If we set up Zeebe broker to manage our workflow, the workflow states get saved in the RocksDB. How do we manage the disaster recovery and region failures for the states stored in the RocksDB? In case you can guide us here.
The RocksDB state is replicated across brokers in the cluster. There is nothing that you need to do about that.
For region failures, you can deploy Zeebe brokers across multiple regions. This incurs a performance overhead, because of the first point: RocksDB state is replicated across brokers in the cluster.
hope this helps,