From owner-freebsd-stable Sat Aug 31 3:15:37 2002 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 8ECE337B400 for ; Sat, 31 Aug 2002 03:15:34 -0700 (PDT) Received: from mail.allcaps.org (h-66-166-142-198.SNDACAGL.covad.net [66.166.142.198]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1E38643E6A for ; Sat, 31 Aug 2002 03:15:34 -0700 (PDT) (envelope-from bsder@mail.allcaps.org) Received: by mail.allcaps.org (Postfix, from userid 501) id 55AEA154EF; Sat, 31 Aug 2002 03:15:33 -0700 (PDT) Received: from localhost (localhost [127.0.0.1]) by mail.allcaps.org (Postfix) with ESMTP id 4BBC4154EE; Sat, 31 Aug 2002 03:15:33 -0700 (PDT) Date: Sat, 31 Aug 2002 03:15:33 -0700 (PDT) From: "Andrew P. Lentvorski" To: BugsGrief@bugsgrief.net Cc: freebsd-stable@freebsd.org Subject: Re: ata problem(s) In-Reply-To: <200208310930.g7V9UcK57531@ogyo.bugsgrief.net> Message-ID: <20020831025133.O55708-100000@mail.allcaps.org> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG You have symptoms similar to what I had. After digging on the web and wading through IBM tech support, my hypothesis was that these symptoms were caused by a thermal recalibration cycle. However, since this bug was intermittent, I never managed to get conclusive proof. This thermal recalibration seems to be the reason why IBM specified that certain drives were only expected to be on about 10 hours per day (it does a thermal recalibrate on power up and then doesn't need to do one again for a while). I had 3 40GB IBM disks which had exactly the same symptoms. Normally the disk makes a pretty good racket (clicks and clunks) just before FreeBSD throws the errors. If my hypothesis is correct, there really isn't any way around this unless someone can suggest a kernel variable tweak that allows more time for disk I/O writes to complete or increases the number of retries before resetting the ATA bus so that the drive can come out of calibration and then complete the transaction. If the drives are still under warranty, harass IBM. Unfortunately, the drive is going to show up as a good drive under their drive tester software (thermal recalibration is very intermittent); so they'll give you grief. I managed to get them to swap all of my drives for new ones, but I just use them as spares now. I dumped my IBM drives from my RAID box and replaced them with another brand (Maxtor, I think). I haven't had any problems since. -a To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message