Message-ID: <985a2672b882183d323bbba37e6357ae.squirrel@sq.gransy.com>
In-Reply-To: <A0B9339E-9D58-4842-A206-050B773360B6@darkstatic.com>
References: <DF2D5436-117C-4D02-9B3C-A55723B7DDE1@darkstatic.com>
	<20110412171855.GA14292@tux>
	<FC3A3A2B-3ECB-41BA-8F94-356D6FED3695@darkstatic.com>
	<4DA496FA.3070908@fuzzy.cz>
	<8F22D592-23C1-4A3C-94A5-48363332ADD3@darkstatic.com>
	<4DA4BFA2.5060601@fuzzy.cz>
	<B93DFBBB-DA56-4044-A508-0B7E4A2CFD28@darkstatic.com>
	<4DA4D40A.4010200@fuzzy.cz>
	<A0B9339E-9D58-4842-A206-050B773360B6@darkstatic.com>
Date: Wed, 13 Apr 2011 16:14:42 +0200
Subject: Re: Performance
From: tv@fuzzy.cz
To: "Ogden" <lists@darkstatic.com>
Cc: "Tomas Vondra" <tv@fuzzy.cz>,
 pgsql-performance@postgresql.org
User-Agent: SquirrelMail/1.4.21
MIME-Version: 1.0
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: 8bit
Importance: Normal

> Thomas,
>
> Thank you for your very detailed and well written description. In
> conclusion, I should keep my random_page_cost (3.0) to a value more than
> seq_page_cost (1.0)? Is this bad practice or will this suffice for my
> setup (where the database is much bigger than the RAM in the system)? Or
> is this not what you are suggesting at all?

Yes, keep it that way. The fact that 'random_page_cost >= seq_page_cost'
generally means that random reads are more expensive than sequential
reads. The actual values are  dependent but 4:1 is usually OK, unless your
db fits into memory etc.

The decrease of performance after descreasing random_page_cost to 3 due to
changes of some execution plans (the index scan becomes slightly less
expensive than seq scan), but in your case it's a false assumption. So
keep it at 4 (you may even try to increase it, just to see if that
improves the performance).

regards
Tomas