From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Wed, 25 Jul 2001 06:00:39 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Wed, 25 Jul 2001 06:00:29 -0400 Received: from fe070.worldonline.dk ([212.54.64.208]:61197 "HELO fe070.worldonline.dk") by vger.kernel.org with SMTP id ; Wed, 25 Jul 2001 06:00:20 -0400 Date: Wed, 25 Jul 2001 11:50:55 +0200 From: Jens Axboe To: Satish Kumar Cc: linux-kernel@vger.kernel.org Subject: Re: bdev manipulation at block layer level Message-ID: <20010725115055.D834@suse.de> In-Reply-To: <20010725091716.71588.qmail@web11502.mail.yahoo.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20010725091716.71588.qmail@web11502.mail.yahoo.com> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Original-Recipient: rfc822;linux-kernel-outgoing On Wed, Jul 25 2001, Satish Kumar wrote: > Hi, > > I have a host PC connected to a disk array with > alternate paths to an end scsi disk. > I attempted load balancing across paths by changing > the b_dev & b_rdev ib the bufer_head list in > ll_rw_block function of ll_rw_blk.c (block layer), and > was successful in load balancing with data integrity, > on linux 2.2.16. > However, when I try the same stunt with a linux 2.4 ( > &2.4.2) kernel, I am seeing corruption occasionally. > Can anyone let me know if I am missing anything ? You must not change b_dev, that will screw the cache! Just change b_rdev to redirect the I/O and it should work. -- Jens Axboe