my vCenter failed over, and won't "fail back to production"

[davidserafini]

Hello - I'm running Veeam Backup & Replication 6.5 on a Windows 2008 R2 Standard SP1 VM. I have two hosts clustered with my primary host running 9 VMs including the vCenter server. The other host houses all my _replicas.

I failed over my vCenter as it was having an issue (slow and not responding). But when I select the "fail back to production option" in VEEAM, it seems to be looking for the vCenter server in order to start the process. well this works fine for other servers but how can I fail back over the vCenter server, if it's looking for itself in order to complete the job. I'm running into a catch 22.

In fact I think I'd rather restore the _replica (as that's what my environment is running from at the moment) back to the original host... as an in-place transfer or whatever. This replica is running just fine but i don't like the looks of running from a rep.

any help or advice would be great. thanks all.

[Rumple]

Its actually not supported to replicate or backup the vCenter server as far as I know because it has to always be available.
However its easy to manually do the failback. Do a permanent failover of the vCenter server
Re-do the replia so its replicating backwards...once its in sync, shutdown the replica, power on the prod and manually remove the snapshots from both prod and replica
Then reverse the replication again so the prod is again replicating to the replica.

[Vitaliy S.]

Its actually not supported to replicate or backup the vCenter server as far as I know because it has to always be available.

Actually it is fully supported to replicate vCenter Server and the backup server itself. If you have a remote site, it is recommended to install additional backup server to manage replication jobs from the other site. When source vCenter Server is not available you can still power-up all the replicated VMs via secondary backup server.

I failed over my vCenter as it was having an issue (slow and not responding). But when I select the "fail back to production option" in VEEAM, it seems to be looking for the vCenter server in order to start the process. well this works fine for other servers but how can I fail back over the vCenter server, if it's looking for itself in order to complete the job. I'm running into a catch 22.

In fact I think I'd rather restore the _replica (as that's what my environment is running from at the moment) back to the original host... as an in-place transfer or whatever. This replica is running just fine but i don't like the looks of running from a rep.

What have you specified as a destination target for the replicated vCenter Server? A standalone ESXi host? Or this host was also managed by the vCenter Server that failed? If you were replicating within the same server, then, indeed, it would be easier to do a permanent failover and then and replicate changes back to the production via new replication job and replica mapping function.

[davidserafini]

Thanks for the replies. I'm new to VEEAM and VMWare in general. The goal for me was to simply add an iSCSI software adaptor to both my hosts. I added to the secondary (we'll call it) no problem as nothing was on that host. but my vCenter was on host #1 (the main host). Not knowing to simply connect vSphere client directly to the host to shut down the servers to put in maintenance mode (no vMotion yet) I manually failed over the vcenter (thinking i needed to, my mistake). When trying to fail back it didn't work (fail back to original location). But seeing i didn't fail it over because of a 'disaster' i just powered off the _replica and powered back on the original. That worked ok (although VEEAM wasn't happy with this as it had Running (1) still in the job list, waiting for it to fail back over to production).

again this was ok until the original started acting up and running like garbage.... so I powered it down and fired up the _replica with is acting fine but I don't want my environment to be running like this. so i should just 'permanently fail over' the server in VEEAM? once that is complete will VEEAM be 'OK' with things? then i just reverse the replication for a few cycles? then do another fail back cycle?

thanks all for the information!

[Vitaliy S.]

so i should just 'permanently fail over' the server in VEEAM? once that is complete will VEEAM be 'OK' with things? then i just reverse the replication for a few cycles? then do another fail back cycle?

Another failback cycle will not be required if you create new replication job to replicate failed over vCenter Server back.

This topic "Question about failover with replication" should give you a bit more insight on the failover/failback process.

[davidserafini]

ok great. so... so far I've created a new job to backup the 'working' _replica back to my primary host. All went well. "success". I then powered down the _repelica and powered on the backed up VM (that is on my primary host) and THAT vm is pulling the same stunts as the original. very slow, (actually the screen is still black after 10 minutes of Powering On. No pings, no response.

What would make this VM that ran perfectly for months start acting like this, as well as the freshly replicated copy of the "working" replica? ...no joy.

any info would be great.

[Vitaliy S.]

Hmm... given that replica is the exact copy of your VM, then I would suspect either issue with the source storage hardware or misconfiguration on the VM side. Replication job just transfers the same image from host to host.