From owner-freebsd-scsi@FreeBSD.ORG Thu Nov 13 12:52:45 2003 Return-Path: Delivered-To: freebsd-scsi@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 612B216A4D0 for ; Thu, 13 Nov 2003 12:52:45 -0800 (PST) Received: from mailman.zeta.org.au (mailman.zeta.org.au [203.26.10.16]) by mx1.FreeBSD.org (Postfix) with ESMTP id 412DB43FA3 for ; Thu, 13 Nov 2003 12:52:42 -0800 (PST) (envelope-from bde@zeta.org.au) Received: from gamplex.bde.org (katana.zip.com.au [61.8.7.246]) by mailman.zeta.org.au (8.9.3p2/8.8.7) with ESMTP id HAA21577; Fri, 14 Nov 2003 07:52:26 +1100 Date: Fri, 14 Nov 2003 07:52:26 +1100 (EST) From: Bruce Evans X-X-Sender: bde@gamplex.bde.org To: Allen Briggs In-Reply-To: <20031113132327.GF13029@canolog.ninthwonder.com> Message-ID: <20031114074826.K3873@gamplex.bde.org> References: <20031112172306.J4572@pooker.samsco.home> <20031113084355.P13503@cvs.imp.ch> <20031113132327.GF13029@canolog.ninthwonder.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII cc: freebsd-scsi@freebsd.org cc: Martin Blapp Subject: Re: Very bad FreeBSD SCSI RAID5 write speed performance X-BeenThere: freebsd-scsi@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: SCSI subsystem List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 13 Nov 2003 20:52:45 -0000 On Thu, 13 Nov 2003, Allen Briggs wrote: > On Thu, Nov 13, 2003 at 06:38:53AM -0500, Gary Stanley wrote: > > root@64:[/tmp/blah]>dd if=/dev/zero of=/tmp/blah/blah > > 89472+0 records in > > 89471+0 records out > > 45809152 bytes transferred in 8.546312 secs (5360108 bytes/sec) > > I suspect you know this, but if you give a larger block size (bs=...), > you should see better performance. Writing to a file on the filesystem > not only brings up the softupdates question, but also your choices for > block and fragment sizes. On a RAID, there's also the question of > stripe size and how that's set with respect to the filesystem block > and fragment sizes. Writing through the file system will cause reblocking so that the small blocks are only seen by the software, especially for big files as in the above, and since CPUs are now much faster than disks there should be very little decrease in i/o speed due to using small blocks. Some CPU cycles will just become unavailable for other things. Bruce