From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Graeme Hinchliffe <graeme(dot)hinchliffe(at)zeninternet(dot)co(dot)uk> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: Massive slowdown when LIMIT applied |
Date: | 2006-08-16 14:23:01 |
Message-ID: | 19785.1155738181@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Graeme Hinchliffe <graeme(dot)hinchliffe(at)zeninternet(dot)co(dot)uk> writes:
> If I do the following:
> SELECT username,acctstarttime FROM radacct WHERE username='user'
> ORDER BY acctstarttime;
> again the results come up slightly slower as would be expected but
> still very quickly.. (a few hundred rows in this case)
> Now, if I do the following:
> SELECT username,acctstarttime FROM radacct WHERE username='user'
> ORDER BY acctstarttime LIMIT 50;
> I have to wait a very long time, IE instead of 2-3 seconds.. can be
> as slow as 10+ minutes.
Have you compared EXPLAIN ANALYZE outputs for the two cases? Presumably
the planner is picking a different plan for the LIMIT case (and guessing
wrong :-()
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | John R. Allgood | 2006-08-16 14:24:35 | Determine Transactions Per Second |
Previous Message | Alvaro Herrera | 2006-08-16 14:22:13 | Re: Timezones -- what comes out does not go in? |