Hi @Luca11 -
So, let me see if I understand your question/post correctly. You use an “all-in-one” Veeam server at site 1 to replicate VMs to site 2. For whatever reason, Veeam server at site 1 is down and all or specific VMs at site 1 are down & so to get things back up & running you power up VMs at site 2 manually without Veeam? And, you’re wanting to failback VMs back to site 1 when the Host at site 1 is back up?
Well, this can’t be done sadly. The way around this is to have 2 Veeam servers → one at site 1 doing backups and one at site 2 doing replication (from site 1 to site 2). That way, when site 1 resources go down, you *do* still have a Veeam server to manage your replication tasks with. You perform failover tasks with the site 2 Veeam server and, when site 1 resources come back online, you again use site 2 Veeam server to failback all VMs you failed over back to site 1.
Let me know if you have further questions.
hi @Luca11 is impossible to install your Veeam on the DR ESXi ?
hi @Luca11 is impossible to install your Veeam on the DR ESXi ?
i would also prefer the input from @Stabz.
Additional, if you have a target RPO of 24h for your backups and replicas, you could also use a backup or backup-copy (maybe in the DR site) repository on the DR site as your source. So you don’t have to read twice from your productive vSphere environment. But this depends on your target RPO.
Best regards Markus
@Luca11 -
About the only thing I know you can do to get a copy of your Replica VMs back to production via Veeam is to create File Copy jobs and copy the VM folders on site 2 Datastore(s) back over to your Datastore(s) on site 1. it would be a slow & manual process. You would probably need to power down the VMs at site 1 & 2, remove the VMs at site 1 from Inventory in vCenter, probably rename the VM folders on the site 1 Datastore (I wouldn’t delete them until your sure the File Copy process works as expected), create a File Copy job to copy VM folders from site 2 over to site 1, add each VM to Inventory at site 1, then power the VMs on. Yes, it is manual, but an option. I recommend pinging Veeam Support to see if they have a better solution just to get you back up and going.
In moving forward, this is my recommendation though → if you have a small virtual environment, and it sounds like you do by the description of your Veeam environment being an all-in-one server, I would at least add an ‘external’ VM LINUX Proxy to each site...site 1 and site 2. By external I mean an additional Proxy outside the Veeam Server default Proxy. If OS cost is an issue, that’s why I suggest adding Linux Proxies because Linux OS is free. I created a nice step-by-step of how to do this in a Hub post (see below):
Use a Veeam Server at site 1 just for backups to keep your backups local, performant, and with relatively quick restores. At site 2 I’d create a separate Veeam Server solely used for Replication. Since Veeam doesn’t require licensing for target Hosts, you can use the same license as your site 1 Veeam Server. Assuming you have good connection between site 1 and 2, use the ‘direct’ approach for your Replication jobs, as well as the new Proxies I suggested to create. And in the event site 1 goes down, you have a Veeam server to manage your replicas with at site 2. You could just use 1 Veeam server if you really wanted to. And if you do continue this route, I recommend placing it in site 2 as others have also suggested...again, so you have a Veeam server to manage at least your replicas with in case site 1 goes down. If you have a really small vSphere and thus Backup environment (I’d say less than 40-50 VMs?), 1 Veeam server is probably enough. I just suggest using 2 as a failsafe.
Hope this helps. Let us know if you still have questions.
A workaround is possible, install a temporary VBR on the DR site and do a new replication with the VM in production from the DR site.
Thank you all! You confirm me that is better to have a Veeam server on my DR site to perform eventually failback.
Have a nice day
