and two Windows 2008 R2 w/SP1 hyper-v hosts and no new hotfixes since SP1, using four Clustered Shared Volumes (CSV's) between the two VH host nodes.

Backups will occastionally fail when switching between one of the two nodes. One of the node's VM will back up fine, only to have othe other nodes vm fail with "0x0008156 - Resources specified for snapsho do not have any valid data on them. Check if the files were deleted or renamed".

My current theory is it's a timing out problem with the different clustered shared volumes and switching the owner during the backup process. There was a hotfix available pre sp1 that was targeting a similar issue but it should be interegrated with 2008 R2 SP1 now (?).

Has anyone else experienced this issue???  I've had four or five support tickets opened the past year trying to resolve the issue, but nothing really helped - the issue goes away for a month or two only to return. 

2 REPLIES 2

stephan_vanheld
Level 5

I have exactly the same issue - backup of all VMs running on the second node fails with the above message.

ahall
Level 3

Glad to hear I'm not the only one out there with this problem

I opened up a ticket yesterday (again) to try to work on the problem

In the meantime I've done the following which works so far

  • Split the single VM job into backups of each node, for me just 2 backup jobs
  • Definied preferred owners and failback preference for each VM, then made sure they are running on the preferred owner
  • Selected only the VM's per node that are running on from the overall cluster name
  • Each VM host was only running VM's on the same clustered shared volume. So it looked like this for me

VM Host 1

  • 8 Guest VMs running in CSV 1 & 2

VM Host 2

  • 5 Guest VM's running on CSV 3 & 4

 

  • By splitting up the VM's and make sure the CSV's don't change ownership between the nodes the issue appears to be isolated. This furthers my theory that it related issue with the CSV's changing ownership between nodes during a backup is causing the dreaded "snapshot not found ...."