After VMware storage vMotion, some VMmage backups randomly fail with following:
avvcbimage Warning <16004>: Soap fault detected, Find VM - NOT ok, Msg:
In the workorder specification, the VMware datastore for the VM is not valid and reference the old datastore name from before the storage vMotion.
<targetlist>
<path backup="true" name="[olddatastore] vmname/vmname.vmdk" diskCapacity="xxx" />
</targetlist>
Dell EMC engineering believe this is a VMware timing issue. In some cases, vCenter will sometimes publishes events (like a storage vMotion) before the actual migration job is completed. This causes Avamar MCS service to look up the VM, capture the current VM configuration, and update our VM cache with this data. The VM configuration becomes out of date within a few seconds.
Perform the below steps for short term solution:
By default, Avamar will run auto update the 'vm cache' 1 hour before backup windows start. This means this issue should be fixed automatically during the next day. Note Auto-sync requires no image backup to be running during this period.
In Avamar, the user can right click on the vCenter domain and choose "sync with vCenter" option to force a cache update with the same limitation that no image backups be running.
Perform the below steps for long term solution:
To workaround this VMware issue, Avamar MCS service waits 5 seconds after the vc publish task time to retrieve the VM configuration.
This feature was added in Avamar bug 265993