DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma;
	h=mime-version:in-reply-to:references:date:message-id:subject:from:to
	:cc:content-type:content-transfer-encoding;
	b=kojk4VxvufTEMfXIgqbN2Fg2gLF0GmSKVyO6EKMdKIwo6qUCJ/XXHbfQtoKKEY6sxK
	yxlOw4oDtXZRmg1TgTDozCKH2OQxcHjR80DPyjqQq2ZiMU8PMtMQmGBm2hDNAJtp3miN
	IeaoEz6nmn1a9+KPpiQOJM4NkZBkaR8T51id0=
MIME-Version: 1.0
In-Reply-To: <20100420184727.GE53489@mr-paradox.net>
References: <20100420173936.GA50886@mr-paradox.net>
	<h2w603c8f071004201044tbc8ce8fenb846b7643a2b9376@mail.gmail.com>
	<20100420180351.GA51940@mr-paradox.net>
	<s2i603c8f071004201115xae430457q88074af464c8964f@mail.gmail.com>
	<20100420184727.GE53489@mr-paradox.net>
Date: Tue, 20 Apr 2010 13:22:36 -0600
Message-ID: <q2udcc563d11004201222se02e858fv50b78b96b9d2c015@mail.gmail.com>
Subject: Re: Very high effective_cache_size == worse performance?
From: Scott Marlowe <scott.marlowe@gmail.com>
To: David Kerr <dmk@mr-paradox.net>
Cc: Robert Haas <robertmhaas@gmail.com>, pgsql-performance@postgresql.org
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

On Tue, Apr 20, 2010 at 12:47 PM, David Kerr <dmk@mr-paradox.net> wrote:
> On Tue, Apr 20, 2010 at 02:15:19PM -0400, Robert Haas wrote:
> - On Tue, Apr 20, 2010 at 2:03 PM, David Kerr <dmk@mr-paradox.net> wrote:
> - > that thought occured to me while I was testing this. I ran a vacuumdb=
 -z
> - > on my database during the load and it didn't impact performance at al=
l.
> -
> - The window to run ANALYZE usefully is pretty short. =A0If you run it
> - before the load is complete, your stats will be wrong. =A0If you run it
> - after the select statements that hit the table are planned, the
> - updated stats won't arrive in time to do any good.
>
> right, but i'm loading 20 million records in 1000 record increments. so
> the analyze should affect all subsequent increments, no?

I keep thinking FK checks are taking a long time because they aren't
cached because in import they went through the ring buffer in pg or
some other way aren't in a buffer but large effective cache size says
it's 99.99% chance or better that it's in cache, and chooses a poor
plan to look them up.  Just a guess.