From: "Dave Dutcher" <dave@tridecap.com>
To: "'Gauri Kanekar'" <meetgaurikanekar@gmail.com>
Cc: <pgsql-performance@postgresql.org>
Subject: Re: Nested Loop
Date: Mon, 26 Mar 2007 10:38:12 -0500
Message-ID: <004101c76fbc$c439b700$2e00a8c0@tridecap.com>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Thread-Index: AcdvuEcMLyIsGqmySDWh4+55xiq5bgAA39Xw
In-Reply-To: <7e4ba9550703260803l16b86ef0w42182ef762837db8@mail.gmail.com>
Importance: Normal

-----Original Message-----
>From: pgsql-performance-owner@postgresql.org On Behalf Of Gauri Kanekar
>Subject: Re: [PERFORM] Nested Loop
>
>join_collapse_limit =3D 1                 # JOINs=20

Is there a reason you have this set to 1?  Postgres can't consider =
multiple
join orders when you do that.  I would try setting that back to the =
default
and seeing if this query is any faster.

Other than that it looked like the problems with the query might be bad
estimates of rows.  One is that postgres expects there to be 1 matching =
row
from rd when there are actually 30.  You might try increasing the =
statistics
targets on rd.sd and rd.sdt, reanalyzing, and seeing if that helps.  =
Also
postgres expects the join of rd and rm to return about 27205 rows when =
it
actually returns 10 million.  I'm not sure what you can do about that.
Maybe if Postgres gets a better estimate for rd it would then estimate =
the
join better.

Dave