On Mon, 2006-10-30 at 22:58 +0100, Andreas Kostyrka wrote:
> Am Montag, den 30.10.2006, 08:18 +0000 schrieb Simon Riggs:
> > On Sun, 2006-10-29 at 00:28 +0200, Andreas Kostyrka wrote:
> > > Any ideas how to make the optimizer handle partitioned tables more
> > > sensible?
> > Yes, those are known inefficiencies in the current implementation which
> > we expect to address for 8.3.
> Any ideas to force the current optimizer to do something sensible?
Brute force & ignorance: PL/pgSQL
Perhaps some other domain knowledge might help you shorten the search?
Thats all for now. It's not a minor fixup and nobody had time to fix
that for 8.2 since other fish were bigger.
In response to
pgsql-performance by date
|Next:||From: Leif Mortenson||Date: 2006-10-31 04:04:12|
|Subject: Index ignored on column containing mostly 0 values|
|Previous:||From: Andreas Kostyrka||Date: 2006-10-30 21:58:15|
|Subject: Re: partitioned table performance|