From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264919AbUHHB5G (ORCPT ); Sat, 7 Aug 2004 21:57:06 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264929AbUHHB5G (ORCPT ); Sat, 7 Aug 2004 21:57:06 -0400 Received: from c66-235-4-168.sea2.cablespeed.com ([66.235.4.168]:17132 "EHLO darklands.zimres.net") by vger.kernel.org with ESMTP id S264919AbUHHB5E (ORCPT ); Sat, 7 Aug 2004 21:57:04 -0400 Date: Sat, 7 Aug 2004 18:54:58 -0700 From: Thomas Zimmerman To: linux-kernel@vger.kernel.org Cc: "Theodore Ts'o" , Buddy Lumpkin Subject: Re: EXT intent logging Message-ID: <20040808015458.GA11279@darklands> Reply-To: Thomas Mail-Followup-To: linux-kernel@vger.kernel.org, Theodore Ts'o , Buddy Lumpkin References: <20040806195615.GA14163@thunk.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20040806195615.GA14163@thunk.org> X-Operating-System: Linux darklands 2.6.8-rc1-ck5-amd64 X-Operating-Status: 19:39:30 up 1:21, 4 users, load average: 0.33, 0.24, 0.22 User-Agent: Mutt/1.5.6i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On 06-Aug 03:56, Theodore Ts'o wrote: > "mount -o data=journal" logs metadata blocks and data blocks into > journal first, and then after the transaction commits, the metadata > and data blocks are written to their final location on disk. The > problem with this is that all your write bandwidth is cut in half > since all block writes get written twice to disk --- once to the > journal, and once to the final location on disk. While you do half your write bandwidth, there arn't any _seeks_ while writing to the journal. For IO workloads where a sync allows you to move on to another job, this can speed up your workload. Mail delivery on an old laptop disk was ~2 as fast using ext3 data=journal... Thomas