From owner-freebsd-hackers Mon Nov 23 16:13:59 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id QAA08604 for freebsd-hackers-outgoing; Mon, 23 Nov 1998 16:13:59 -0800 (PST) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from wolf.com (ns1.wolf.com [207.137.58.20]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id QAA08599 for ; Mon, 23 Nov 1998 16:13:57 -0800 (PST) (envelope-from dan@wolf.com) Received: (qmail 11317 invoked by uid 100); 24 Nov 1998 00:22:58 -0000 Message-ID: <19981123162257.A8896@wolf.com> Date: Mon, 23 Nov 1998 16:22:57 -0800 From: Dan Mahoney To: Florin Nicolescu , FreeBSD-hackers Subject: Re: sever ide hdd crash References: <199811232346.BAA00552@nick.ro> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.93i In-Reply-To: <199811232346.BAA00552@nick.ro>; from Florin Nicolescu on Tue, Nov 24, 1998 at 01:46:27AM +0200 Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG I normally hesitate to recommend such low-level steps, but my next recommendation would be to get your hands on a good digital voltmeter and look at the output of the PC power supply. If you have an unused disk-drive power connector, monitor the voltages there while the machine is running - preferably, while it is performing the same type of operation that has been leading up to system failures. Look at the DC voltage levels, and look also at any AC component that might be present. If you have tried swapping out everything except the case and the CPU, then that might be the problem - since the power supply stays with the case. BUT BE CAREFUL! Doing anything with the system running is risky, and messing around with the power connectors while power is applied makes it very easy to toast your system if you jumper the wrong points together. > I posted about two weeks ago a possible kernel bug regarding the hdd crush I have > experienced. I couldn't continue this discussion earlier, because I couldn't make my > computer running until now. > All the answers suggested something like > >What you have there sounds like a plain, old-fashioned, simple, hard disk failure. Not > >space aliens, not the ghost of Elvis trying to communicate with the living through your > >hard drive, but what the spanish call "el seagate muerte" Your drive is dead, Jim. > Maybe I was exaggerating a little (I had just lost my two month work in a quite big > project), but it wasn't so simple. After two days, I changed my burned hdd (Quantum > Fireball 4.3Gb) with a brand new IBM 4.3Gb. When I came home and start reinstalling the > os, I just witness the second hdd crash. I changed again the hdd (also a IBM 4.3 Gb) > together (as suggested by my supplier) with the ide atapi cdrom (the two crashes > happened during cdrom copy) and with the cables. Again, trying to install my box, I > start receiving messages like: > wd0: interrupt timeout: > wd0: status 50 error 0 > wd0: interrupt timeout: > wd0: status 50 error 1 > I canceled everything. After rebooting, for about five minutes the hdd wasn't seen by > bios, and after that was seen being 435Mb. After a low level format it was ok. > Finally, I succeeded installing my box using a lpt Iomega ZIP drive (I got the dist > copied on ZIP, than I put it on a dos partition, ...). With this occasion, I saw the > famous blue error window of M$ Windows stating hdd write error during cdrom access. > Now I changed everything in my computer (except the ATX case and the CPU) but I am > seeing the same errors (the only thing that is quasi-the-same is the cdrom brand, the > rest is brand new - I insisted in breaking the seal personally). > Does anyone have any clue about this? I traced some previous threads about wd timeouts, > but they were no help. > The only chance to have it accessing the cdrom for more that 1 minute (but finally get > the same error) is to disable both UDMA and PIO mode for wd and wcd. > > Thanks, > Florin. > > BTW: I don't belive in aliens nor in ghosts or vampires :) > > -- > ------------------------------------------------------------------ > | Florin-Nicolae Nicolescu | > | University of Bucharest, Faculty of Mathematics | > | Bucharest,Romania | > ------------------------------------------------------------------ > | Friends don't let friends use Windows. | > | Double your hard drive space instantly! Delete Windows! | > ------------------------------------------------------------------ > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-hackers" in the body of the message To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message