From: | "A(dot) Kretschmer" <andreas(dot)kretschmer(at)schollglas(dot)com> |
---|---|
To: | pgsql-performance(at)postgresql(dot)org |
Subject: | Re: Query timing increased from 3s to 55s when used as a function instead of select |
Date: | 2010-05-25 10:30:26 |
Message-ID: | 20100525103026.GA25341@a-kretschmer.de |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-performance |
In response to Tyler Hildebrandt :
> We're using a function that when run as a select statement outside of the
> function takes roughly 1.5s to complete whereas running an identical
> query within a function is taking around 55s to complete.
>
> select * from fn_medirota_get_staff_leave_summary(6);
> Time: 57375.477 ms
I think, your problem is here:
SELECT INTO current_user * FROM
fn_medirota_validate_rota_master(in_currentuser);
The planner has no knowledge about how many rows this functions returns
if he don't know the actual parameter. Because of this, this query
enforce a seq-scan. Try to rewrite that to something like:
execute 'select * from fn_medirota_validate_rota_master(' ||
in_currentuser' || ')' into current_user
*untested*
HTH, Andreas
--
Andreas Kretschmer
Kontakt: Heynitz: 035242/47150, D1: 0160/7141639 (mehr: -> Header)
GnuPG: 0x31720C99, 1006 CCB4 A326 1D42 6431 2EB0 389D 1DC2 3172 0C99
From | Date | Subject | |
---|---|---|---|
Next Message | Andres Freund | 2010-05-25 10:41:58 | Re: performance of temporary vs. regular tables |
Previous Message | Tyler Hildebrandt | 2010-05-25 09:59:43 | Query timing increased from 3s to 55s when used as a function instead of select |