Friday's synthetic full backups repeatedly missed

[Basti]

Hello Everybody

We have the following backup configuration:
4 VMs backed up on weekdays (mo-fr) at 22:00 with synthetic full backups scheduled for friday.
All 4 backups are configured to keep 7 restore points.

The backups are chained (schedule -> run automatically -> after this job: **) to avoid having to worry about their respective time of completion.
Now, as others have reported, backups starting as late as saturday or sunday do not create synthetic fulls.
Which in turn leads to indefinite incremental chains.
Allowing synthetic fulls on saturday and sunday on the other hand creates additional/unwanted backups - thus ignoring the base setting.

As far as I can tell, the only solution to this problem is to start all 4 jobs simultaneously.
Would this be the recommended way to go and what are potential drawbacks of simulteneous starts?

Thanks for your feedback.
Sebastian

[HannesK]

Hello,
and welcome to the forums.

In general it's recommended to avoid job chaining, because it creates dependencies and non-optimal resource usage (with 4 VMs it's probably irrelevant )

In most situations it also makes sense to put all 4 VMs in one job. By doing that, the configuration is easier (again, probably not a big deal with 4 VMs, but for example 400 jobs for 400 VMs would be painful)

Best regards
Hannes

[Basti]

Hello Hannes
Thanks for your recommendation.
I assume it would break the current backup file structure if I put all VMs into a single job - right?

Take care
Sebastian

[HannesK]

Hello,
nothing breaks The software creates a new full backup for the VMs that are added to the job.

Best regards,
Hannes

[Basti]

I just created and launched a new backup job including all 4 VMs.

Veeam creates a new folder in the target repository and merges all VMs into a single file. The existing backups are left untouched.
That's what I meant by *breaking* the backup file structure - it would start a new backup chain rather than continuing the existing one.

Cheers, Sebastian

[foggy]

If you hadn't created a new job but rather added the missing VMs into the existing one, it wouldn't have broken anything but continue using the existing backup chain, just adding the new VMs data. Also, if you prefer to have a separate backup chain for each VM, enable the per-VM backup chains setting in the repository settings.