Design question

[fslot]

I am moving away from backup exec and have 2 backup servers available to me for Veeam BR enterprise with storage and tape drives attached to each. I have about 100 VM's to back up on one Vcenter, about 100TB of total data.

Any recommendations on design? Use both servers or only 1? And any preference on local sql express vs using enterprise sql db on another server?

[Rick.Vanover]

Hi fslot, welcome to the Veeam Forums.

I would generally recommend using both servers, and ensuring our beloved 3-2-1 rule is in place. It says to have 3 different copies of important data, on 2 different media, and 1 being off-site. Add bonus points for 1 being offline / air-gapped.

So if you can build around those principles with the servers, that may help as a starting mindset.

Beyond that we generally recommend building backup jobs on vSphere tags (set on a policy) and having the best possible backup storage for the first stop of the 3-2-1 rule for performance.

[veremin]

Are going to archive to tapes only backups created by Veeam Backup & Replication or also some files (say, residing on a file share)? In the later case consider switching to full-blown SQL server. Thanks!

[fslot]

I have 1 vcenter with about 100 VM's and about 100TB i want to back up to disk and then offload to tape once a week. I have 2 servers that i am currently using for backup exec and they each have their own local attached storage and local attached tape drives. with Veeam would I set these 2 servers up independently to back up half or so of the VM's? Or is there a way to set it up so Veeam is aware of both backup servers and splits the load for me? Do i set one as a proxy for the other?

[foggy]

You can combine them into a scale-out repository for spreading the load. However, initial Rick's suggestion was to have copies on both of them to ensure the 3-2-1 rule, I guess.

[fslot]

so with the scale out repository I install Veeam B&R on one server, add the 2nd server as a scale out repository, and backup jobs created on the first server will be spread out to run across both servers simultaneously and automatically?

[foggy]

To be entirely correct, you create repositories on both servers, add them as extents to a scale-out repository and point your jobs to scale-out repository. Data will be split between extents according to the configured policy automatically.

[fslot]

So I would install Veeam B&R on both servers, create repositories on both, and add them both as extents to a scale out repository? Which server are the jobs created on? Is Enterprise manager used/needed in this case?

[foggy]

You do not need to install full Veeam B&R on both, you just create repositories on both servers in a single Veeam B&R installation.

[fslot]

ah, i see. Thanks!