"Restore points to keep" being ignored?

[kkoehn]

We're using B&R 5.0.1 and are noticing that the number of restore points to keep are being ignored. I've got them set to 21 and most of our backup jobs have 30+ restore points and we're running out of space! We've had to change backup types during the life of the jobs but the numbers just seem to keep adding up.

Is there a way to reconcile this or is recreating the jobs and deleting the backups the only way to fix this? I've even set them to 1 point, 5 points, etc. and it doesn't change anything.

Thanks!

[Gostev]

Hi, you need to contact our support directly to troubleshoot this. Restore points should be removed automatically. Thanks.

[kkoehn]

Thanks.. will do in the morning!

[Alwin]

i acknowledge this behavior

[Gostev]

This may happen when some VMs which were previously backed up by this job are no longer being backed up (for example, because they are no longer fall into the job scope). VM level retention in Veeam Backup attempts to preserve last X restore points of each VM (according to your retention policy), including thoses no longer being backed up. So, backup files containing corresponding VM data (restore points) will stay on disk until you acknowledge that VM is no longer needed. You do that by deleting VM from the corresponding backup under Backups node. Doing this will immediately clean up all backup files which were kept because of this VM.

Basically, our product only sees that VM it was backing up previously is gone. We cannot know if VM was deleted on purpose, or accidentally (in which case you will want to restore it later). Without keeping the data (v4 style) we had customers loosing data when they were spotting that VM is gone too late (after it was removed by retention). Of course, data loss would be impossible with local or offsite copy of backups most customers do for long term retention, however according to amount of similar support cases for v4, there are too many people not practicing this.

Anyway, we are looking into adding an option automated retention policy (number of days to keep deleted VM data) to the next update, so that every customer could choose how long to keep deleted VM data before it gets removed. I think 30 days would make it a good default value, or is it too much?

Thanks!

[kkoehn]

I would agree that an option to remove deleted VMs would be a good option. I would also say that automatic reconciliation would be good as well instead of failing an entire job because you forgot to remove a VM from a job.

In response to my original problem... the VMs are still alive and still in place -- they were not deleted. The restore points are just hanging around on disk and not being removed per the restore point removal schedule.

I've opened up a support ticket so we'll see what happens.

Thanks!

[Gostev]

Thank you, the logs will show what is going on immediately. By the way, if you are using incremental backup, also make sure you are not overlooking its specifics... see my explanation in the following topic: Rollback points and full backups.

[Chriss]

I am also a little confused here. What is the best option to use, "Reverse Incremental", or "Incremental"? I only want 8 restore points locally on disk as every weekend we write the back-ups on disk to a tape, and send to off-site storage.
Is there a difference in performance between the 2, as it seems to me, that the "Reverse Incremental" is a lot slower.

[Vitaliy S.]

Chris, please look through the thread mentioned above and our F.A.Q. topic that covers all the differences between those modes (see v5 Choosing Backup Mode.zip file). Hope it helps!