Replication Failover Considerations

[jdodd6]

While setting up Veeam Replication as a primary method of disaster recovery failover for two Hyper-V clusters, there are two design issues I have been trying to plan around when creating a failover plan and process documentation.

1. Failover of the Veeam B&R server itself.
   • The plan is to restore the configuration backup to an empty PostgreSQL database on a Disaster Recovery Veeam B&R server.
2. Loss of the replication job metadata.
   • The backup repo holding the replication metadata is at the primary site. Can a Veeam replication failover take place successfully if this repo and metadata are missing?

[Mildur]

Hi Joel

1.) Your plan works, but our best practice is to run the backup server in your remote site. This way, you can immediately start your replicas without having to deal with configuration restore first.

2.) The replication metadata is not required to power on Replica VMs with the Veeam backup console.

The purpose of the metadata is explained in our helpcenter.

Best,
Fabian

[jdodd6]

Are there any plans for developing a clustered instance of Veeam?

[Mildur]

Yes, we will introduce a High Availability feature for the backup server in the upcoming release v13.0.1 (Q4 2025).

This HA feature (active/passive) will be available exclusively for our Veeam Software Appliance (Backup Server on Linux). Configuration will be replicated from the primary backup server to the secondary one.
If the primary backup server becomes unavailable or during planned downtime at the primary site, the secondary backup server is able to take over the entire backup infrastructure.

- Veeam Backup & Replication v13 - Whats New, Page 8

Best,
Fabian