Still recommend Jobs by OS Type/Edition for deduplication?

[nd39475]

Does Veeam still recommend backup jobs to generally have all the same OS Edition VM's for space-saving? or is that a negligible thing now?

[PetrM]

Hello,

Deduplication is performed by both Source Data Mover only for virtual disk currently being processed and by Target Data Mover for blocks which belong to the same backup chain so its efficiency depends on per-VM option is enabled or not.

If you use per-VM: only virtual disks of the same VM will be deduplicated. If the option is disabled: virtual disks of all VMs in the same job will be deduplicated.
Therefore, adding of the same OS edition into the same job will allow you to increase the dedupe ratio if per-vm is disabled.

Thanks!

[dirkdigs]

Hello is it still a Veeam best practice (performance wise) to group all of the same version of OS together in the same backup job?
Forgive me it's been a while since I've worked with Veeam.
For example if we have 50 total VM's to back up .

and we have a mix of 2008 OS , 2012 OS and 2016 OS should i be grouping the backups jobs by OS version? Does this provide any sort of performance gains?
Any other suggestions that i should be aware of when setting up new backup jobs for this size of infrastructure. ?

thank you .
PS: This is a ESXI/VMware infrastructure .

[PetrM]

Hello Mike,

I've moved your post to the already existing topic. It makes sense to group jobs by OS version from space-saving perspective, please see the answer above.

I would also recommend to download this white paper and review the section about vSphere tuning in our best practices guide. Also, you may play with The Restore Point Simulator to get better estimations of required free space on the backup repository.

Thanks!