From owner-freebsd-hackers Wed Jul 3 11:32:44 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id LAA08304 for hackers-outgoing; Wed, 3 Jul 1996 11:32:44 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.211]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id LAA08289 for ; Wed, 3 Jul 1996 11:32:42 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id LAA10899; Wed, 3 Jul 1996 11:26:01 -0700 From: Terry Lambert Message-Id: <199607031826.LAA10899@phaeton.artisoft.com> Subject: Re: async flag to mount(8) To: alex@fa.tdktca.com Date: Wed, 3 Jul 1996 11:26:01 -0700 (MST) Cc: dunn@harborcom.net, hackers@freebsd.org In-Reply-To: <199607031339.IAA08875@fa.tdktca.com> from "Alex Nash" at Jul 3, 96 08:39:17 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-hackers@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > > Has anyone ever had the stomach to mount a fs async? What have been > > your experiences? > > At the risk of experiencing Terry's wrath :), async is very useful, > and under *some* scenarios (like rm -rf), very fast. I think (almost) > everyone agrees that it's risky, but for filesystems such as /tmp one > really shouldn't care. Matt Day (next cubicle over from mine) has implemented soft updates, though without some of my FS changes, they can not be integrated into FreeBSD. An FS using soft updates operates within ~5% of memory speed, according to the Ganger/Patt paper from the University of Michigan. This is faster than pure async because of write gathering during the soft update timer period. Async sucks. Get over it. It is only useful until soft updates are integrated, and arguably a more dangerous cure than the disease even before that time. Regards, Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.