| From: | Martijn van Oosterhout <kleptog(at)svana(dot)org> |
|---|---|
| To: | Kolus Maximiliano <Kolus(dot)maximiliano(at)bcr(dot)com(dot)ar> |
| Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, pgsql-general(at)postgresql(dot)org |
| Subject: | Re: problem with huge joins |
| Date: | 2003-11-01 00:43:21 |
| Message-ID: | 20031101004321.GD5475@svana.org |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
On Fri, Oct 31, 2003 at 01:17:45PM -0300, Kolus Maximiliano wrote:
> > > Total runtime: 317046.69 msec
>
> Total runtime: 216001.94 msec
>
> A lot better! Thanks!
>
> > The hash indexes are a waste of time for this :-(
>
> Which kind should I use?
My guess that for large joins it can be more useful to use indexes to
presort a table and use highly efficient merge joins instead. However, hash
indexes cannot be used to sort a table, only for lookups.
btrees are more highly optimised and can produce sorted output.
Give it a shot.
--
Martijn van Oosterhout <kleptog(at)svana(dot)org> http://svana.org/kleptog/
> "All that is needed for the forces of evil to triumph is for enough good
> men to do nothing." - Edmond Burke
> "The penalty good people pay for not being interested in politics is to be
> governed by people worse than themselves." - Plato
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Ryan Mack | 2003-11-01 01:36:21 | PostgreSQL License Question |
| Previous Message | Martijn van Oosterhout | 2003-11-01 00:31:01 | Re: slow query performance |