Re: vacuum / analyze parent tables on partitioned tables.

> autovacuum is still enabled, but we want to avoid autovacuum to hit right
in the day / when the etl is
> working. So that's why we want to shedule it by hand, before autovacuum
starts.

I see.

> Is it a good idea to exclude those tables then? And run only vacuum /
analyze on the child tables, an
> wait for autovacuum to analyze the parent tables if it ever would become
necesarry..

Sure, you could do that.
Maybe somebody on the list would know if autovacuum/analyze on the parent
table
will scan the children - I believe it does not.
You could also test it yourself.

Yours,
Laurenz Albe

On Wed, Jan 23, 2013 at 1:38 PM, Albe Laurenz <laurenz(dot)albe(at)wien(dot)gv(dot)at>wrote:

> Bert wrote:
> > I wrote a script to make sure all tables are vacuumed and analyzed every
> evening. This works very
> > well.
>
> Autovacuum doesn't do the job for you?
>
> That would save you from worries like the ones you have.
>
> > Does anyone has an idea why in this case the vacuum/analyze takes almost
> as long on the parent table
> > as on the biggest child table? (the other child tables are smaller than
> f_tranaction_1, and their
> > vacuum/analyze time is much shorter).
>
> ANALYZE on the parent table gathers statistics about the child
> tables as well, so it scans them too.
>
> You still need to ANALYZE and VACUUM the child tables though.
>
> Yours,
> Laurenz Albe
>

--
Bert Desmet
0477/305361