From owner-freebsd-hackers Mon Mar 29 9:31: 7 1999 Delivered-To: freebsd-hackers@freebsd.org Received: from smtp2.vnet.net (smtp2.vnet.net [166.82.1.32]) by hub.freebsd.org (Postfix) with ESMTP id 92C5614FA0 for ; Mon, 29 Mar 1999 09:30:59 -0800 (PST) (envelope-from rivers@dignus.com) Received: from dignus.com (ponds.vnet.net [166.82.177.48]) by smtp2.vnet.net (8.9.1a/8.9.1) with ESMTP id MAA14614; Mon, 29 Mar 1999 12:32:05 -0500 (EST) Received: from lakes.dignus.com (lakes.dignus.com [10.0.0.3]) by dignus.com (8.9.2/8.8.5) with ESMTP id MAA00896; Mon, 29 Mar 1999 12:30:25 -0500 (EST) Received: (from rivers@localhost) by lakes.dignus.com (8.9.2/8.6.9) id MAA11166; Mon, 29 Mar 1999 12:30:25 -0500 (EST) Date: Mon, 29 Mar 1999 12:30:25 -0500 (EST) From: Thomas David Rivers Message-Id: <199903291730.MAA11166@lakes.dignus.com> To: avalon@coombs.anu.edu.au, mjacob@feral.com Subject: Re: another ufs panic.. Cc: hackers@FreeBSD.ORG In-Reply-To: <199903291723.DAA01298@cheops.anu.edu.au> Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG > > In some mail from Matthew Jacob, sie said: > > > > > In article <199903291550.BAA28981@cheops.anu.edu.au>, > > > Darren Reed wrote: > > > > Well, I toggled the internal BIOS termination setting and > > > > it still crashes. > > > > > > Don't be insulted if this is too obvious, but ... have you run fsck > > > on all your filesystems since you fixed the termination and cabling? > > > There could be a lot of residual damage in your filesystems from > > > earlier errors. > > > > > > > And don't believe the 'clean' bit if you've had I/O errors. > > well, I've been newfs'ing the destination partitions each time, if that > answers that question, which is where the trouble is showing up. > Well - you're not going to like this, but at one time the reproduction I had with 2.2.5 took the following steps: 1) Write 0xff all over the disk partition 2) newfs the partition 3) Do an fsck to find that 0x00 wasn't properly written, some inodes had 0xff in them... (that is, fsck of a newfs'd partition reported errors) So - if this is the same problem I had, doing a newfs doesn't reliably clean things up. (I actually got it narrowed down to writing a single 0xff in one spot on the disk.) But - I did find that fsck was able to repair things. I believe because it did things in a different order than newfs did. I suppose the moral is - if you're having these kind of problems, don't trust newfs either... :-( - Dave Rivers - To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message