Replication between clusters

Allow one cluster to mirror another. Could be similar to read-only replica, but instead of a single node that is known to the source cluster, the mirror would be a read-only replicated cluster.

One use of it is to enable faster reads. If the "operational" source cluster is in one region, and the application is read-intensive, replicating events to another regions and allowing reads from the mirror can offload reads from the operational cluster. It might not be a valid use case because ESDB is normally used with read-write transactions. There aren't many cases where the read-write disbalance can be observed. It might change if ESDB will store current state and allow queries.

More frequent demand is to replicate data from one cluster to another for disaster recovery (DR) purposes. Using cluster mirroring in this scenario would also imply that ESDB needs to support fail-over from the source to the mirror (enable writes to the mirror and stop replication) when the disaster actually happens.