During the last weeks I did a lot of research on High Availability for Postgres and MariaDB. Both by my own research and also recommended by vendor's, replications seems to have more advantages.
Some benefits as using a standby node for backup; scaling; the fact it is simpler (no need for storage replication) and the flexibility...
However; our own DBA's don't like replication. Our Oracle DBA mentions replication is only useful in addition to clustering.
Is replication a valid (or even better) alternative for active-passive clustering? What do most new HA implementations use?
-- a Linux architect