From owner-freebsd-questions@FreeBSD.ORG Mon Feb 8 19:37:33 2010 Return-Path: Delivered-To: freebsd-questions@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id E30761065692 for ; Mon, 8 Feb 2010 19:37:33 +0000 (UTC) (envelope-from bf1783@googlemail.com) Received: from mail-ew0-f211.google.com (mail-ew0-f211.google.com [209.85.219.211]) by mx1.freebsd.org (Postfix) with ESMTP id 7089B8FC1D for ; Mon, 8 Feb 2010 19:37:33 +0000 (UTC) Received: by ewy3 with SMTP id 3so3872969ewy.13 for ; Mon, 08 Feb 2010 11:37:31 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=googlemail.com; s=gamma; h=domainkey-signature:mime-version:received:reply-to:date:message-id :subject:from:to:content-type; bh=U1/PjpiZr31FaJ0lCUzmWUkCzYWNKifmJZ4oYGnk1ls=; b=sapyIuLMqHPkw5jJU54fr6sJj+XWXTz/07Eb6vfjNUApP3u0jV1i/RH435qogLVBRq QF4ektXI/OVrd6PeAhYY+i4myXTsdTqawlLnzWqgkcDRCVoLua1HpD7xSXeNcRnh/jjl 0VeeL2yMdM7Bdx3VAWkOcAGfV2n8luys8xLbA= DomainKey-Signature: a=rsa-sha1; c=nofws; d=googlemail.com; s=gamma; h=mime-version:reply-to:date:message-id:subject:from:to:content-type; b=j53fsY4ogkbJNbgP6oh+wkgdLBwHU6/0XJyVFQhVzYJvtunDX8hK3iU2vJDZh8ovB2 DfgDQY0NDqi7kOcQNzSD4wixDxbxwcvItrqYmtqVponAexfjxZkBtqGhvMcTF55iMBWm 2KFX82gmNZ0usumLX2xL8dymLal3tAI2qJObQ= MIME-Version: 1.0 Received: by 10.216.85.143 with SMTP id u15mr2431744wee.205.1265657850734; Mon, 08 Feb 2010 11:37:30 -0800 (PST) Date: Mon, 8 Feb 2010 14:37:30 -0500 Message-ID: From: "b. f." To: freebsd-questions@FreeBSD.org Content-Type: text/plain; charset=ISO-8859-1 Cc: Subject: Re: Should root partition be first partition? X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: bf1783@gmail.com List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 08 Feb 2010 19:37:34 -0000 > You can even >leave gaps between partitions if you want, but that is pretty crazy >since it just wastes some of the available space. > >There have been quite a lot of recommendations on how to lay out a disk >for best performance, based on the observation that disk access times >vary depending on how far away the data is from the spindle, and the >expected usage patterns for the partition. Like any such advice, it >has tended to become less valid over time. Modern disks really don't >have any physical meaning to the Cylinder/Head/Sector style addressing >schemes[*] nowadays -- and you're pretty much bound to be using LBA >style addressing anyhow. Also, machines nowadays have so much RAM that >(a) swap is hardly ever used and (b) access to popular files is >frequently answered out of VM caches rathe than needing disk IO. Layout is still important, and leaving some blank space may not be so crazy. Here I'm thinking not so much of ordering (although one would probably be best served by the recommended default ordering), but of alignment, size, raid/stripe/concat configuration, and file system block and fragment size selection. Witness the (as much as tenfold) performance difference from simple changes, highlighted in the recent thread entitled 'File system blocks alignment' on freebsd-arch@ during December 2009 - January 2010, beginning with: http://lists.freebsd.org/pipermail/freebsd-arch/2009-December/009770.html If you're laying out a new disk, you may as well take a few minutes and get the most out of it, even if you're not going to invest in a lot of new hardware. b.