Poul Kristensen <bcc5226(at)gmail(dot)com> writes:
> Replica(hot_standby) server:
> Is it normal behavior that PostgreSQL 9.6 does a very very slow
> recovery restart point at 1/77472E5 in the pg_log
> while doing this
> postgres: startup process recovering 000000010000000100000077(this
> process has run several days)
If that's the standby server, that process will persist as long as it's in
standby; it's what's responsible for applying WAL received by the
walreceiver process. You should have been able to connect long before
now, though.
regards, tom lane