Re: Broken hint bits (freeze)

Vladimir Borodin <root(at)simply(dot)name> writes:

> > 6 июня 2017 г., в 23:30, Sergey Burladyan <eshkinkot(at)gmail(dot)com> написал(а):
> >
> > Dmitriy Sarafannikov <dsarafannikov(at)yandex(dot)ru> writes:
> >
> >> Starting and stopping master after running pg_upgrade but before rsync to collect statistics
> >> was a bad idea.
> >
> > But, starting and stopping master after running pg_upgrade is *required*
> > by documentation:
> > https://www.postgresql.org/docs/9.6/static/pgupgrade.html
> >> f. Start and stop the new master cluster
> >> In the new master cluster, change wal_level to replica in the postgresql.conf file and then start and stop the cluster.
> >
> > and there is no any suggestion to disable autovacuum for it.

> Yep. This should probably be fixed in the documentation?

I think so. There is some problem in pg_upgrade documentation, nothing about:
1. preventing heap change by vacuum, analyze, something also when master
restarted after pg_upgrade but before rsync
2. log-shipping only standby cannot shutdown at the same checkpoint with
master

I try to start discuss about this: https://www.postgresql.org/message-id/87y3ta49zp.fsf%40seb.koffice.internal
but without luck :-)

PS: I CC'd Bruce here.

--
Sergey Burladyan