Date: Thu, 8 Feb 2001 12:08:31 +0000 From: Geoff Buckingham <geoffb@chuggalug.clues.com> To: Dan Phoenix <dphoenix@bravenet.com> Cc: Julian Elischer <julian@elischer.org>, Matt Dillon <dillon@earth.backplane.com>, Andrew Reilly <areilly@bigpond.net.au>, Alfred Perlstein <bright@wintelcom.net>, Andre Oppermann <oppermann@monzoon.net>, Rik van Riel <riel@conectiva.com.br>, Mike Silbersack <silby@silby.com>, Poul-Henning Kamp <phk@critter.freebsd.dk>, Charles Randall <crandall@matchlogic.com>, Jos Backus <josb@cncdsl.com>, freebsd-hackers@FreeBSD.ORG Subject: Re: vinum and qmail (RE: qmail IO problems) Message-ID: <20010208120831.C61928@chuggalug.clues.com> In-Reply-To: <Pine.BSO.4.21.0102080336030.3495-100000@gandalf.bravenet.com>; from Dan Phoenix on Thu, Feb 08, 2001 at 03:41:59AM -0800 References: <3A826301.8AC3771C@elischer.org> <Pine.BSO.4.21.0102080336030.3495-100000@gandalf.bravenet.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Feb 08, 2001 at 03:41:59AM -0800, Dan Phoenix wrote: > > > > Yes I did and it made some real differences. I enabled it on /usr as well > as /var and mounted /var with the noatime option. Doing not bad for the > amount of email it is pushing....thing is this I/O problem never use to be > an issue....but with growth constantly happening it has come to a hardware > based solution. What I have recommended to the company is a scsi card in > that machine with 2 scsi drives.....I will raid 0 then together with ccd > or venim and mount it as /var....turn existing /var into extra swap > space.....although i may thing of something else...as I don;t think it > needs a gig of swap... and that should fix the I/O issue incredibly. > Right now I have split up the load also between2 machines so that has > helped out incredibly....but systat -vmstat is still always showing 100% > disk usage so I will have to remedy the problem. Then I plan on moving all > mail back to that one machine and beating the shit right out of that > freebsd machine to see what freebsd can really handle. If anyone has some > nice newbie docs :) on ccd or venim would be greatly appreciated. > > A word of warning on this, when striping on an even number of drives using a power of 2 as as the stripe size, it is very easy to concentrate meta data on one drive, thereby doing away with much of your performance gain. Workarounds include striping at cluster size (16 or 32MB usually) using and odd number of disks (for non raid 3/5), though or experimentation. There should be a number of mails on the subject in the archives of the scsi mailing list, look for myself or greg lehay to find the thread. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010208120831.C61928>