From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263695AbTKACfx (ORCPT ); Fri, 31 Oct 2003 21:35:53 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263700AbTKACfw (ORCPT ); Fri, 31 Oct 2003 21:35:52 -0500 Received: from hawk.mail.pas.earthlink.net ([207.217.120.22]:6876 "EHLO hawk.mail.pas.earthlink.net") by vger.kernel.org with ESMTP id S263695AbTKACfv (ORCPT ); Fri, 31 Oct 2003 21:35:51 -0500 Subject: md wierdness with 2.6.0test9 From: Brad Langhorst To: linux-kernel@vger.kernel.org Content-Type: text/plain Message-Id: <1067654148.19557.409.camel@up> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.5 Date: Fri, 31 Oct 2003 21:35:48 -0500 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I have a pretty simple little raid 1 /dev/hda1 and /dev/hdc1 should be mirrors of each other. i created the raid like this mdadm --create --level 1 -n 2 /dev/md0 /dev/hdc1 missing (this is from memory - just to give you an idea of how this happened) i also created /dev/md1 for swap mdadm --create --level 1 -n 2 /dev/md1 /dev/hdc2 /dev/hda2 now when i go to add /dev/hda1 to the array all looks fine mdadm --zero-superblock /dev/hda1 mdadm --add /dev/md0 /dev/hda1 array syncs up but i see this wierdness (it exists across reboots) from mdadm --details /dev/md0 Number Major Minor RaidDevice State 0 22 1 0 active sync /dev/hdc1 1 0 0 -1 removed 2 3 1 1 spare /dev/hda1 rebooting to a 2.4 series kernel appears to fix this problem... In my search for information I read this post http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=utf-8&threadm=rvg0.5x4.17%40gated-at.bofh.it&rnum=4&prev=/groups%3Fq%3Dmdadm%2Bfaulty%2Bremoved%26hl%3Den%26lr%3D%26ie%3DUTF-8%26oe%3Dutf-8%26selm%3Drvg0.5x4.17%2540gated-at.bofh.it%26rnum%3D4 which talks about patching md.c to fix a similar problem... In test9 the patch discussed is already applied so it either doesn't fix the problem or this problem is not the same problem best wishes brad