To: "Kevin Grittner" <Kevin.Grittner@wicourts.gov>
cc: pgsql-performance@postgresql.org
Subject: Re: Planner reluctant to start from subquery 
In-reply-to: <43E0AA7E.EE98.0025.0@wicourts.gov> 
References: <43E0AA7E.EE98.0025.0@wicourts.gov>
Comments: In-reply-to "Kevin Grittner" <Kevin.Grittner@wicourts.gov>
	message dated "Wed, 01 Feb 2006 12:33:02 -0600"
Date: Wed, 01 Feb 2006 14:34:24 -0500
Message-ID: <3759.1138822464@sss.pgh.pa.us>
From: Tom Lane <tgl@sss.pgh.pa.us>

"Kevin Grittner" <Kevin.Grittner@wicourts.gov> writes:
> We do have a few queries where PostgreSQL is several orders of
> magnitude slower.  It appears that the reason it is choosing a bad plan
> is that it is reluctant to start from a subquery when there is an outer
> join in the FROM clause.

AFAICT this case doesn't really hinge on the outer join at all.  The
problem is that EXISTS subqueries aren't well optimized.  I would have
expected an equivalent IN clause to work better.  In fact, I'm not
clear why the planner isn't finding the cheapest plan (which it does
estimate as cheapest) from the IN version you posted.  What PG version
is this exactly?

> ... The third query is the fastest, but isn't
> portable enough for our mixed environment.

Not really relevant to the problem, but what's wrong with it?  Looks
like standard SQL to me.

			regards, tom lane