Feature Request: Auto-Merge Multiple Jobs to one Repodata

[Poweruser]

Actually i have two jobs.
One running daily full to tape by reverse incremental mo-fr.
another job is running manually by time to time by clicking run.

Both jobs do a full backup. The first Job makes synthetic stuff, the later one makes a real full job.
Nevertheless i got 2 times the same data, which wastes repo space.

Because i dont have fixed times where the real full backup is run, i cant set "create full backup" for some days in the incr. job.
So i have to run a second job. (both hyper-v backup)

What Veeam does ist writing the same backup 2 times to repository.
How can i circumvent this.. veeam needs to merge both and make use of the full real backup for its reverse inc.

I hope you got me :-)

[PTide]

Hi,

There are some uncertainties which I hope you could clarify for me

One running daily full to tape by reverse incremental mo-fr.

Did you meant that you actually have two jobs - one is backup job in reverse incremental mode, and another one is backup to tape job running daily full to tape?

another job is running manually by time to time by clicking run.

Is that a backup to tape job, or just a backup job?

Thanks

[Poweruser]

First i describe my jobs.
I got 6 Jobs where 1 Job is the configuration backup which is not a "real" job.

It goes like this:

Every Monday or ... or Friday start Job at 23:00: // normal mo-fr job, reverse incr.
Do Backup Hyper-V selected VMs (all selected) to Repository using reverse incremental and friday full.
After
Do Backup Repo to Tape

Every Tuesday or ... or Saturday start Job at 4:00 (its not monday, because of midnight..):
DO configuration backup (file->backup..) to repo

Every Tuesday or ... or Saturday start Job at 4:10 (its not monday, because of midnight..):
Do Tapejob Configuration from Repo to Tape

Every Tuesday or ... or Saturday start Job at 6:00
Do create new mediaset from mediapool monday-friday with retention 6d

-----

manually there are 2 other jobs:

Every user-manual-start() DO // ABC JOB, full backup native
(not yet implemented by Veeam, because i cant run it before prejob) execute hyper-v vm batch on VM which checks SQL using exe file.
Prejob shutdown all VMs in the right reverse startup (by delay from hyper-v manager setting) order.
DO backup hyper-v VMs (all manual selected) to repo
postjob start hyper-v VMs in right startup order (by delay from hyper-v manager setting)
AFTER
DO tape job for this repo to tape from medial pool ABC with retention 2 months, eject tape
//dunno what todo with mediaset, maybe i set it to create new every backup i think. does not matter anyway i think.

I think thats a detailed enough descr. of my jobs.
Now you see, i got the same data from ABC and mo-fr job. anyway it would be a pleasure to have the ABC job deleted right after its written to tape. much better would be a merge to the main mo-fr repo.
as friday should be a full backup and mo, tue, .. should be full (reverse), too. so its a waste of space.

Maybe i could remove the second (ABC) job and do a tape backup only, but then its not up2date because i run ABC at weekend and the other job on mo-fr.

[PTide]

Why do you have two backup jobs to backup the same VMs, in the first place? As I get it - the vendor's recommendation was to shutdown the VMs before backup. Therefore the question is - why would you use a simple scheduled backup job then? Also the idea of merging backups into a single chain is a little bit on a dangerous and complicated side of design. Anyway, you request has been heard and noted, although I can't give you any ETA.

Thanks

[Poweruser]

i do two different jobs
one online vss with sql server off only while nobody working.
and the ABC with all vms offline.
and yes i could change it to one job only, but then i wont have offline backup.
a merge would save much discspace.
but if merge is risky, just add a direct to tape job or delete first job after taping so space is not wasted.