Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Subject: Re: slow query using sub select 
Date: Tue, 23 May 2006 09:26:37 -0400
Message-ID: <47668A1334CDBF46927C1A0DFEB223D340EF94@mail.optiosoftware.com>
Thread-Topic: [PERFORM] slow query using sub select 
thread-index: AcZ99H95e0LzlgA4T722/LypaaE1kwAd8SiQ
From: "Tim Jones" <TJones@optio.com>
To: "Tom Lane" <tgl@sss.pgh.pa.us>
Cc: <pgsql-performance@postgresql.org>

that worked like a champ nice call as always!=20

thanks

Tim Jones
Healthcare Project Manager
Optio Software, Inc.
(770) 576-3555

-----Original Message-----
From: Tom Lane [mailto:tgl@sss.pgh.pa.us]=20
Sent: Monday, May 22, 2006 7:07 PM
To: Tim Jones
Cc: pgsql-performance@postgresql.org
Subject: Re: [PERFORM] slow query using sub select=20

"Tim Jones" <TJones@optio.com> writes:
>   I am having a problem with a sub select query being kinda slow.  The

> query is as follows:
=20
> select batterycode, batterydescription, observationdate from Battery=20
> t1 where patientidentifier=3D611802158 and observationdate =3D (select
> max(observationdate) from Battery t2 where=20
> t2.batterycode=3Dt1.batterycode and patientidentifier=3D611802158) =
order
by batterydescription.

Yeah, this is essentially impossible for the planner to optimize,
because it doesn't see any way to de-correlate the subselect, so it does
it over again for every row.  You might find it works better if you cast
the thing as a SELECT DISTINCT ON problem (look at the "weather report"
example in the SELECT reference page).

			regards, tom lane