Sometimes VMware’s Data Protection (VDP) appliance fails. Some VMs may not be backup up properly, when VDP thinks that there are ghost snapshots attached on these VMs.
In our case, VDP failed with the error “VDP: Operation failed due to existing snapshot”. Checking out the related datastore, we noticed that there was an old vmdk snapshot file in the VM datastore folder. The question was whether this snapshot file is linked to the base image or not.
First I tried to consolidate the files via the snapshot manager, I even tried to create a snapshot and then “deleted all” snapshots, but nothing fixed the problem.
Somehow, you may use “vmkfstools -e -v10” command from a ssh terminal session to verify the linking between the vmdk files but I didn’t figure out how to do it properly.
An easy way to verify whether the snapshot file is linked or not, is to migrate the VM to another datastore. By doing so, the VM related files move to the new datastore and the garbage files remain in the old datastore. Browsing the old datastore you may see the ghost snapshot file that produced the problem. You may delete the old datastore folder when you finish the entire troubleshooting session and everything works alright.
Even though I migrated the VM to another datastore the problem still remained, and VDP backup failed with the error “VDP: An unexpected error occurred with the following error code: 10058.”.
Finally, I migrated the VM to another host and VDP backup finished successfully.