From owner-freebsd-doc@FreeBSD.ORG Tue Nov 6 02:27:20 2012 Return-Path: Delivered-To: doc@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id 501CDBAB for ; Tue, 6 Nov 2012 02:27:20 +0000 (UTC) (envelope-from wblock@wonkity.com) Received: from wonkity.com (wonkity.com [67.158.26.137]) by mx1.freebsd.org (Postfix) with ESMTP id EA9048FC14 for ; Tue, 6 Nov 2012 02:26:49 +0000 (UTC) Received: from wonkity.com (localhost [127.0.0.1]) by wonkity.com (8.14.5/8.14.5) with ESMTP id qA62Qhht088700 for ; Mon, 5 Nov 2012 19:26:43 -0700 (MST) (envelope-from wblock@wonkity.com) Received: from localhost (wblock@localhost) by wonkity.com (8.14.5/8.14.5/Submit) with ESMTP id qA62QD3v088697; Mon, 5 Nov 2012 19:26:43 -0700 (MST) (envelope-from wblock@wonkity.com) Date: Mon, 5 Nov 2012 19:26:13 -0700 (MST) From: Warren Block To: Alexander Yerenkow Subject: Re: what's the deal with changing release notes? In-Reply-To: Message-ID: References: User-Agent: Alpine 2.00 (BSF 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.2.7 (wonkity.com [127.0.0.1]); Mon, 05 Nov 2012 19:26:44 -0700 (MST) Cc: doc@freebsd.org X-BeenThere: freebsd-doc@freebsd.org X-Mailman-Version: 2.1.14 Precedence: list List-Id: Documentation project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 06 Nov 2012 02:27:20 -0000 On Sun, 4 Nov 2012, Alexander Yerenkow wrote: > Can I add my 2c at this point? Where did 10240 appeared from? Did someone > tried to find some optimal size for most of flashes/cards? I always try to > use some large number, like 1m or even more. I know by sad experience that > writing flash with default bs can take much longer than with large bs value. > Very quick and dirty test (from dev/urandom to 1gb partition on sd card) > give me that speed numbers (bs, speed): > 128m = 11,38 > 8m = 11,42 > 512k = 11,46 > 10k = 1,10 > > So, default advice on how write those images are unoptimised. Drop in speed > by factor of ten without any reason not god as for me. Agreed. The 10K number also looks like a magic number, and it is not. Huge buffer sizes don't really improve speed, as your benchmark shows. There's no significant difference between 512K and 128M. I would suggest using 64K or 128K for the value. These are large enough to give speed benefits yet still small enough to work even if memory is very limited. Note: FreeBSD's dd is not case-sensitive about unit modifiers, but Linux is, so the example should stick to upper case "K".