Restoring unstructured file share from older rotating media

[jcoehoorn]

Hoping this is a simple one and I missed something obvious.

I'm reviewing our internal "playbook" documentation: the notes on how to do certain types of restores to have ready in case of emergency.
I found an issue where the documentation is incomplete due to a recent change, and I also can't find the way to actually accomplish the task to correct it.

The issue in question is a job backing up a file share. The share shows in the Veeam inventory in the "unstructured data" section. I can easily restore these files from the current/online repository, but this repository also uses offline/air gapped rotating drives and is marked as "This repository is backed by rotating drives", where the drives rotate out periodically.

I'm working on the documentation for what to do if we need to restore from older media... say if the repository is corrupted, or if the item we want to restore is past the retention span of the current repository but we know the item would still be on the older media.

In my test, I swap back in the prior media drive I want to use. However, I am unable to find a way to do a restore from this drive for anything from the file share. Everything I try tells me "Unable to perform restore because ... backup metadata is not available." I'm not able to import that old backup, because from what I can see there's no vbk file. Looking on the drive, I can see a vsource file and bunch of metadata and data, but no vbk, and nothing in the Veeam console that wants to use the vsource.

Note we also backup some VMs into the same rotating repository, and I am able to restore these VMs from the older media drive just fine, because the VMs **do** have vbk files and I can import the backups. As a final quirk, these backups are encrypted. However, I have the decryption password ready. We are running version 12.1.2.172.

What am I missing to be able to use file share data from the older media?

[david.domask]

Hi Joel,

With rotated drives, the information about the backups in the repository is not updated until the job runs again as there is no "Rescan" option for rotated drives.

So the backups from the rotated drive will be visible under two circumstances:

1. The job has run after the drive change and recognized the change; the backups on that drive will be present in Backups > Disk
2. Attach the drive, assign a new drive letter, and add it as a new repository (backed by rotated drives is not required); during the repository creation, check the boxes for steps 2 and 3 on this step of the wizard, and when completing the repository creation, it will import the backups for restore. Remove the repository once the restore is completed.

[jcoehoorn]

Okay, so I did this, and it still only finds the VM backups. The imported backups don't show the file shares.

This also seems like a lot more work than it should be. Why no rescan?

[david.domask]

Hi Joel,

Hrm that is not expected, so I'd ask let's get a Support case created and export logs from the Veeam Server itself (use the 3rd radio option, select the Veeam server), and upload those to the case.

The procedure should work as above -- to confirm, you checked the boxes as noted above to scan for backups in the new repository?

As for the behavior of rotated drives here, right now it's by design for how the rotated drive mechanism works; it is a few steps, granted, and will discuss with our technical writers on documenting the procedure.