From owner-freebsd-hackers Fri Jan 29 04:58:33 1999 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id EAA12975 for freebsd-hackers-outgoing; Fri, 29 Jan 1999 04:58:33 -0800 (PST) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from ns1.yes.no (ns1.yes.no [195.204.136.10]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id EAA12954 for ; Fri, 29 Jan 1999 04:58:29 -0800 (PST) (envelope-from eivind@bitbox.follo.net) Received: from bitbox.follo.net (bitbox.follo.net [195.204.143.218]) by ns1.yes.no (8.9.1a/8.9.1) with ESMTP id NAA13648 for ; Fri, 29 Jan 1999 13:58:27 +0100 (CET) Received: (from eivind@localhost) by bitbox.follo.net (8.8.8/8.8.6) id NAA35584 for hackers@freebsd.org; Fri, 29 Jan 1999 13:58:26 +0100 (MET) Date: Fri, 29 Jan 1999 13:58:26 +0100 From: Eivind Eklund To: hackers@FreeBSD.ORG Subject: Re: DPT wobbles Message-ID: <19990129135825.E34859@bitbox.follo.net> References: <19990128162102.A10431@emma.eng.uct.ac.za> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.95.1i In-Reply-To: <19990128162102.A10431@emma.eng.uct.ac.za>; from Shaun Courtney on Thu, Jan 28, 1999 at 04:21:03PM +0200 Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Thu, Jan 28, 1999 at 04:21:03PM +0200, Shaun Courtney wrote: > Hi > > We are using 2.2.8-Release with an Asus p2B motherboard and a dpt 2144uw > raid controller. Connected on the scsi chain are 3 ibm 4Gig, 50 pin > drives that form the raid array and an hp dat drive. After days of > operation the dpt controller starts beeping indicating a drive failure, > if we rebuild the array without replacing the "failed" drive it > continues to work for another couple of days and then starts beeping > again - often it is an entirely different drive that has failed. We have > replaced all three drives and have put the machine onto a UPS, but still > the problem occurs. Can anyone help? Check your cabling. Intermittent failures usually comes from marginal cabling; the DPT drives the drives very hard. Another possibility is vibration reinforcement between the drives, but 3 drives sounds like awfully few for this problem, and usually the drives themselves will die. You can solve this type of problem by not using metal casing. I think a problem like yours would not occur if you ran with ram-to-drive ECC; have you set up your disks with 528 byte sectors, and equipped the DPT with ECC RAM from DPT? This gives you an extremely good error path, as you have error correction *everywhere*. Of course, with this you'd likely just have lost speed instead of finding your bad cabling (unless you read your logs carefully), which in some ways are worse... Eivind. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message