From mboxrd@z Thu Jan 1 00:00:00 1970 From: Phil Turmel Subject: Re: raid10 recovery assistance requested Date: Sun, 22 Sep 2013 23:27:13 -0400 Message-ID: <523FB511.5070001@turmel.org> References: <523A7DB9.9070600@hardwarefreak.com> <523F6692.9090407@turmel.org> <523F7C62.4040305@turmel.org> <523FAD81.9070005@turmel.org> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: Sender: linux-raid-owner@vger.kernel.org To: Dave Gomboc Cc: linux-raid@vger.kernel.org List-Id: linux-raid.ids On 09/22/2013 11:19 PM, Dave Gomboc wrote: > [snipped] >> However, based on the data you've reported, sda1 and sdc1 are identical, >> down to the superblock bytes. And sdb1 and sdd1 are also identical, >> down to the superblock bytes. >> >> Are you sure there were no misdirected "dd" operations? >> >> Phil > > I did not start using ddrescue until after I was no longer able to > boot the raid array. When I did the ddrescuing, I specifically used > the /dev/disk/by-id symlinks, not the sdX names, precisely in order to > avoid accidents. I was extremely alarmed by the inability to boot the > raid10, and took extra care to make sure I was doing it properly. Ok. If you'll generate the following for each member, we'll see what we can do: dd if=/dev/sdX1 bs=512 skip=2048 count=16 2>/dev/null |hexdump -C dd if=/dev/sdX1 bs=512 skip=3072 count=16 2>/dev/null |hexdump -C (These will be smaller than the other report... which didn't make it to the list.) Phil