Automate filesystem cleanups

[raronson@dga.org]

We use Exagrid as a storage target in two sites with B&R 9.5 Update 4. We're using Exagrid to replicate backups from our remote location to our central site which we then copy to tape once a week. Data at the remote site is deduped at that site then replicated to a similar repository which is local to the tape system. The replicated data has to be rehydrated before it can go to the tape. its around 10 tb, takes nearly 24 hours. Nothing can change the way that functions.

My problem is every week the tape copy is backing up the entire replicated repository to tape. I can't figure out how to only copy the most recent versions from the replicated data to the tape with a scheduled copy process. If I could either come up with a way to backup only the most recent data it would save hours. The other option I can see is to remove all but the most recent backup from the local repository before my tape copy. The favored outcome is to reduce the time to write data during the tape copy, any methodology youp can recommend would be appreciated.

[PetrM]

Hi Rob!

We're using Exagrid to replicate backups from our remote location

Do I understand correctly that you're talking about backups created by backup jobs on some remote location and not about backups produced by backup copy job?

In the first case, may be you can use backup copy job to transfer only actual VM state from remote site and then to use this backup copy as a source job for backup to tape?
Backup copy approach is described in this article as one of possible options to follow 3-2-1 rule.

However, please keep in mind that you lose Exagrid replication benefits following this recommendation and also will need to schedule a virtual full.
One more consideration is that in this case the backup chain on tape will be always one restore point shorter than a chain on a disk, schedule needs to be adjusted accordingly.

Thanks!