Protecting the Veeam Backup & Replication server

Andre - I do both. In the event I need to recover VBR, as long as no jobs were running at the time of my VBR image backup , I would probably use my image backup to recover. Though Veeam recommends config backup, and I do backup my config as well, to me, it’s quicker (less steps) doing image-based restore.

I also do image backups for potential recovery of files I have on the VBR server (scripts, files, etc). In all the yrs I’ve used Veeam, which is about 15yrs or more, I’ve thankfully never had to recover Veeam by either means. I have done the config backup Migrate restore option last year when I was building a new environment and wanted to migrate away from SQL Xprs to PostgreSQL. It was pretty seamless/easy.

If I were teaching Veeam classes such as you, I would share the same thing...to only use Config Backup. Although, I’d be honest and share I’ve done both.

Best.

Only use configuration backup myself here and it is sent offsite to immutable object storage for all our sites.  I have done recoveries or migrations this way and it is fast.

Hi Andre, thanks for this - i completely agree with this.

But what is your opinion regarding replication of the VBR server, because it may also causes problems?

For sure only this VM in a job, only when no other job is running etc. but still..it could cause issues?!

I like the replication approach, maybe to an offsite DR ESX Host, because it’s faster to boot the VM.

What do you think?

matze

I don’t backup our regular VBRs but I do backup the ones running my VBOs. You don’t need anything but the config to restore but I’m a fan of options if possible.

Rep or backup snaps of VBRs might freeze I/O. 

I only do config database backups and those previously landed on a DFS share that was replicated between sites, but since that share is no longer replicated, I need to move it to a different repo.  Likely looking at our Object storage repo. Often I’ll put that in a VCC repo hosted by our Service Provider Console, but since the console is on the same infra as our primary compute, the cloud will be the better solution.

I see, thanks for your comments and honesty :-).
I believe that more than 95% do only image backup, I hope you continue without the need to use restore, by the way, what is your name?

Great to hear that Cris, thank you for sharing your experience .

Hi Matze, 

the user guide recommend not use BJ or RJ either, but I need to find out if using CDP will be fine or not (I’ll do that).

we have more options to protect, we have the possibility to use vSphere Replication or Fault Tolerance I will check if there is some documents saying that is possible or not using VMware features. 

https://helpcenter.veeam.com/docs/backup/vsphere/vbr_config.html?zoom_highlight=Do%20not%20back%20up&ver=120

Thank you for you comments. 

Thank yu Dereck for shariong your solution

That’s right, snaping it may freeze the I/O.

I have my Veeam server in SRM so if my main site goes down I can spin it up at the DR site.  That being said, I also have a copy of my config backup and keep that at both sites as well. Even a simple script to copy those files to immutable cloud storage can be a life saver. 

You can take an old server, install Veeam on it and power it off waiting for a disaster as a time saver if you need. 

VM backup MIGHT work, but is not the recommended way. Config backup WILL work. 

I see Scott, but the manual says do not use “snap technologies” becuase you can have some issues, such as: disconnect from database, VBR and iSCSI network. it is not a restore matter, is is also lossing connections during backup time.