Safe to Delete?

[kkoehn]

We've been using B&R 5.0.1 for quite a while now and have had good luck with it -- but we've hit a few snags.

The first problem is the size of the backups and long-term storage of them. Our Exchange server, for instance, is reaching almost 2TB of storage in a 30 day window -- not good! I've changed the backups to Forever Incrementals with a full synthetic on Saturday and set to WAN encryption for the smallest backup in hopes we can calm it down in the future. The problem I ran into is I attempted to clean up previous backups by turning on roll-up and setting the previous restore points to 5 and performed a backup. However, it just ignores the previous TB or so of backups. Is there any way to get rid of them?

Here is one example of what I see:
1-7 VIB
1-6 VIB
1-4 VIB
1-3 VIB
1-2 VIB
1-2 VBK
1-1 VIB
12-31 VIB
... (more VIBs)
12-11 VBK

My question is... are any of these safe to delete? Why isn't the 'restore points to keep' being honored here? I've already got 15TB of storage -- management would shoot me if I need to ask for more

Thanks!

[Gostev]

Generally speaking, it is safe to delete any full backup chain that you do not want to be able to restore from but it is highly recommended to let backup job retention policy to handle this instead. Also, this would only mean that your retention policy is not set correctly, and should really be reduced from 30 days.

What do you mean by "restore points to keep being honored here"? According to what you wrote above, I can see 30 days retention working normally. You will see your 12-11 VBK and the corresponding VIB chain deleted automatically in a couple of weeks, freeing up the disk space.

Also, if backup target disk space consumption concerns you, you may simply consider using different backup modes? Incremental backup mode is least effective for as for disk space consumption due to the need to keep multiple full backups. Please check you backup modes comparison table in sticky FAQ topic for more information on how different backup modes our product provides compare to each other.

Thanks!

[Wil]

[merged]

Good afternoon,
Our backup jobs are stored on shared drives that, are constantly running out of disk space. My question is, which files can I delete? Should I be deleting any files? What would be the proper way to delete them. Numerous times our backup jobs failed due to this issue. Our backup jobs are setup to keep 14 restore points. We run incremental daily and a full once a month. Any advice would be appreciated. Thanks,

[tsightler]

If you are using 14 restore points with forward incrementals and monthly fulls then you are going to have times when there are 45 incremantals a two full backups on disk at one time. Setting the restore point to 14 requires the system to keep at least 14 restore points on disk at all times, however, an incremental restore point is of no value without it's corresponding full backup, so on day 28 the only way to keep 14 restore points is to keep all 28 restore points. Not until the next full backup does a new chain start and then 14 days after that point the old chain could be deleted.

If you only want to keep 14 restore points there are really three options, increase the frequency of fulls (maybe to once a week, which would still cause you to vary between 14-21 restore points), configure a synthetic full to transform old restore points to rollbacks, or switch to reverse incremental backups. There are advantages and disadvantages to each choice so you'll have to pick the option that best fits your environment.

[Wil]

Thank you for the responds. I may just reduce the number of restore points. Getting more storage space is not going to happen.
I’ll have to look into the difference between reverse incrementals and transforming old restore points to rollback.
Also, how did you come up with the 45 incrementals and 2 full backups.

[Vitaliy S.]

Also, how did you come up with the 45 incrementals and 2 full backups.

Please search our forums for existing topics explaining this behavior. By the way here is one of them: Job Set for 14 Mount Points - 46 exist

[tsightler]

Also, how did you come up with the 45 incrementals and 2 full backups.

Well, just think about it. You're only running 1 full backup a month, the rest are incrementals, however, incremental backups are not useful without their corresponding full backups so Veeam cannot delete the first full backup and it's incrementals until your second full backup has at least 13 incrementals (thus satisfying the requirement for 14 restore points). At that point Veeam would delete the previous months worth of backups. Since a month potentially has 31 days, then 31 (first months full + 30 incrementals) + 14 (second months full + 13 incrementals) = 45 restore points.