Re: SQL Function Performance

From: Michael Fuhr <mike(at)fuhr(dot)org>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: Adnan DURSUN <a_dursun(at)hotmail(dot)com>, pgsql-performance(at)postgresql(dot)org
Subject: Re: SQL Function Performance
Date: 2006-02-14 01:31:54
Message-ID: 20060214013154.GA74883@winnie.fuhr.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

On Mon, Feb 13, 2006 at 07:57:07PM -0500, Tom Lane wrote:
> "Adnan DURSUN" <a_dursun(at)hotmail(dot)com> writes:
> >>>> EXPLAIN ANALYZE EXECUTE stmt (...);
>
> > Here is the EXPLAIN ANALYZE output for prepared statement :
>
> This is exactly the same as the other plan --- you did not parameterize
> the query. To see what's going on, you need to insert PREPARE
> parameters in the places where the function uses plpgsql variables.

Actually it was an SQL function, but that also does PREPARE/EXECUTE,
right?

Adnan, what Tom is saying is that I requested this (simplified):

PREPARE stmt (integer) AS SELECT * FROM foo WHERE id = $1;
EXPLAIN ANALYZE EXECUTE stmt (12345);

but instead you appear to have done this:

PREPARE stmt AS SELECT * FROM foo WHERE id = 12345;
EXPLAIN ANALYZE EXECUTE stmt;

We can tell because if you had done it the first way (parameterized)
then the EXPLAIN ANALYZE output would have shown the parameters as
$1, $2, $3, etc., which it didn't.

--
Michael Fuhr

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Chris 2006-02-14 02:32:32 Re: Optimizing performance of a like '%...%' condition
Previous Message Tom Lane 2006-02-14 00:57:07 Re: SQL Function Performance