From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail.saout.de ([127.0.0.1]) by localhost (mail.saout.de [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id dYcJQ3zD5lqL for ; Sun, 8 Apr 2012 22:43:38 +0200 (CEST) Received: from mail-wi0-f178.google.com (mail-wi0-f178.google.com [209.85.212.178]) (using TLSv1 with cipher ECDHE-RSA-RC4-SHA (128/128 bits)) (No client certificate requested) by mail.saout.de (Postfix) with ESMTPS for ; Sun, 8 Apr 2012 22:43:38 +0200 (CEST) Received: by wibhq7 with SMTP id hq7so1880688wib.1 for ; Sun, 08 Apr 2012 13:43:37 -0700 (PDT) Message-ID: <4F81F876.1010909@googlemail.com> Date: Sun, 08 Apr 2012 22:43:34 +0200 From: artificial11000 MIME-Version: 1.0 References: <4F818CEF.3030706@googlemail.com> <20120408154753.GA31917@tansi.org> <4F81BB7A.7050508@googlemail.com> <20120408203530.GB4029@tansi.org> In-Reply-To: <20120408203530.GB4029@tansi.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: Re: [dm-crypt] can't open luks device after raid 5 rebuild List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: dm-crypt@saout.de Hi guys, thanks for your effort. Meanwhile I had to start over and format the whole device. I have a backup. Allthough it is not really up to date, I can work with it. Thanks again artificial On 04/08/2012 10:35 PM, Arno Wagner wrote: > On Sun, Apr 08, 2012 at 06:23:22PM +0200, artificial11000 wrote: >> Hi, >> >> as I noticed that I can't access the device, I recreated the raid manually. >> mdadm --create --level=5 -n 3 /dev/md127 /dev/sdb /dev/sdc /dev/sdd >> >> Could this damage the key-slot? >> Are the metadata of the RAID relevant to this? >> I read that old versions of mdadm created RAIDs with metadata=0.90 >> and the newer versions with metadata=1.20. > > This may be an additional problem or the only problem. Metadata 0.90 > is at the end of the device. Metadata 1.20 is 4k from the start of > the device. The other versions are in still other places. > (I have no idea what messed up process created these thorougly > insane changes.) > > So, yes, 4k offset is right in the first key-stripe. If so, > then your data is pemanently gone unless you have a header-backup. > In this case it is also possible that the array actually assembled > right and the problem is _only_ the new raid metadata block. > > Arno >