Heikki,
Yes, I did restart postmaster after setting autovacuum=off
as I've mentioned before I've checked it using show autovacuum.
It was vacuuming a table from the old database while
the server was running pg_dump -U postgres old_base | psql new_base postgres
Sorry for missing the CC, I had a bad night sleep (actually dump-restored db =)
On Fri, 04 May 2007 11:57:25 +0100
Heikki Linnakangas <heikki@enterprisedb.com> wrote:
> Please CC the list to allow others to participate in the discussion, and
> to have it archived for future readers...
>
> Evgeny Gridasov wrote:
> > I've seen it analyzing and vacuuming some tables for a long time...
> > Is it OK?
> > The reason I'm asking I was dump-restoring a huge database, and disabled
> > autovacuum in config file.
>
> Well, to avoid XID wrap-around, you have to run (auto)vacuum eventually.
> If you don't mind that your clog files will take up more disk space, as
> you probably don't if you have a huge database anyways, you can increase
> autovacuum_freeze_max_age so you don't need to run it as often.
>
> But I'm starting to wonder if it was the XID wraparound that triggered
> the autovacuum after all. You did restart postmaster after setting
> autovacuum=off in the config file, right? That's required for the
> setting to take effect.
>
> What table is it vacuuming? A recently restored table shouldn't need to
> be vacuumed for a long time.
>
> --
> Heikki Linnakangas
> EnterpriseDB http://www.enterprisedb.com
>
