Backup Restore from Tape to Repository - Structure Manipulation

[mamosorre84]

Hi all,

I have a question about the "restore from tape to repository" default behavior: is it possible to keep the same structure objects/point of the tape job?

Let me explain, the scenario is (values are for example):

-10 backup to tape jobs with GFS (12 monthly + 3 yearly)
-every job has 10 VMs

The goal is: import all tape backups to a Veeam repository and decommission the tape library.

When I restore job 1, for example, from tape to repo, I can choose all 10 VMs from the same point, but not every point per VM.

So, the result is that, for the same job, I will have separate "imported object" per point!

Is it the normal behavior?

If yes, is it possible to consolidate these objects, using powershell or manipulating the db, into a single object with multiple VMs/point like the original job to tape structure?

Sorry for the length of the post, I didn't know how to explain it in a more simple way

Thanks

Marco S.

[david.domask]

Hi Marco,

Currently the tape restore to a staging repository places and imports the backup file(s) restored from tape from the given restore point.

However, you can use Restore Entire Tapes to folder, and select multiple tapes for this operation. For example, I restored from my GFS media pool, just locating the necessary tapes and pointing them to a directory with enough free space on one of my repositories. All of the points were restored into the same folder, similar to what you'd see on a normal backup repository. You will need to Import the restored backups by VBK using this method.

So with your configuration, it sounds like you have the machines split across 10 tape jobs; does each job use the same GFS Media Pool + tapes or each job has its own media pool + tapes? Mostly this is just about convenience, as if all jobs share the same media pool, you need to restore entire tapes and cannot specify individual workloads to restore all of.

If you are comfortable with Powershell, you can also utilize the method in this thread to return points for a given machine. The backup file names have a fairly structured format, so you can validate on the file name (assuming Per-VM backups) to understand which backup files relate to which workload.

However, at the time being, you can only import individual backup chains (VBK and the VIB files that depend on them) as separate backups in the UI. Currently there is not a supported means to "stitch" the backup back together, so each backup chain will show as a separate backup when imported.

[mamosorre84]

Hi David,

each job use the same GFS media pool.

Thank you for your suggestions, I'm going to test both options.

Marco S.