Date: Fri, 20 Oct 2006 14:18:57 +0200
From: Tobias Brox <tobias@nordicbet.com>
To: "Jim C. Nasby" <jim@nasby.net>
Cc: Tobias Brox <tobias@nordicbet.com>,
	pgsql-performance@postgresql.org
Subject: Re: Swappiness setting on a linux pg server
Message-ID: <20061020121857.GA11885@oppetid.no>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Organization: Group Nordicbet
User-Agent: Mutt/1.5.9i

[Jim C. Nasby - Thu at 11:31:26AM -0500]
> The issue with pg_xlog is you don't need bandwidth... you need super-low
> latency. The best way to accomplish that is to get a battery-backed RAID
> controller that you can enable write caching on. In fact, if the
> controller is good enough, you can theoretically get away with just
> building one big RAID10 and letting the controller provide the
> low-latency fsyncs that pg_xlog depends on.

I was talking a bit about our system administrator.  We're running 4
disks in raid 1+0 for the database and 2 disks in raid 1 for the WALs
and for OS.  He wasn't really sure if we had write cacheing on the RAID
controller or not.  He pasted me some lines from the dmesg:

 sda: asking for cache data failed
 sda: assuming drive cache: write through
 failed line is expected from these controllers
 0000:02:0e.0 RAID bus controller: Dell PowerEdge Expandable RAID controller 4 (rev 06)

I think we're going to move system logs, temporary files and backup
files from the wal-disks to the db-disks.  Since our database aren't on
fire for the moment, I suppose we'll wait moving the rest of the OS :-)