From owner-freebsd-stable@FreeBSD.ORG Fri Jan 16 16:02:55 2004 Return-Path: Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id E5D0216A4CE for ; Fri, 16 Jan 2004 16:02:55 -0800 (PST) Received: from carver.gumbysoft.com (carver.gumbysoft.com [66.220.23.50]) by mx1.FreeBSD.org (Postfix) with ESMTP id DBE7743D31 for ; Fri, 16 Jan 2004 16:02:54 -0800 (PST) (envelope-from dwhite@gumbysoft.com) Received: by carver.gumbysoft.com (Postfix, from userid 1000) id CE91672DBF; Fri, 16 Jan 2004 16:02:54 -0800 (PST) Received: from localhost (localhost [127.0.0.1]) by carver.gumbysoft.com (Postfix) with ESMTP id CC79372DB5; Fri, 16 Jan 2004 16:02:54 -0800 (PST) Date: Fri, 16 Jan 2004 16:02:54 -0800 (PST) From: Doug White To: Tom Glover In-Reply-To: Message-ID: <20040116155912.H93165@carver.gumbysoft.com> References: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII cc: stable@freebsd.org Subject: Re: Promise controller and 5.2 X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 17 Jan 2004 00:02:56 -0000 On Fri, 16 Jan 2004, Tom Glover wrote: > > I am running 5.2-RELEASE with a Promise S150 TX4 Serial ATA RAID > controller and a couple of WDC SATA drives at RAID 1. I am noticing some > behavior that does not seem right to me. > > Systems boots fine on the two mirrored drives. If I disconnect the drive > on channel 2 then the system continues to run. However if I disconnect the > drive on channel 1 it dies. Can you explain "dies"? > If I reboot after the drive on channel 1 fails I can recover and boot off > the drive on channel 2. However, the Promise S150 BIOS throws up an error > message about the drive failure that requires manual intervention before > boot can proceed. This is normal for Promise and most ATA RAID, since thats likely to be the only warning you'll get about a missing disk. Also, some operating systems are not good about bumping the generation numbers AT ALL, and you can end up with unsynchronized mirrors if you continue. With the mirror, you won't lose your data, but its not designed to be fault-tolerant in a runtime sense. I don't know why you get differing failure modes between disks, though -- perhaps you aren't using the right disk device? You should be using ar0. > There does not appear to be any way around that. This, IMIHO, is brain > dead and makes the controller useless for any remote operation. If this > is SOP for Promise then can someone suggest a SATA RAID controller that > behaves in a more sane manner? You've discovered that ATA RAID is software RAID with hardware assist. If you want more transparent ATA RAID, you want 3ware. -- Doug White | FreeBSD: The Power to Serve dwhite@gumbysoft.com | www.FreeBSD.org