Best way to resolve low disk space on rotated backup copy disk?

[TimWalker]

Hi All
Our full backup sizes are in the region of 4 tb. For offline backup copies, we have a couple of 8tb drives that we rotate and store in a safe - it's working well enough, ReFS fast copy seems to be working so we're basically happy with it.

We keep 15 restore points on the disks - ideally though this means we have a month of backups available across the two drives.

Anyhow, we were closed over Christmas so over that period I left both drives disconnected.

Now, on coming back, somewhat understandably the latest increments are a bit bigger, and I've managed to pretty much fill the drives.

I tried turning down the number of Restore Points to keep for the job, applied the new settings, but the job doesn't want to merge the old backup files - instead it wants to carry on with the latest copy job and fill the drive (I know this because one of the drives filled, and I ended up having to just manually delete the backup files and trigger a full backup).

Is there some way to force it to merge the old backups? Basically I want to delete the oldest restore points but hopefully retain some older ones (and avoid a lengthy full backup, as these drives aren't quick!)

I did try deleting one of the less important servers from the backup list, but this doesn't seem to actually release any disk space (the job contains several servers).

What's the best way to free up space without just junking all the old backup copies? Obviously I have two other copies of the data but would like to keep three...

[foggy]

Hi Tim, retention applies after a successful job run, so you first need to free space to let the job run successfully and free some space. Manually deleting some older backups is the only way to do this. Removing VMs from backup also does not release the space immediately, corresponding blocks are just marked as free to be further reused.

[TimWalker]

Not really possible then with a forever forward incremental, as deleting any backup file will kill the chain - so basically you have to kill the chain either way (unless you have loads of free storage on your rotating drives!)

Would be nice if there was a way to do this! Not sure what the best way is.

For now I've split off a less important server onto a secondary job with less restore points, so at least perhaps in future I only have to zap that one to get myself out of the pickle. Shame that this will lose a little bit of dedupe efficiency.

[foggy]

We have a requirement to address such cases in a more elegant manner logged for future versions.

[mike008]

@foggy Any update on this functionality being integrated in VBR? This is a frequent need in our environments.

[foggy]

Hi Michael, as far as I know, there's currently no particular ETA for this functionality. Thanks!