From owner-freebsd-hackers Tue Feb 6 10: 6:37 2001 Delivered-To: freebsd-hackers@freebsd.org Received: from earth.backplane.com (earth-nat-cw.backplane.com [208.161.114.67]) by hub.freebsd.org (Postfix) with ESMTP id 5337137B4EC for ; Tue, 6 Feb 2001 10:06:20 -0800 (PST) Received: (from dillon@localhost) by earth.backplane.com (8.11.1/8.9.3) id f16I3m262539; Tue, 6 Feb 2001 10:03:48 -0800 (PST) (envelope-from dillon) Date: Tue, 6 Feb 2001 10:03:48 -0800 (PST) From: Matt Dillon Message-Id: <200102061803.f16I3m262539@earth.backplane.com> To: Poul-Henning Kamp Cc: Charles Randall , Dan Phoenix , Alfred Perlstein , Jos Backus , freebsd-hackers@FreeBSD.ORG Subject: Re: soft updates and qmail (RE: qmail IO problems) References: <36239.981482344@critter> Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG :> with fsync(), so softupdates is not going to be too much worse then :> other FS's. : :Actually, if you don't use fsync you do loose more work with :softupdates than if you use plain UFS. : :Softupdates can delay directory updates which plain UFS will runs :synchronously, and consequently you can loose stuff you throught :you had. : :-- :Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 Yes, that's true. Except if you crash in the middle of doing significant directory updates, especially when you are deleting AND creating files in the same directory at a high rate, there's a good chance that fsck will blow the entire directory away as being corrupt with a normal UFS mount. The synchronous metadata updates that a normal UFS mount will do only reduces the probability of significant loss of data for *LIGHTLY* treaded directories. Any directory under heavy use ... well, softupdates or a log structured filesystem coupled with judicious fsync() use is your only real choice. -Matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message