2 Drives, 2 Jobs?

[john.aubrey]

Hello,
I'm trying to figure out the best way to taking our synthetic full backups we do friday, and writing them Saturday/Sunday to take. Right now we don't use "Per-VM" backups. We have a tape library with 2 drives in it. I'm trying to wrap my head around the best practices: https://helpcenter.veeam.com/archive/ba ... ing.html#2
I enable the "Enable parallel processing for tape jobs using this media pool" and "Enable parallel processing of backup chains within a single job".

I want it to write to both drives at the same time, and be as best practice as possible. Do I create 2 jobs and ballpark total storage and decide the jobs up as eventually as possible? Can I do one job and it will split the the backups up as one drive become available. It appears from what I understand, either of these will work? Any best practices? I'm leaning to on job and let it figure it out.

BTW we are on V10

[HannesK]

Hello,

1) I would start using per-VM backups. Because if you want to restore one VM / file, it will save a lot time.
2) if you point both jobs to the same source, it will write it to tape twice. (I hope I did not misunderstood your "ballpark total storage")
3) You can do it with one job, if you have per-VM chains (my recommendation from 1)

Best regards,
Hannes

PS: BTW we are on V10 -> oh, then I would kick out tape and go with object storage. I'm too young for tape

[john.aubrey]

Thanks for answering my questions. It appears that per-VM is a backup repository level feature. Any harm in changing it? We will be moving to a new backup repository in lime 3 weeks. Would it be good to wait until then?

[veremin]

I would wait for repository migration, then, as enabling per-VM backup chains requires new active full backup (either manual or scheduled). So, I would do that after migration is over. Thanks!

[nathano]

a fellow tape user, nice to see

how many backup jobs are you selecting within the tape job?

[john.aubrey]

I would wait for repository migration, then, as enabling per-VM backup chains requires new active full backup (either manual or scheduled). So, I would do that after migration is over. Thanks!

Is there any downside to per-VM? Sounds like I might lose my space savings to storing the servers as a group (dedup) vs each their own. Is this the only draw back (looking to see if it will be slower in any aspect)

[HannesK]

Is this the only draw back

yep (at least I never heard anything else)