From: Josh Berkus <josh@agliodbs.com>
Reply-To: josh@agliodbs.com
Organization: Aglio Database Solutions
To: pgsql-performance@postgresql.org
Subject: Re: Feature Request --- was: PostgreSQL Performance Tuning
Date: Fri, 27 Apr 2007 12:11:11 -0700
User-Agent: KMail/1.8
Cc: Bill Moran <wmoran@collaborativefusion.com>,
	Dan Harris <fbsd@drivefaster.net>
References: 
 <OFC96167D5.6BB63AC2-ON652572C9.003B5263-652572C9.003E4142@lntinfotech.com>
	<46323744.5020703@drivefaster.net>
	<20070427141127.cdac56fb.wmoran@collaborativefusion.com>
In-Reply-To: <20070427141127.cdac56fb.wmoran@collaborativefusion.com>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <200704271211.12439.josh@agliodbs.com>

Bill,

> The only one that seems practical (to me) is random_page_cost.  The
> others are all configuration options that I (as a DBA) want to be able
> to decide for myself. 

Actually, random_page_cost *should* be a constant "4.0" or "3.5", which 
represents the approximate ratio of seek/scan speed which has been 
relatively constant across 6 years of HDD technology.  The only reason we 
make it a configuration variable is that there's defects in our cost model 
which cause users to want to tinker with it.

Mind you, that's gotten better in recent versions as well.  Lately I mostly 
tinker with effective_cache_size and the various cpu_* stats rather than 
modifying random_page_cost.

-- 
--Josh

Josh Berkus
PostgreSQL @ Sun
San Francisco