Surebackup on two servers

[frankive]

Isit possible to run surebackup on two servers? we have a customer with a hyper-v cluster, but they don't have enough ram to run full out surebackupjobs.
Is it possible to run the surebackup onver f.ex 2 hyper-v servers? any challenge to the networking part?

edit:
I also see that surebackup takes more and more place while it is running. f.ex a 20 GB hyper-c free partition on a hyper-v host was increasing down to 6-7 GB while running a surebackup job.
What is cuasing this?

[PTide]

Hi,

Unfortunately there is no multi-host lab configuration for Hyper-V. You can try to split the job into two separate and sync them, is that an option?

May I ask you which server did you specify as a mount server in the settings of the repository where the backups reside at?

Thanks

[frankive]

will tro the sync.

the repository says the veeam server itself is the mounting server. I was hoping all writing was performed to that server and not the hyper-v host itself.
Could it be the paged memory growing with the addition of the vm in the surebackup running perhaps?

[PTide]

Could it be the paged memory growing with the addition of the vm in the surebackup running perhaps?

It might be. It's worth checking the page file size before and after SureBackup.

Thanks

[frankive]

it seems it is the .avhdx files that is growing.

Is it ok to but the veeam virtual lab on the veeamserver itself, and still use the production hyper-v servers as host for the surebackup vms?

[PTide]

I've double-checked something. That's a snapshot file. Every time a VM is started directly from a backup (instant recovery/surebackup) all changes are written into the snapshot. So mount server location has nothing to do with the problem, sorry for confusion.

Thanks

[frankive]

just for the record, i guess the snapshot file always will be on the hyper-v host the vm are started in the surebackup for performance?

[Mike Resseler]

Frank,

Correct. And as Pavel said, the avhdx is the "differencing" disk where all the changes are written. When you stop the job, it literally discards the changes that happened (as in... it throws away the AVHD(X)

Cheers
Mike

[PTide]

Another thing to keep in mind - avhdx files are also created during replication jobs, if you're going to use replication then I'd recommend you to either expand the existing storage where VMs' virtual disks reside on or to move the virtual disks to a storage with a higher capacity. Backup jobs can create avhdx files (depends on the version of Hyper-V) as well, so expanding an existing storage won't hurt anyway. If an avhdx gets too big and no free space left on the device the whole job will fail. Although older Hyper-Vs (2008R2, 2012) do not create avhdx during the backup and use shadow copy instead, they still might crash due to the same reasons.

Thanks