Message-ID: <50512.63.227.74.41.1097295204.squirrel@63.227.74.41>
In-Reply-To: <23596.1097293763@sss.pgh.pa.us>
References: <51319.63.227.74.41.1097293075.squirrel@63.227.74.41>
	<23596.1097293763@sss.pgh.pa.us>
Date: Fri, 8 Oct 2004 23:13:02 -0500 (CDT)
Subject: Re: Impact of foreign keys on a simple count(*) ?
From: "C. Bensend" <benny@bennyvision.com>
To: pgsql-sql@postgresql.org
User-Agent: SquirrelMail/1.4.3a
MIME-Version: 1.0
Content-Type: text/plain;charset=iso-8859-1
Content-Transfer-Encoding: 8bit
Importance: Normal


> Foreign keys have zero, nada, zilch to do with the performance of
> count(*).

OK, I just wanted to make sure.

> The only plausible theory I can think of for the performance
> difference is that in your "new" database the table has been through
> several mass updates, leading to a whole lot of dead rows and a much
> larger physical table size to scan through.  I'd suggest a VACUUM
> VERBOSE on both old and new copies of the table to get an idea of the
> relative physical sizes.  You might need a VACUUM FULL to get the new
> table back down to a reasonable size...

Oh good lord.

*waves the Idiot Flag as he slinks back under his rock*

Thank you, Tom, for your patience.  That was it.  I'm a potatohead,
but I'm learning.  :)

Benny, swearing he DID do a vacuum, but obviously not...


-- 
"Even if a man chops off your hand with a sword, you still have two nice,
sharp bones to stick in his eyes."
                                                      -- .sig on Slashdot