Datto Backup + Veeam

Posted by GhostNode@reddit | sysadmin | View on Reddit | 4 comments

I just took over a new environment. In it is a Hyper-V VM running RedHat that I just started backing up with a new Datto. They were only doing file-level backup of this VM prior. The VM hasn't been rebooted in over a year, and while the Datto backups succeed, and I can mount and access the files in the backups, they fail to boot in instant VR, or via a restore to the Hyper-V Host. I'm not sure if the production VM has a corrupt file system (I'm not afraid to reboot it), or if the issue is just with how Datto is backing up the VM. Due to... reasons.. there is also a Veeam backup solution in this env. I know other RHEL VMs on this host are backing up, and restoring to Hyper-V properly with host-level Veeam backups. I'm inclined to add this VM to a job and see if that backup will restore. Question being: If I pause Datto backups before kicking off a Veema job, does ayone foresee issues with the two solutions running on top one another?

Reply to Post

4 Comments

[-]

GhostNode@reddit (OP)

\*\*UPDATE\*\* Suspended the Datto backups for that VM. Added the VM to the Veeam backup job. Ran it. It completed without issue. Restored it to a different Hyper-V host, boots without problem. Issue seems to be the Datto backups and the linux Datto backup agent, not the prod VM (wipes sweat from brow), and no conflicts were experienced when running the Veeam backup job on a VM with the Datto agent installed. \-Hope this helps some googler in the future. Good luck. Praise be to dog.

[-]

ledow@reddit

I wouldn't touch Datto with a bargepole. An MSP was brought into my workplace ("to show us best practices") and tried to force my in-house team to use it and it was a pile of absolute junk. We never got it working anywhere near reliably and it was atrocious for things like Bitlocker, TPM, restoration and Linux VM support. After several instances were NOT ONE of our "instant restore" VMs were required and literally never cleanly restored once (and often required stupendous intervention to do things with Secure Boot and resetting UEFI keys etc. in the VMs to make them work, despite the fact that the VMs they were backing up worked just fine and our previous backup software - Veeam! - worked immediately and perfectly with them), we abandoned their product entirely and sought refunds. They used to lock our production VMs into states that meant we couldn't back them up with anything (thereby requiring our entire cluster to restart!), they never restored reliably - not even once. You will get VM locking issues with Datto... their backup process is basically a load of hacked-together nonsense. Datto appears to be a bunch of junk, top to bottom. And that's when you are just using it on its own, let alone when you try to use it combination with another backup solution. We binned it, and it contributed massively to our removing that MSP entirely and falling back to our previous Veeam-only solution (along with the time that the MSP dialled into our system without asking, APPLIED a VM checkpoint that was months old, thereby rolling back our complete systems back months in time, then denied all knowledge until the logs were shown and also our recordings of us WORKING ON OUR OWN SERVERS, and then them logging in and kicking us all out without asking, and then in their remote sessions their technicians went about "applying" the old checkpoints instead of "deleting" them, and then hastily logged on when they realised what they'd done). Do yourself a favour - throw Datto away and stop using it. We had their expensive appliance that was supposed to be able to host EVERY VM we had on our cluster in an emergency - it never worked for even a single VM - the full cloud backup package and we had to uninstall our previous Veeam because of all the bad interactions with it, and we ended up undoing all that and just using Veeam as we always had (some £100,000 later!).

[-]

Party_Trifle4640@reddit

I’ve seen similar situations in mixed backup environments. I work for a VAR and have helped customers clean up legacy setups like this before, especially when backup tools have been layered over time without a reset. A few quick thoughts: If Datto and Veeam are doing host-level backups, they might conflict at the Hyper-V snapshot level depending on how VSS is handled. Pausing Datto while testing Veeam on that VM is a smart move to avoid I/O issues. If the VM file system is corrupt, Datto may have been backing up bad blocks all along, Veeam might give you better feedback during backup or restore, especially with SureBackup. If you’re thinking about switching or consolidating, I can help scope a clean backup and DR strategy that avoids these headaches, whether that’s staying with Veeam, moving off Datto, or using a hybrid for now. Shoot me a dm if you need more support!

[-]

GhostNode@reddit (OP)

Thanks Mate. Datto is using agent-based backups, so as long as I pause the Datto backups while the Veeam host-level snapshot-based backup runs, I'm \*thinking\* I should be in the clear. Was wishing on a star someone here is going to be like "Matter of fact I've done that, and its fine!" or "Dear god what ever you do, don't do THAT!" I'm planning on seeing if I can repair the filesystem in the VM that won't boot, which will give me a little more hope, and will keep my eyes on this post to see who says what. I appreciate your input!