From owner-freebsd-advocacy Tue Mar 2 12:36:49 1999 Delivered-To: freebsd-advocacy@freebsd.org Received: from iquest3.iquest.net (iquest3.iquest.net [209.43.20.203]) by hub.freebsd.org (Postfix) with SMTP id E5E5414DFD for ; Tue, 2 Mar 1999 12:36:17 -0800 (PST) (envelope-from toor@y.dyson.net) Received: (qmail 6692 invoked from network); 2 Mar 1999 20:35:48 -0000 Received: from dyson.iquest.net (HELO y.dyson.net) (198.70.144.127) by iquest3.iquest.net with SMTP; 2 Mar 1999 20:35:48 -0000 Received: (from toor@localhost) by y.dyson.net (8.9.1/8.9.1) id PAA66502; Tue, 2 Mar 1999 15:35:47 -0500 (EST) Message-Id: <199903022035.PAA66502@y.dyson.net> Subject: Re: The Linux PR firestorm disaster (w.r.t. FreeBSD) In-Reply-To: <19990302184800.B4386@foobar.franken.de> from Harold Gutch at "Mar 2, 99 06:48:00 pm" To: logix@foobar.franken.de (Harold Gutch) Date: Tue, 2 Mar 1999 15:35:47 -0500 (EST) Cc: dwilde1@thuntek.net, kheuer@gwdu60.gwdg.de, freebsd-advocacy@FreeBSD.ORG From: "John S. Dyson" Reply-To: dyson@iquest.net X-Mailer: ELM [version 2.4ME+ PL38 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-advocacy@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Harold Gutch said: > On Tue, Mar 02, 1999 at 07:55:15AM -0800, Don Wilde wrote: > > FreeBSD uses synchronous writes of data, whereas Linux, by default, uses > > asynchronous writes. This means that Linux doesn't write immediately: > > your data is saved in RAM until the machine is less busy, and then it > > writes. If your machine dies, you've lost the data. FreeBSD makes sure > > your data is saved immediately, and it's so much more efficient that > > even with safer writing policies FreeBSD is faster. > > I saw a posting by Martin Cracauer to a German BSD-newsgroup a > couple of days ago, which can be summed up to "normal synchronous > writes (the 'classic' FreeBSD thing) are slow, asynchronous > writes (what Linux does) are dangerous - softupdates are a little > slower than asynchronous writes, but ensure the reliability of > synchronous writes. > In many cases, it seems that softupdates is slower than full async. However, softupdates, being much more intelligent, can avoid doing data I/O at all under certain circumstances. (Temp files.) Async isn't quite as smart. -- John | Never try to teach a pig to sing, dyson@iquest.net | it makes one look stupid jdyson@nc.com | and it irritates the pig. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message