For zero data loss protection, there must be synchronous communication between your primary database and the Far Sync instance. The Far Sync instance receives redo from the primary in a synchronous manner and forwards it immediately to all the standby databases in an asynchronous manner. This setup also reduces the overhead on the primary database, because it only has to send the redo to the Far Sync instance rather than all the standby databases. If a Far Sync instance fails, Data Guard automatically uses asynchronous transport to the secondary database from the primary database to maintain near-zero data loss protection. For added resiliency, customers may deploy multiple Far Sync instances per each database instance (primary and secondaries).
0 commit comments