Message-ID: <4DA26A30.6010208@2ndQuadrant.com>
Date: Sun, 10 Apr 2011 22:40:48 -0400
From: Greg Smith <greg@2ndQuadrant.com>
User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US;
	rv:1.9.1.16) Gecko/20110307 Icedove/3.0.11
MIME-Version: 1.0
To: pgsql-performance@postgresql.org
Subject: Re: optimizer parameters
References: <77760.3733.qm@web38606.mail.mud.yahoo.com>
	<4DA23CFE.7020109@fuzzy.cz>
In-Reply-To: <4DA23CFE.7020109@fuzzy.cz>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

On 04/10/2011 07:27 PM, Tomas Vondra wrote:
> But if you need more details, then I quess the best approach to get it
> is to read the sources (search for the cost estimation etc.).
>    

There's a small fully worked out example of this in my book too, where I 
duplicate the optimizer's EXPLAIN cost computations for a simple query.  
The main subtle thing most people don't appreciate fully is how much the 
optimizer takes into account two things:  the selectivity of operators, 
and the expected size of the data being moved around.  For example, a 
lot of the confusion around "why didn't it use my index?" comes from not 
noticing the size of the index involved, and therefore how expensive its 
page cost is considered to be.

-- 
Greg Smith   2ndQuadrant US    greg@2ndQuadrant.com   Baltimore, MD
PostgreSQL Training, Services, and 24x7 Support  www.2ndQuadrant.us
"PostgreSQL 9.0 High Performance": http://www.2ndQuadrant.com/books