There are definitely users running Temporal/Cadence with Aurora so that works. As for the replication strategy you mentioned, I’m not sure what the best approach is. Let me tag someone who does.
This area has always been confusing, are you saying with XDC settings, each cluster can have its own database and temporal will replicate all info from the primary namespace/domain to secondary automatically?
Each cluster is practically independent with its own database.
Each namespace can be active in one cluster at a time.
Different namespaces can be active in different clusters. So there is no failover cluster command, only failover namespace one.
Data is replicated asynchronously (currently using Kafka) between clusters.
In case of inconsistencies caused by network partitions and replication delays Temporal performs conflict resolution ensuring that no signals are lost.