Proper way to back up Veeam server itself when used as a replication proxy for hot add

[veeeammeupscotty]

We have two veeam servers that replicate VMs back and forth.

Scenario:

- Server 1 completes backup job
- Server 1 starts replication job to server 2 using hot add on server 2
- Server 2 starts backup job which includes itself
- Because replication is ongoing and the replica disks were hot-added, replication fails for the VMs that were replicating when backup job started


I've thought about adjusting the job timing to minimize overlap, but since the durations are variable, it's difficult to achieve and still fit in the window. My next idea is to try and add disk exclusions for veeam server 2 so the replica disks don't get picked up, but this isn't ideal either because if a local disk is added to the veeam server in the future, it wouldn't be automatically backed up. It might be worth mentioning that this only became an issue on V11 since upgrading from 9.5, so I'm curious if there was a specific design or behavior change that could be contributing.

[PetrM]

Hello,

According to our best practices, it's recommended to back up the configuration of Veeam server but not Veeam server itself. Instead, you may prepare a VM on which you will restore configuration in case of disaster.

One more point is that it's inadvisable to use one Veeam server as a proxy for another Veeam server as it could provoke compatibility issues after upgrade. I'd suggest to separate both instances of Veeam B&R from each other.

Thanks!

[veeeammeupscotty]

So even if another Windows VM serves as a proxy, the same issue would present itself if that proxy is backed up during replication would it not? A physical proxy would not, but then hot-add wouldn't work. On this page it doesn't say proxies shouldn't be backed up, only that CBT will be disabled:

If you back up proxies that use the Virtual appliance (Hot-Add) mode to process VM data, the change block tracking mechanism (CBT) will be disabled.

In the link you sent, I do not see mention that it is not best practice to back up the Veeam server, but only that is is recommended to configure a configuration backup job.

[PetrM]

If you back up Hot-Add proxy, then it will be better to schedule a job for this proxy when it does not process any VMs. You may see the following limitations are listed in the second link:

It is not recommended to back up the backup server configuration using backup jobs in Veeam Backup & Replication. For backup, Veeam Backup & Replication uses VM snapshots. During snapshot creation and commit, the VM freezes for some time, which can potentially lead to the following consequences:

• Disconnection from the configuration database. For more information, see the Veeam KB1681 article.

• Disconnection from remote Veeam Backup & Replication agents.

• Disconnection from network storage (for example, storage presented through iSCSI) and so on.

Thanks!

[jnash]

Hi,

Currently have Veeam 12 running on a windows VM.

Is it possible to set up a replication job within our Production environment to replicate this Veeam server to our DR environment?

I have tried but get the "Resetting CBT per job settings for active fulls failed Details: The operation is not allowed in the current state.
Failed to create VM snapshot." error.

[PetrM]

Hello!

It would be best to keep the backup server on DR site and back up its configuration. A backup server can orchestrate backup and replication jobs on your production site as well, just install a proxy server in the production site and add it to Backup Infrastructure of the backup server on DR.

Thanks!

[StoopidMonkey]

If the VBR server is at the DR site and the DR site goes offline, how then would we run the backups at the primary site?

[PetrM]

Hello,

You will need to restore a configuration backup on the primary site and continue backups.

Thanks!