| From: | Stephan Szabo <sszabo(at)megazone23(dot)bigpanda(dot)com> |
|---|---|
| To: | Patrick Hatcher <PHatcher(at)macys(dot)com> |
| Cc: | <sfpug(at)postgresql(dot)org>, <sfpug-owner(at)postgresql(dot)org> |
| Subject: | Re: Patrick's Evil Query |
| Date: | 2003-01-21 00:17:02 |
| Message-ID: | 20030120161414.D70744-100000@megazone23.bigpanda.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | sfpug |
On Mon, 20 Jan 2003, Patrick Hatcher wrote:
>
> I thought I would ask this question here before sending to the Performance
> list:
> So I run my Evil query using a condition of gmmid = 6. It takes about 90
> secs to run. However, if I run this and add a LIMIT 100 clause, the query
> runs in less than 5 secs.
Well, some not insignificant portion of that time is being spent making
the complicated select list at the top level. That part would certainly
be sped up if there were less rows to do it for.
Also, does it give the same plan? It is possible that it might choose a
different plan if it only wants 100 rows out.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Josh Berkus | 2003-01-21 16:16:08 | Re: Reporting: Oracle-like features? |
| Previous Message | Patrick Hatcher | 2003-01-20 23:55:19 | Re: Patrick's Evil Query |