Veeam Backup server DR

[markotsg]

Hello
We have Veeam backup setup in place and would like to start replicating vms as well.

Our setup is Veeam 9.5 on Vmware 6.0. Veeam backup, proxy role reside on single physical server, and other 4 physical servers are Repository servers and contribute to storage pool.
Backups are done locally to these servers, but then backup copy jobs are done to Cloud.


We are working on new DR environment (this will be separate cluster in same Virtual center) and we will be using Veeam to replicate the servers from production to new DR cluster.
For this i will be designing protection plan to cordinate the replication and recovery in correct priority order.
With this in mind, i would like to re-design our current veeam setup so that in the event of primary site loss, we can still access Veeam console and Veeam backup server role to cordinate the recovery and failover plan recovery on DR.
What would you recommend to make this possible? In the event of site loss, i can build a new veeam server and restore the config but this is manual process, not sure if separating Backup role on another server (that will be residing in DR) might be better way?


Also as mentioned 4 physical servers contribute to overall storage pool, would it be possible to split them across sites? Production 2x physical servers veeam proxy, DR, 2 Physical servers proxy
In the event of site loss, we may loose 50% storage space for backups but not 100% like with current design

I am relatively new to Veeam, and your advice would be really appreciated

thanks all

Marko

[DGrinev]

Hi

not sure if separating Backup role on another server (that will be residing in DR) might be better way?

That's the recommended approach to deploy backup servers at the DR site and manage replication jobs remotely.
It allows you to failover quickly without manual intervention in case of disaster.

Also as mentioned 4 physical servers contribute to overall storage pool, would it be possible to split them across sites?

Sure, you can have proxy servers at both sites with repositories and that allows you to keep backup traffic local at each site.

Thanks!

[PetrM]

By the way, remote replica from backup might be useful in this scenario because it minimizes workload on the production environment by using existing backup chain on a repository as a data source. You do not need to create a VM snapshot for replication and transfer the same data twice.

Thanks!

[markotsg]

thank you, do we build a new server and migrate Veeam backup server role to this server? Server will reside in DR location

This will be perfect from the point of replication failover, but how would it affect server backups? if Backup server resides in DR and all 4 proxys in main sites?

[DGrinev]

Since all responsible components for backup and replication are present at the production site, you shouldn't face any issues.
The backup server is acting as a remote management center for the whole backup infrastructure. Thanks!

[markotsg]

What would be the best way to achieve above, and migrate the backup server role to new dedicated virtual machine , (this will reside in DR) and remaining physical servers will remain in prod DC?

Will this new setup affect running backups and backup performance?

[PetrM]

I'd suggest to take a look at this option to migrate configuration database to another server.

I wouldn't expect that backup performance could be affected as long as you have proxy servers and repositories residing on your primary site.
Just don't forget to make sure that these proxy servers are specified in job settings.

Thanks!

[markotsg]

Thank you very much, i will take a look at your link.

So i can go ahead with either scenarios below, do you see one been better then the other?

scenario 1: Keep All proxy and repository servers in Prod site - just build a new server (Vm) in DR and install vm as Backup server, nothing else required Veeam wise in DR?
Scenario 2: Split 50% 2x Veeam hosts (proxy and repository) and move them to DR, Build a new VM as Backup server in DR, leave 2x Veeam servers (50%) Proxy and repository in Prod

[PetrM]

The first scenario seems to be ok.
Not sure about advantages of the second one, it allows achieving the same goal but it looks over complicated.

Thanks!

[GregorS]

You could also keep the Veeam backup server role in a VM on the production site and replicate it over to the disaster recovery site. You will only need to split your physical backup repositories between sites if you also have some VMs that are actually running on the DR site and you want to back them up too imho. For replication you already have your outbound proxies on PROD, you would also want a (virtual) proxy on DR for inbound traffic.