A previously-working vRanger backup fails with the above error (almost instantly).
The storage location on one of the VM's disks (if a multiple-disk VM) may be "obsolete" in the vRanger and/or vCenter databases. Upon examination of the "Edit Settings" page for the VM, the path of one or more of the .vmdk disks may read "[ ] /{datastore UUID}/{VM Path}/{VMName.vmdk}". Note the " [ ] " "blank" brackets in the datastore name field. This indicates that VMWare cannot resolve this datastore UUID, so it either no longer exists, has been reformatted/re-added (and received a new UUID in the process), or is corrupt.
1) Best solution: Storage vMotion the VM disk(s) in question to another datastore, then back to the proper datastore. The backup should then run normally, as this will refresh the datastore registration information.
2) If not licensed for Storage VMotion, the VM should be shut down, the offending disk(s) removed (NOT deleted, just removed from the VM), then the .vmdk(s) in question should be re-added to the VM with the "use existing disk" option in the VMware "Edit Settings" page. Then, the machine may be brought online, the vRanger inventory refreshed, and the backup job should then run normally again.
