From owner-freebsd-questions@FreeBSD.ORG Tue Oct 5 14:47:44 2004 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 84CCA16A4CE for ; Tue, 5 Oct 2004 14:47:44 +0000 (GMT) Received: from internet.potentialtech.com (h-66-167-251-6.phlapafg.covad.net [66.167.251.6]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5AC4F43D5A for ; Tue, 5 Oct 2004 14:47:44 +0000 (GMT) (envelope-from wmoran@potentialtech.com) Received: from working.potentialtech.com (pa-plum-cmts1e-68-68-113-64.pittpa.adelphia.net [68.68.113.64]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by internet.potentialtech.com (Postfix) with ESMTP id 98C0969A39; Tue, 5 Oct 2004 10:47:43 -0400 (EDT) Date: Tue, 5 Oct 2004 10:47:42 -0400 From: Bill Moran To: Rae Message-Id: <20041005104742.5babb2e2.wmoran@potentialtech.com> In-Reply-To: <2ede6f3204100507344ba2f245@mail.gmail.com> References: <2ede6f3204100507344ba2f245@mail.gmail.com> Organization: Potential Technologies X-Mailer: Sylpheed version 0.9.12 (GTK+ 1.2.10; i386-portbld-freebsd4.9) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit cc: freebsd-questions@freebsd.org Subject: Re: would async mount improve IO speed? X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 05 Oct 2004 14:47:44 -0000 Rae wrote: > File systems used to be mounted with soft-updates(?) option. > I disable it with "tunefs -n disable ....." and mount all file system > with "async" option but I can feel any difference. > some said it's dangerous. Does it matter even if I'm not running any > heavy load server? Async greatly increases the possiblity that a hard crash (i.e. power outage) will corrupt the disk such that it will be impossible to repair the filesystem, or that data will be corrupted or lost. If you have perfect power, and you can be 100% sure that your system will never hard boot, then there is no danger in async. Softupdates has 90% of the performance of async, with 90% of the safety of sync. This is why you won't see a big performance improvement unless you specifically test for it. Since softupdates is so fast, I can see only a few edge cases where it makes sense to use async. -- Bill Moran Potential Technologies http://www.potentialtech.com