Multiple restore points in the same tape job with daily synthetic full

[migeauxy]

Dear Forum

I'm facing an Tape job configuration issue that has been lasting for long, and I hoped that with version 10 it would have been solved, but no

Here is the situation :
VBR 10
vSphere 6.7
Disk Target Dell Datadomain
a bunch of VM jobs ( some with a lot of VMs), configured as incremental, with Daily Synthetic full ( it has been configured this way - do'nt know if its best practice or not )
Monthly tape export as secondary job
the tape job is configured to "Process latest full backup chain only" selected, and "Prevent this job from being interrupted by source backup jobs" is not selected

Here is the thing :
my monthly tape job runs longer than a day, and each day at 22h00 the normal daily backup starts.
So my tape job is "paused", because "changes in source backup files detected, rescan required" and "Prevent this job from being interrupted by source backup jobs" is not selected
and when my daily backup is done, tape job resumes

BUT ...

I see in the tape job log, that some of my Vms are processed more than once, sometimes twice, sometimes 3times, because, I suppose, when tape job resumes, well it takes the "latest full backup chain" . and as the latest full is actually from the latest backup (synthetic full as it is incremental with daily synthetic full), it looks like the tape job re-exports the same VM, but another restore point ( the most recent one )
I see in the log the percentage of each restore point sent to the tape, and it goes up to 100% each time, so it's not because the VM restore point is paused at the backup job start. restore point export is complete, but I get another one the day after, and sometimes the day after after

result : Long time running Tape Jobs, physical tapes filling up more than necessary, cannot do maintenance on Backup infrastructure for a week long .....

is there a way to tell the tape export : Export to tape only the restore points that were present AT THE TAPE JOB STARTUP !! and do not recatalog each day what are the latest synthetic fulls ....

some of my tape exports last 4-5 days because of this, and some of the VMs are exported multiple times to tape when it's not necessary

Thanks for your advice

regards

[Dima P.]

Hello Yannick,

The described behavior is unexpected (unless I miss something obvious) - in tape jobs tasks should be restarted if we started processing. When tape job starts it must process only the existing restore points on disk, ignoring the ones that were created while tape job was running what you have on disk. Can I ask you to raise a support case and share the case ID with me? Thank you!

[lyapkost]

Hi Yannick. Have you considered using GFS media pool to achieve you goal? You can enable Monthly media set only, and it should perfectly fit to your case.