Veeam servers in both HQ and DR locations.

[arbitrix]

Customer has two sites: HQ and DR.

Each site has a Hyper-V server, a Veeam server and a Linux Veeam repository.

Intuitively this is what I would like to do:

• HQ Veeam runs a backup job from HQ Hyper-V to the HQ repo.
• HQ Veeam runs a backup copy job from HQ repo to DR repo.
• DR Veeam performs a "replica-from-backup" onto DR Hyper-V from the DR repo (i.e. updates the replica from the backup copies written by HQ Veeam).

Unfortunately a lot of online folks advise against having a single repository accessed by multiple Veeam servers, even though it isn't entirely prohibited.

In light of this, what is the best way to

• take backups from HQ to HQ
• ship backup copies from HQ to DR
• maintain replicas in DR

while avoiding multiple transfers of the same data over the slow HQ-DR link?

It would seem the following approach doubles the bandwidth requirements on the HQ-DR link:

• HQ Veeam runs a backup job from HQ Hyper-V to the HQ repo.
• HQ Veeam runs a backup copy job from HQ repo to DR repo.
• DR Veeam replicates from the HQ Hyper-V to DR Hyper-V.

What am I missing?

[foggy]

You can achieve what you're after with a single Veeam B&R instance, no actual need to have a separate one in each location. The same instance can be responsible for regular local backups, subsequent backup copies, and replication from those.

Here are more details on why sharing components between Veeam B&R instances is not recommended. And here's a bit on how to implement that in case you still go this path.

[arbitrix]

Thanks @foggy,

The main reason to have a VBR server in both HQ and DR locations is for when the HQ-DR link is down.

I know we must have a VBR server in DR in case HQ becomes unavailable and we need to re-IP/activate replicas.

But when the HQ-DR link is down we must continue to take backups in HQ so that also requires a VBR server in HQ.

[foggy]

You can sync Veeam B&R configuration from DR to HQ to be able to quickly fire up a new instance and continue backing up in such a case.

[PetrM]

Hello,

Agree with Foggy, you may deploy just one VBR instance on DR to orchestrate replication jobs from HQ to DR and backup jobs on HQ.
You could run configuration backup from DR VBR server to HQ repository and restore configuration on HQ using "migrate" wizard to continue running backups if the HQ-DR link is down.

Once the link is up again: turn off restored HQ VBR and rescan repository on DR VBR to update configuration database with actual backups which were created during the HQ-DR link outage.

Thanks!