Re: Query performance question on a large table

=?iso-8859-2?Q?Egy=FCd_Csaba?= <csegyud(at)vnet(dot)hu> writes:
> here is a sample query:
> select mertido, fomeazon, ertektipus, mertertek from t_me30 where fomeazon
> in (select distinct fomeazon from t_fome where lower(inuse) = 'igen') and
> mertido like '2003-12-17%' and ertektipus in ('+MW') order by mertido,
> fomeazon, ertektipus;

> Ohh, I nearly forgot the config: Linux 7.1; Postgres 7.3.2;

The first thing you ought to do is move to PG 7.4. "foo IN (SELECT ...)"
generally works a lot better under 7.4 than prior releases. I'd suggest
dropping the "DISTINCT" when using 7.4, too.

regards, tom lane