Why old backups are not being removed

[sandroalves]

Hi,

why are old backups not being removed?

Before it would normally overwrite, but after I enabled GFS and create synthetic full it didn't overwrite anymore.

Thanks.

[Egor Yakovlev]

Hi Sandro.

- When you have Forever-Forward incremental chain(no periodic fulls) Veeam can just cleanup oldest restore points. You see as many backup files on the disk as set under Retention policy.
- As soon as you enable periodic fulls, Veeam switches to "classic" Forward-Incremental backup, where retention applies differently. On your screenshot you have 5 restore points in the latest Full+incrementals chain(5/1->5/5) - in this case if Veeam deletes what you have marked yellow on a screenshot(old backups), your backup chain will breach set Retention Policy(RP=7), leaving you with less restore points than you wanted. Chain will be cleared in 2 days, when you get 7 restore points to leave after cleanup. With this type of a chain you can expect more restore points on disk.
- We have a wonderful animated retention KB that explains all modes and cleanup methods.

/Cheers!

[sandroalves]

Hi Egor,

I enabled (synthetic) after a training that I received because it is a good practice, but I did not expect that it would increase the consumption of space like that.

this means that he will always keep a previous chair similar to the retention policy, that is, doubling the space.

if i disable synthetic full in the next backup will i always only have one full (.vbk) and the incremental ones (.vib) always?

Thanks.

[Egor Yakovlev]

Well, if you want to use Backup Job GFS as on 1st screenshot, Periodic Full must be enabled(or triggered via script) and job will keep more restore points to satisfy Retention Policy settings.
Note that it does not really "doubles" the space(just in your 7+7 case), as total number of restore points on disk will be equal to [Retention Policy period + Periodic Full period]. For example, with RP = 30 and Weekly Full, you will have 37 restore points on disk before retention can cleanup oldest 7-points chain.

As far as I remember, switching from "classic" to forever-forward mode(aka disable periodic fulls) shall be accompanied with manual Full Backup to start new chain retention type.

[Gostev]

this means that he will always keep a previous chair similar to the retention policy, that is, doubling the space.

This is why you should use ReFS or XFS for your backup repository, which is also something we recommend at the training... with these, synthetic fulls do not take any physical disk space at all, and previous chains take very little - because most blocks are reused in the following week's restore points anyway.