Re: random_page_costs - are defaults of 4.0 realistic for SCSI RAID 1

From: Greg Smith <gsmith(at)gregsmith(dot)com>
To: pgsql-performance(at)postgresql(dot)org
Subject: Re: random_page_costs - are defaults of 4.0 realistic for SCSI RAID 1
Date: 2007-09-12 00:10:50
Message-ID: Pine.GSO.4.64.0709112005300.28383@westnet.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

On Wed, 12 Sep 2007, Gregory Stark wrote:

> Also, it helps to run a "vmstat 1" in another window and watch the bi and bo
> columns.

Recently on Linux systems I've been using dstat (
http://dag.wieers.com/home-made/dstat/ ) for live monitoring in this sort
of situation. Once you get the command line parameters right, you can get
data for each of the major disks on your system that keep the columns
human readable (like switching from KB/s to MB/s as appropriate) as
activity goes up and down combined with the standard vmstat data.

I still use vmstat/iostat if I want to archive or parse the data, but if
I'm watching it I always use dstat now.

--
* Greg Smith gsmith(at)gregsmith(dot)com http://www.gregsmith.com Baltimore, MD

In response to

Responses

Browse pgsql-performance by date

  From Date Subject
Next Message Joshua D. Drake 2007-09-12 00:46:09 Re: random_page_costs - are defaults of 4.0 realistic for SCSI RAID 1
Previous Message david 2007-09-11 23:51:40 Re: SAN vs Internal Disks