Regular Repositories not creating Synthetic Fulls

[rogerdu]

I have a regular repository that seems to not be creating its weekly synthetic full backups... There are 4 jobs that run against the repository and of the ones that are linked (3 jobs), the two that have the setting "Run After This other Job" never create their weekly synthetic Full backup (even though the check box is there in the advanced options).

The 2 jobs that seem to never create Synthetic fulls both follow another job (Start after...) and have a secondary job (Backup Copy to a local repository)

We are running 9.5.0.1038

[foggy]

Hi Roger, check the time when the jobs actually start. In case they do not start on the day synthetic full is scheduled at (but start on the next day, for example, due to being chained to other jobs), synthetic full will not be triggered.

[rogerdu]

So, in this case, if the Synthetic Full is slated to run on Saturday, I should not have a Saturday evening chained job?

Interesting how the parent job runs fine... I have over a year of incrementals chewing up space where I should only have a month...

Is there any way to truncate the series of incrementals at a given date so I can keep 4 weeks of incrementals with their associated weekly synthetic full backups?

Thanks in advance,

Roger

[DGrinev]

Hi Roger,

I'd recommend you to get rid of job chaining and use Veeam native resource distribution through limiting proxy\repository concurrent tasks.
1. You can just wait for the next round of synthetic operations without job chaining and let the retention policy to remove the past year of increments, but it brings you another month of backups.
2. Another approach is you can try to disable any periodic fulls in order to switch to Forever Forward Incremental backup method and leave the retention policy to keep 30 RPs, that will shorten the backup chain by injecting exceeded RPs into the existing full.
Since then, you can schedule periodic Synthetic fulls on the desired day, but without job chaining. Thanks!

[rogerdu]

The problem with that is the customer wants Monthly full backups kept for 7 years... I'll go with option 1 but stagger the start times so the Gold, then Silver then Bronze jobs start in a staggered manner.

[haslund]

Rogerdu, you mentioned utilizing backup copy jobs earlier. Why not go with the suggestion by Dgrinev and change the backup job to forever forward incremental and modify restore points to 30. Then handle the 7 years of monthly fulls in the backup copy job(s)? Backup Copy jobs support GFS and it's very easy to handle the scheduling.