From owner-freebsd-fs@FreeBSD.ORG Sat Sep 4 12:09:04 2010 Return-Path: Delivered-To: freebsd-fs@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id ABEA210656B0 for ; Sat, 4 Sep 2010 12:09:04 +0000 (UTC) (envelope-from marco@tolstoy.tols.org) Received: from tolstoy.tols.org (tolstoy.tols.org [IPv6:2a02:898:0:20::57:1]) by mx1.freebsd.org (Postfix) with ESMTP id 445FC8FC0A for ; Sat, 4 Sep 2010 12:09:03 +0000 (UTC) Received: from tolstoy.tols.org (localhost [127.0.0.1]) by tolstoy.tols.org (8.14.4/8.14.4) with ESMTP id o84C907t040724 for ; Sat, 4 Sep 2010 12:09:00 GMT (envelope-from marco@tolstoy.tols.org) X-Virus-Status: Clean X-Virus-Scanned: clamav-milter 0.96.1 at tolstoy.tols.org Received: (from marco@localhost) by tolstoy.tols.org (8.14.4/8.14.4/Submit) id o84C906Q040723 for freebsd-fs@freebsd.org; Sat, 4 Sep 2010 14:09:00 +0200 (CEST) (envelope-from marco) Date: Sat, 4 Sep 2010 14:09:00 +0200 From: Marco van Tol To: freebsd-fs Message-ID: <20100904120900.GB40103@tolstoy.tols.org> Mail-Followup-To: freebsd-fs References: <20100903130646.GD19666@tolstoy.tols.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.4.2.3i X-Spam-Status: No, score=-2.9 required=5.0 tests=ALL_TRUSTED,BAYES_00 autolearn=ham version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on tolstoy.tols.org Subject: Re: just another sad story in zfs tuning city X-BeenThere: freebsd-fs@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Filesystems List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 04 Sep 2010 12:09:04 -0000 On Sat, Sep 04, 2010 at 09:25:44AM +0900, Randy Bush wrote: > >> i am tempted to just boot without the zfs memory hacks in loader > >> conf. any warnings on doing so? any other clues also gladly > >> accepted. > > > > Sticking a wet finger in the air, and assuming /dev/mirror/boota holds a > > UFS filesystem that's just used to boot from I'd say give kmem_size 1.5x > > physical, arc_max 1.0x physical, and arc_min 0.5x physical memory to > > have it compete with the UFS page cache. After that applications that > > run can't use more then (0.5x physical memory plus swap space). > > thanks. it ain't crashed yet. running some large builds to see how it > performs. > > but what about just taking the defaults, no parms? one would hope that > this would be successful, as that is the simple path. The defaults go well under what I wrote, and I'm pretty sure for good reasons. I _think_ that has to do with that the defaults have to be prepared for lots more situations then those we setup and tune the values for. And ofcourse the arc goes into wired memory as opposed to page cache in active/inactive under UFS. I'm not an expert enough to go into lots more detail, but I think what I write is accurate. :) Marco -- The first step to better times is to imagine them. - www.chinese-fortune-cookie.com