From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265680AbTIJTmV (ORCPT ); Wed, 10 Sep 2003 15:42:21 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265677AbTIJTlo (ORCPT ); Wed, 10 Sep 2003 15:41:44 -0400 Received: from voyager.st-peter.stw.uni-erlangen.de ([131.188.24.132]:43709 "EHLO voyager.st-peter.stw.uni-erlangen.de") by vger.kernel.org with ESMTP id S265664AbTIJTk0 (ORCPT ); Wed, 10 Sep 2003 15:40:26 -0400 X-Mailbox-Line: From galia@st-peter.stw.uni-erlangen.de Wed Sep 10 21:40:23 2003 Message-ID: <1063222822.3f5f7e2696eda@secure.st-peter.stw.uni-erlangen.de> Date: Wed, 10 Sep 2003 21:40:22 +0200 From: Svetoslav Slavtchev To: Linux Kernel Mailing List Subject: Re: wierd raid 1 problem(sort of) MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7BIT User-Agent: Internet Messaging Program (IMP) 3.2.1 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, i'm seeing similar problems with the bad side that the files never recover :( this is a Epox 8k9a3+ , 4 IBM deskstar on the HPT374 as masters in raid-1 & raid-5, with or without lvm1( resently recreated with LVM2) when i use md( or lvm over md) the data gets corrupted , if i use the partitions which are under the md as normal partitions there is no coruption, if i create md over them again - again corruption dma seems to be OK since 2.4.21ac1 & 2.4.22pre in case i use acpi=off pci=noacpi which also allows me to use usb and onboard lan the same setup had never had problems with erlier kernels on a Epox 8k5a3+ but it died (i moved the drives from dead board to the new one, which is basicly the same only uses KT333 instead of KT400 in the 8k9a3+) best, svetljo PS. please CC me as i'm not subscribed to the list --