From owner-freebsd-fs@FreeBSD.ORG Thu Nov 18 18:04:04 2010 Return-Path: Delivered-To: fs@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id A482710657C1 for ; Thu, 18 Nov 2010 18:04:04 +0000 (UTC) (envelope-from dougb@FreeBSD.org) Received: from mail2.fluidhosting.com (mx23.fluidhosting.com [204.14.89.6]) by mx1.freebsd.org (Postfix) with ESMTP id 345BE8FC18 for ; Thu, 18 Nov 2010 18:04:03 +0000 (UTC) Received: (qmail 25251 invoked by uid 399); 18 Nov 2010 18:04:02 -0000 Received: from localhost (HELO doug-optiplex.ka9q.net) (dougb@dougbarton.us@127.0.0.1) by localhost with ESMTPAM; 18 Nov 2010 18:04:02 -0000 X-Originating-IP: 127.0.0.1 X-Sender: dougb@dougbarton.us Message-ID: <4CE56A90.3020200@FreeBSD.org> Date: Thu, 18 Nov 2010 10:04:00 -0800 From: Doug Barton Organization: http://SupersetSolutions.com/ User-Agent: Mozilla/5.0 (X11; U; FreeBSD amd64; en-US; rv:1.9.2.12) Gecko/20101028 Thunderbird/3.1.6 MIME-Version: 1.0 To: Aditya Sarawgi References: <20100929031825.L683@besplex.bde.org> <20100929084801.M948@besplex.bde.org> <20100929041650.GA1553@aditya> <201009290917.05269.jhb@freebsd.org> <20100929202526.GA1564@aditya> <4CD0A3E8.4080304@FreeBSD.org> <4CD201AE.3040409@FreeBSD.org> <20101108174327.GC2066@earth> In-Reply-To: <20101108174327.GC2066@earth> X-Enigmail-Version: 1.1.2 OpenPGP: id=1A1ABC84 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: fs@freebsd.org Subject: Re: ext2fs now extremely slow 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: Thu, 18 Nov 2010 18:04:04 -0000 On 11/08/2010 09:43, Aditya Sarawgi wrote: > I have attached the patch. Finally got a chance to test it. There was a huge .rej file for _alloc.c, so I regenerated the patch against HEAD: http://people.freebsd.org/~dougb/ext2fs_prealloc-r215457.diff In light testing so far it seems ok, and it is slightly better from a performance standpoint, but still not as fast as 8.1. My test is csup'ing the ports tree without using -s. Admittedly that's a rather non-scientific test in that the network affects it, however I'm on a pretty fast link, and I'm forcing IPv4 just to be safe. This is the use case Times in seconds: 8.1 averages around 300 HEAD pre-patch average around 450 HEAD post-patch average around 370 So a good improvement to be sure, but still a ways to go. Doug -- Nothin' ever doesn't change, but nothin' changes much. -- OK Go Breadth of IT experience, and depth of knowledge in the DNS. Yours for the right price. :) http://SupersetSolutions.com/