From owner-freebsd-hackers Sun May 16 17:17:20 1999 Delivered-To: freebsd-hackers@freebsd.org Received: from jane.lfn.org (brinternational.com [209.16.92.9]) by hub.freebsd.org (Postfix) with SMTP id E75E014C25 for ; Sun, 16 May 1999 17:17:18 -0700 (PDT) (envelope-from caj@lfn.org) Received: (qmail 2622 invoked by uid 100); 17 May 1999 00:17:17 -0000 Date: Sun, 16 May 1999 19:17:17 -0500 (CDT) From: Craig Johnston To: freebsd-hackers@freebsd.org Subject: Re: fsck and large file systems In-Reply-To: <199905162317.QAA75741@bubba.whistle.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Sun, 16 May 1999, Archie Cobbs wrote: > Alex Le Heux writes: > > Maybe I'm completely wrong here, but didn't I read somewhere that with > > softupdates it would theoretically be possible to boot the system before > > the fsck and fsck while it's running? > > Yes.. if you make the assumptions that: > > 1. There are no bugs in the soft updates code > 2. Your disk always writes its blocks atomically > > Then it follows that the only disk inconsistencies that are possible > during any boot-up are minor ones (eg, bitmap entry showing an allocated > block that's not pointed to by anything) that can be fixed by a background > daemon running at its leisure. As a consequence of doing this, your may > have free, but temporarily unavailable, disk blocks for a while until > the daemon finishes. Can we assume that the above is a goal we're aiming for at some point in the foreseeable future? That'd be real sweet. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message