WAN Replication Modes
In clusters connected with WAN Replication, a cluster can have one of two roles: Active or Passive, conceptually explained in the previous section above.
With these roles, there are two modes of WAN Replication:
Active-Passive: This mode can be used for failover scenarios where you want to replicate an active cluster to one or more passive clusters, for the purpose of maintaining a backup. If the active cluster becomes unavailable, you may redirect user traffic to the passive clusters. Once the active cluster becomes available again, you may again redirect user traffic to the active cluster. The active cluster may have however been started empty or might have simply missed any updates that have happened on the passive clusters. Any updates made on the passive clusters will not be replicated back to the active cluster and if the data is out-of-sync, it will not be synchronized. If you require that these updates be copied back to the active cluster, you may consider using active-active mode instead.
Active-Active: Every cluster is equal, each cluster replicates to all other clusters. This is normally used to connect different clients to different clusters for the sake of the shortest path between client and server, hence gaining increased performance. An example use case would be geographically distributed applications. If the same key is modified on multiple clusters when using this mode of replication, data - including any relevant metadata such as expiry information - across all clusters eventually matches the last replicated transaction, assuming a transparent WAN merging policy.