Rollback and reducing consumed space.

[BenG]

G'day All,

We have one server(Backup) that does not have a lot of spare disk capacity and it is housed in a remote location, so increasing it is not an easy option.

When we originally configured the Veeam Backup job, we mistakenly included a very large VMDK(200GB), that had archival data on it.
As a result, the initial backup was 300+GB, after a couple of more backup jobs the spare capacity has dropped below 50GB.

Once we realised our mistake, we edited the job to exclude the 200GBVMDK and assumed that after the retention count (14) expired, we would get the disk space back.
However that has not happened as we expected, it does process the retention points but not the original backup.

Now, I know we could just configure a new backup job etc and wait for it to do the appropriate number of backups before deleting the old one, but as there is not sufficient space to do this on the server, its not an option. Nor will it be possible to archive the data as bandwidth to the site is quite low.

So, is there anything we can do to manually resolve this issue?

On a related question, is it possible to selectively remove/manage rollback points?
ie if we have 1 VBK file and 10 VRB files that span a period of 2 weeks (lets number them 1-10 with 10 being the most recent).
Is it possible to selectively delete/copy/archive individual VRB files and still maintain the integrity of the backup?
For example if we grabbed the VBK file and #5VRB would that be sufficient to restore to the date of #5 or do we also require #1-4 as well?


Rgds Ben

[Gostev]

Hi Ben, as for your last question - rollback (VRB) chain cannot be broken in the middle, as these are reversed increments. If you need restore date #5, you have to have VRBs number 1 through 5, and actual VBK. So, you always want to delete the oldest VRB you have.

As first your first question, with the current version, the only way to remove VM from VBK is to perform Full Backup on the job. If you do not have space to host new full (without that 200GB VM), then your only option is delete backup and re-create the job.

Thanks!

[BenG]

G'day Gostev,
Thanks for clarifying these two items, much appreciated.
We will see if we can attach a NAS box to the local LAN and move the a copy across there as our SLA states we will keep 2 weeks.

Rgds Ben

[pearsondan99]

We had the same issue - initally our replication was 2 x jobs with 3 x rollback points per job which chewed up all of the space on the replication target, the only way we could clear the old snapshots at the target was to delete the job and recreate with only 1 rollback point which does fit the allocated space on the target. It would be handy as a future enhancement if you could remove snapshots safely without having to delete and re create the job (avoiding the intial full re-replication)

Dan

[Gostev]

Dan, you can delete oldest unnecessary rollback point manually at any time even with the current version. Or, you can change job's retention policy and it will do this for you automatically during the next job run. You do not need to recreate the job for this.