From owner-freebsd-hackers Tue Feb 6 11:24: 4 2001 Delivered-To: freebsd-hackers@freebsd.org Received: from mailtoaster1.pipeline.ch (mailtoaster1.pipeline.ch [62.48.0.70]) by hub.freebsd.org (Postfix) with SMTP id 3FCA837B491 for ; Tue, 6 Feb 2001 11:23:42 -0800 (PST) Received: (qmail 11919 invoked from network); 6 Feb 2001 19:20:30 -0000 Received: from unknown (HELO monzoon.net) ([195.134.133.140]) (envelope-sender ) by mailtoaster1.pipeline.ch (qmail-ldap-1.03) with SMTP for ; 6 Feb 2001 19:20:30 -0000 Message-ID: <3A804EF1.F44B1F07@monzoon.net> Date: Tue, 06 Feb 2001 20:22:25 +0100 From: Andre Oppermann X-Mailer: Mozilla 4.74 [en] (Windows NT 5.0; U) X-Accept-Language: en MIME-Version: 1.0 To: Matt Dillon Cc: Poul-Henning Kamp , Charles Randall , Dan Phoenix , Alfred Perlstein , Jos Backus , freebsd-hackers@FreeBSD.ORG Subject: Re: soft updates and qmail (RE: qmail IO problems) References: <36088.981481432@critter> <200102061750.f16HoZD62214@earth.backplane.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Matt Dillon wrote: > > :> softupdates disk may wind up unwinding 'more' of the last few moments > :> worth of operations then a normal filesystem would. And, I might add, > :> Reiser is the same way. > :> > :> The only way to guarentee that file data is written to disk, with any > :> filesystem no matter how it is mounted (even sync mounted filesystems), > :> is by calling fsync(). > :> > :> So I would stick with softupdates. > : > :... provided that qmail calls fsync(2). > : > :-- > :Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 > > It doesn't matter whether qmail calls fsync or not as far as using > softupdates goes. No filesytem will guarentee stable storage > with fsync(), so softupdates is not going to be too much worse then > other FS's. So one might as well use softupdates. When/if qmail gets > its act together and calls fsync() properly, softupdates will guarentee > a level of consistency that only ReiserFS or XFS can even come close > to matching. A normal EXT2FS or FFS filesystem, even with fsync(), > *EVEN* with synchronous mounts, cannot guarentee file consistency > because it is still possible to corrupt the directory and loose the > file that was fsync()'d if a crash were to occur just after the fsync(). No, see Julians previous mail. fsync() only returns when all associated inodes and directory entries up the root are made. Qmail does the right thing. -- Andre To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message