Source Vcenter Lost & Failover

[SachmoNZ]

Hi there,

I've got a concern and hoping people with further experience with Veeam than I can offer some insight. I'm really spelling it out so people get an idea of my concern and how I got to worrying about it.

The one line concern is "If replicating between two VCenters, and source VCenter is lost, will initiating a failover be able to skip past powering off the source servers and carry on RE-IP and bringing up the DR replicas?"

If that doesn't give you enough info here is a bit more detail..

The setup is as below

Source VCenter = NZVC
VM1
VM2
VM3

Are all replicated to Destination Vcenter = AUVC

We have RE-IP rules in the replication job and failover plan which all work great.

We run all our jobs, replications and management from the NZ backup/proxy/wan acceleration server
We have the proxy/wan role installed on a VM at the DR site and a powershell script that exports the config from the active server in NZ over to DR. We have successfully tested restoring this without issue whilst the NZ server is offline.

However my concern is if we have restored the Veeam console, and the SOURCE VC is down and we initiate a failover - veeam normally shuts down the source VM which will fail, and that's fine.. but will it stop there and not attempt to RE-IP the servers and basically leave us having to power on the replicas manually and re-iping them ourselves?

Unfortunately I don't have a lab i could emulate this easily on (also time pressures) so help is appreciated.

Cheers
Sash

[foggy]

Hi Sash, the fact is Veeam B&R doesn't care about source VMs at the failover (unless you're doing planned failover, but you're not meaning it, if I'm getting your scenario right). Typically you do failover in case your source site/VMs is/are down, so no need to shut them down. So if the backup server has access to vCenter Server where replicas are registered, you should be fine. That's why, btw, it is generally recommended to place backup server responsible for replication in the target site, just to be able to quickly failover in case of total source site loss.