>>>
>>> But try comparing pg_current_xlog_location on the master to pg_last_xlog_receive_location on the replica. You could
alsolook into using the pg_stat_replication view.
>>>
>>
>>
>> This becomes challenging after the master is shutdown...
>
> Right. Thus the need for the sequence I mentioned: shut down all services accessing the database, check the status,
shutdown the master...
I could be missing something obvious, but how does this stop e.g. checkpoints, autovacuum etc from occuring and
generatingnew wal that the standby might like? It seems like stopping services accessing the DB is likely to cause a
lengthyautovacuum to begin, for example.
Graeme Bell
