Speeding up my backup

[rene@aksessikt.no]

Hi Y'll
I'm just bringing online my new Hyper-v failover cluster on 3 DL380.
this cluster is using shared storage that's supplied by a CiB from DataOn, storage is made availeble using storage Spaces.
storage cluster (csv) talks to the Hyper-v cluster trough 10gbe nett (failover through 2 mellanox switches)
All this works great. i got 80 something servers on this setup. And VEEAM 9.5 release3 works great.
Inside this setup i got 2 windows2016 servers that are clustered and use a VHDS disk set as shared storage, they supply the fileserver role. This works great.

I got a dedicated server (DL380 with 180 gb ram and 2 cpu core's with 12 processors each)
i'm thinking of putting a mellanox nic in this server and insert it in the storage nett. I'm also looking to run more than 4 threats (jobs) paralell since I've all these processors an loads of memory.
I'm backup to a build in raid 5 set (that's fast and direct conncted) and the less critical stuff to a iSCSI connected NAS.

- is there any gain by inserting the server in the storage network? and special way of configuring?
- How can I run more than 4 job? I increased the maximum jobs on jobs, repository

Rene

[nmdange]

It is pretty easy to do. Just add your backup server as an off-host backup proxy, make sure the computer account has access to the SOFS file shares, and also make sure the "File Server VSS Agent" role is installed on your SOFS cluster nodes.
https://helpcenter.veeam.com/docs/backu ... tml?ver=95

Concurrency is dependent on number of objects, not number of jobs. There are a number of places concurrency can be limited, including in the on-host or off-host proxy, the backup repository, and the number of VSS snapshots on each CSV. The easiest way to tell is to look at the job while it's running. For VMs that haven't started, you'll see a message that says "Resource not ready: <resource name>" which will tell you what the VM is waiting on.