Le 28 août 2017 6:52 AM, "Don Seiler" <don(at)seiler(dot)us> a écrit :
So, that's kind of simple:
* first replaying archived WAL
* when no more archived WAL, streaming
* when error, back to WAL replay
After this last scenario, will the standby periodically retry to connect to
the primary to resume streaming replication? It will it at least try it it
runs out of WAL files?
It will resume streaming when it runs out of WAL files to replay.