Camunda 8 currently only provides a scheduled backup feature, which initiates a backup task at a specific point in time to back up web applications, Zeebe indices, and Zeebe brokers in sequence, ultimately completing a full system backup.
The issue with this backup method is that if the system fails, any data generated between the last successful backup and the time of failure will be lost and cannot be recovered.
Are there any methods to perform real-time backups for Camunda 8? One approach I can think of is creating dual clusters, where the downstream cluster replicates the Raft log in real time to back up the data of the primary cluster. The nodes in the downstream cluster could be attached to the primary cluster’s Raft replication group as a new role (different from leader or follower), replicating the logs without participating in voting. This way, if the primary cluster fails for some reason, the system could switch to the backup cluster in real time for operation.
I wonder if the official team has any plans to address the backup scenario described above?
Hints: Use the Ask AI feature in 