I would like to bring up again that it would be nice if vacuum/analyze recorded the last action and stats in a table
(pg_vacuum) within each database. Then dba's can make better decisions how vacuum/analyze is working as well as
tuning the fsm guc's. This should be a real easy additions to the system. While I am wishing, why not change explain
analyze to put out the last analyze timestamp, this should tell people if the stats are old or not, I have to believe
that this info could also help the autovac process.
Jim
---------- Original Message -----------
From: Robert Treat <xzilla@users.sourceforge.net>
To: pgsql-general@postgresql.org
Cc: Bruce Momjian <pgman@candle.pha.pa.us>, "Matthew T. O'Connor" <matthew@zeut.net>, Will Reese <wreese@rackspace.com>
Sent: Fri, 28 Apr 2006 15:15:05 -0400
Subject: Re: [GENERAL] Autovacuum Logging
> On Thursday 27 April 2006 11:59, Bruce Momjian wrote:
> > Well, one problem now is that everytime pg_autovacuum opens a database,
> > a message is placed in the logs:
> >
> > LOG: autovacuum: processing database "test"
> > LOG: autovacuum: processing database "test"
> > LOG: autovacuum: processing database "test"
> > LOG: autovacuum: processing database "test"
> >
> > This is certainly not desirable. I changed the message to DEBUG1 for
> > 8.2 with the attached patch.
>
> IMHO this is less optimal, since the default installation will provide no way
> for users to verify that autovacuum is running. Currently we can direct them
> to look in thier log output, this way we would have to tell them to crank up
> logging to DEBUG1 and then look, which will of course contain a whole lot of
> other chatter at that level.
>
> The best solution I have seen so far is the idea of a GUC to control
> autovacuums chatter level. I hope we get that becuase the above will be a
> regression imho.
>
> --
> Robert Treat
> Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL
>
> ---------------------------(end of broadcast)---------------------------
> TIP 6: explain analyze is your friend
------- End of Original Message -------
