Re: Docs about buffers and sortmem setting - Mailing list pgsql-performance

From Andrew Sullivan
Subject Re: Docs about buffers and sortmem setting
Date
Msg-id 20021114130119.G9625@mail.libertyrms.com
Whole thread Raw
In response to Re: Docs about buffers and sortmem setting  (Neil Conway <neilc@samurai.com>)
List pgsql-performance
On Thu, Nov 14, 2002 at 12:20:49PM -0500, Neil Conway wrote:
> Well, part of the reason is that a lot of the data in shared_buffers
> has to be effectively duplicated in the kernel's I/O caches, because
> it's frequently accessed. So while I'd think the cost of fetching a
> page from the buffer pool is lower than from the OS' cache, increasing
> the size of the Postgres buffer pool effectively decreases the total
> amount of RAM available for caching.

Well, yes, but on a machine with 16 G and a data set < 16 G, that's
not the issue.  A 1G shared buffer is too big anyway, according to
our experience: it's fast at the beginning, but performance degrades.
I don't know why.

A

--
----
Andrew Sullivan                         204-4141 Yonge Street
Liberty RMS                           Toronto, Ontario Canada
<andrew@libertyrms.info>                              M2P 2A8
                                         +1 416 646 3304 x110


pgsql-performance by date:

Previous
From: pginfo
Date:
Subject: Sort time
Next
From: "Henrik Steffen"
Date:
Subject: Re: [GENERAL] Upgrade to dual processor machine?