Best Way to Restore a Specific VM from Hourly Snapshots in Apache CloudStack 4.19.1.2

[pankajfromcomhard]

We are using SAN storage as the primary NFS storage in our Apache CloudStack 4.19.1.2 environment. To ensure data safety and quick recovery, we take hourly snapshots of our NFS storage. In case of an issue, we restore the snapshot and mount it as a new NFS share within Apache CloudStack. This allows us to browse files as they were one hour earlier.

Is there any efficient method to restore a specific VM without restoring the entire NFS mount? Any best practices or automation suggestions from the community would be highly appreciated.

[boring-cyborg[bot]]

Thanks for opening your first issue here! Be sure to follow the issue template!

[sbrueseke]

I am not sure I get your goal. Can you please explain it in more detail? Are you using Instance-Snapshots or Volume-Snapshots? Do you want to do a full recovery or is your goal to recover on file level?

[pankajfromcomhard]

We are using NetApp SAN storage as the primary NFS storage in our Apache CloudStack 4.19.1.2 environment. Our SAN has the capability to take hourly snapshots of the NFS volume, ensuring data safety and quick recovery.

In case one of our VMs crashes or gets infected by a virus, we create a clone of the NFS repository in our NetApp SAN and mount it as a different NFS share within Apache CloudStack. This allows us to access the data as it was one hour earlier.

Currently, we can see both storage locations in Apache CloudStack and can browse the VM’s disk files (qcow2) from an hour ago. However, there is no direct option to move large files (100-500GB) between the two NFS mounts efficiently within CloudStack.

What would be the best way to restore a specific VM from this snapshot without manually copying large disk files? Are there any best practices or tools to streamline this process?