You can move Writable Volumes from one storage to another. You can also take backups of the volumes and restore them.
You can back up a single volume, or multiple volumes, but you can only restore a single volume. You can also schedule App Volumes Manager to take regular backups of the Writable Volumes.
Storage Groups are not supported for the move, back up, and restore operations.
Thin Provisioning and Writable Volumes
If you move or back up a writable to a storage that does not support thin provisioning, the volume is expanded and becomes a flat VMDK on the destination storage. You can see the expanded size on the Backup Writable and Move Writable screens. For example, Move Writable Volume for user1 - 400.00 MB (expanded: 10.00 GB)
If you restore a volume that is flat, it will remain flat, and will not get converted back to a sparse volume.
Using Shared Datastores for Writable Volumes Operations
You must have a shared datastore between the source and destination vCenter Server to move and back up volumes across different vCenter Servers and if the source volumes are located in a non-shared datastore.
To back up a Writable Volume across different vCenter Servers, with the volume located in a non-shared datastore, you must first move the volume to a shared datastore, and then back up the volume to the destination datastore.
See Support for Shared Datastores for information about shared datastores.
Moving a Large Number of Writable Volumes
If you are moving a large number of Writable Volumes, the operation could take a long time depending on the underlying infrastructure. You can reduce the time taken to move this large number of Writable Volumes by changing the default value of the max_marshall_writable_jobs and marshal_writables_downtime variables.
For example, if you want move 30 volumes in 6 minutes (360 seconds), set the following variables in the database settings table:
max_marshall_writable_jobs = 30
marshal_writables_downtime = 6
In the above case, it takes 12 seconds (360/30) to move each Writable Volume. However, the speed of the movement depends on the total size of the Writable Volume and the infrastructure configuration. The infrastructure includes the underlying datastore, target datastore, and network bandwidth. If the infrastructure does not support the speed set by the variables, the move jobs will pile up and will be completed in a time supported by the underlying infrastructure.
The actual time taken to complete the operation in the above example is also slightly more than 6 minutes since App Volumes might perform a refresh on the existing Writable Volumes or perform other sync jobs, while completing the move.
Therefore, set the values depending on the number of Writable Volumes to be moved and the underlying infrastructure.
The worker_threads_count variable also may need to be modified depending on your setting.
