From owner-cvs-all Wed Jul 22 00:50:34 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id AAA01767 for cvs-all-outgoing; Wed, 22 Jul 1998 00:50:34 -0700 (PDT) (envelope-from owner-cvs-all@FreeBSD.ORG) Received: from ns.mt.sri.com (sri-gw.MT.net [206.127.105.141]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id AAA01736; Wed, 22 Jul 1998 00:50:27 -0700 (PDT) (envelope-from nate@mt.sri.com) Received: from mt.sri.com (rocky.mt.sri.com [206.127.76.100]) by ns.mt.sri.com (8.8.8/8.8.8) with SMTP id BAA21501; Wed, 22 Jul 1998 01:47:18 -0600 (MDT) (envelope-from nate@rocky.mt.sri.com) Received: by mt.sri.com (SMI-8.6/SMI-SVR4) id BAA17271; Wed, 22 Jul 1998 01:47:16 -0600 Date: Wed, 22 Jul 1998 01:47:16 -0600 Message-Id: <199807220747.BAA17271@mt.sri.com> From: Nate Williams MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: "Jordan K. Hubbard" Cc: Bruce Evans , asami@FreeBSD.ORG, committers@FreeBSD.ORG, mike@smith.net.au, stable@FreeBSD.ORG Subject: Re: Tree tagging put off by ~12 hours. In-Reply-To: <12229.901089421@time.cdrom.com> References: <199807220541.PAA25672@godzilla.zeta.org.au> <12229.901089421@time.cdrom.com> X-Mailer: VM 6.29 under 19.15 XEmacs Lucid Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk [ Bringing the Cc list down to size ] > You still don't seem to understand. "Worse" is freefall crashing and > denying service for a problem which is still unknown but seems fairly > non-fatal (at least unless it panics the system). Just because it continues on doesn't mean it's 'non-fatal'. The original FreeBSD box was my 486/33, and it suffered these 'non-fatal' NMI's as well. Unfortunately, I found out much too late that I was corrupting memory, disk, *AND* my backups. Mine was due to bad cache that I replaced ~1.5 years in after my system became so unstable as to be unusable. Almost every file on that box that was older than 24 hours would become corrupted, and cause panic's on a regular (2-3 times/day) basis. After this, I decided to fix the problem rather than bandaid it. The NMI's on the laptop (which NMI_POWERFAIL is all about) is a completely different animal. It's a function of a *really* old BIOS that exists on one machine, and that's PHK's. No other piece of hardware exhibits this problem, and that's where it generates an NMI if you resume from suspended mode. The 'fix' you put in place can potentially corrupt the CVS repository and cause even more instability on freefall than already exists. By panic'ing you are (hopefully) limiting the damage that can be done instead of allowing it to continue. (And yes, for the most part the system will continue. For months even, but the machine and all data on it will eventually become useless.) > Since you can't fix it, don't make it worse by making stupid > suggestions. :-) Since you aren't willing to fix it, don't make it worse by implementing wrong solutions. :) Been there, done that, would hate to see the CVS tree spammed world-wide by hardware problems on the master repository. If it continues, I certainly hope the Repository on the upcoming CDROM is ok.... Nate To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message