From owner-cvs-all Thu Jan 11 15:24: 2 2001 Delivered-To: cvs-all@freebsd.org Received: from ns.yogotech.com (ns.yogotech.com [206.127.123.66]) by hub.freebsd.org (Postfix) with ESMTP id D29B737B400; Thu, 11 Jan 2001 15:23:37 -0800 (PST) Received: from nomad.yogotech.com (nomad.yogotech.com [206.127.123.131]) by ns.yogotech.com (8.9.3/8.9.3) with ESMTP id QAA24182; Thu, 11 Jan 2001 16:23:14 -0700 (MST) (envelope-from nate@nomad.yogotech.com) Received: (from nate@localhost) by nomad.yogotech.com (8.8.8/8.8.8) id QAA05808; Thu, 11 Jan 2001 16:23:13 -0700 (MST) (envelope-from nate) From: Nate Williams MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <14942.16481.241677.610222@nomad.yogotech.com> Date: Thu, 11 Jan 2001 16:23:13 -0700 (MST) To: Doug Barton Cc: Matt Dillon , Sheldon Hearn , , , Subject: Re: cvs commit: src/etc crontab rc src/etc/defaults rc.conf src/etc/mtree BSD.root.dist src/libexec Makefile src/libexec/save-entropy Makefile save-entropy.sh In-Reply-To: References: <200101111912.f0BJCst72747@earth.backplane.com> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Reply-To: nate@yogotech.com (Nate Williams) Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG [ I've stayed out of this so far, and I probably shouldn't jump in, but I needed to state this one thing. ] > Yes, periodically writing things into / is "non-traditional" to say > the least, but I don't think it's going to set anyone's house on fire > either. Just to bring things back into perspective, periodically writing things to / is actually going to burn someone *really* badly, especially in -current. Because the root's FS state is constantly changing with this code (we-re not re-using files, we're rotating files, etc...), the chances of having a latent kernel bug corrupt your FS went up by a couple orders of magnitude. The chances of having this code running when 'something bad happens' which causes root filesystem are now higher than what I consider acceptable. Crashing a kernel in the middle of writing to the file system tends to cause bad things to happen. (Soft-updates minimizes this, but soft-updates don't work well on /, and I don't yet trust softupdates on /). Not-writing to / is akin to having all of the binaries in / static. It's not necessary when everything works fine, but when things go bad, boy is it ever nice. Finally, with -current not being confused with -stable at all lately (ie; crashes are a fairly common occurance), I would hate to see Yarrow et. al get blamed for trashed FreeBSD root partitions. The real bugs are the kernel bugs that causes the crash, but Yarrow will get blamed for unbootable FreeBSD boxes. 'Nuff said. Nate To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message