I have a Hyper-V server running 2012. It has 4 VMs all running 2012. The host has a removable cartridge drive which contains a VHD for each VM which is attached within each VM as a drive used by Windows Server Backup.
The problem is that sometimes, but not often, the VHD will become detached and so the backup fails. The VMs are not rebooting (which would require reattaching the VHD) so that's not the cause.
I've looked in the event log and the only errors other than the failed backup are in the System Log. Event 137 from ntfs which says "the default transaction resource manager on volume <volumeID> encountered a non-retryable error and could not start"; and Event 14 from volsnap which says "the shadow copies of volume <volumeID> were aborted because of an IO failure on volume <volumeID>".
I tend to think the errors came after the VHDs were no longer available, but I can't be certain.
I have the same setup at other locations and they don't have this problem of the VHD randomly detaching.
Anyone run into this before and know how to fix it?
Jonathan