From mboxrd@z Thu Jan 1 00:00:00 1970 From: Wade Pinkston Subject: Re: raid not creating Date: Sun, 29 Jan 2006 21:02:35 -0800 Message-ID: <43DD9DEB.7060502@bugs.osu.edu> References: <57285.192.55.52.1.1138383357.squirrel@bugs.osu.edu> <17370.35932.856350.327787@cse.unsw.edu.au> <43DAD418.9040803@bugs.osu.edu> <17373.14913.427109.184586@cse.unsw.edu.au> <43DD7A27.6070806@bugs.osu.edu> <17373.33567.411470.796966@cse.unsw.edu.au> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <17373.33567.411470.796966@cse.unsw.edu.au> Sender: linux-raid-owner@vger.kernel.org To: linux-raid@vger.kernel.org List-Id: linux-raid.ids -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Neil Brown wrote: > On Sunday January 29, blata@bugs.osu.edu wrote: > >>>Neil Brown wrote: >>> >>> >>>>On Friday January 27, blata@bugs.osu.edu wrote: >>> >>> >>>>># mdadm --create --verbose /dev/md4 --level=1 --raid-devices=2 /dev/hde3 >>>>>/dev/hdg3 >>>>>mdadm: /dev/hde3 appears to contain an ext2fs file system >>>>> size=35214480K mtime=Thu Jan 26 13:38:20 2006 >>>>>mdadm: /dev/hdg3 appears to contain an ext2fs file system >>>>> size=35214480K mtime=Thu Jan 26 13:38:20 2006 >>>>>mdadm: size set to 35214400K >>>>>Continue creating array? y >>>>>mdadm: SET_ARRAY_INFO failed for /dev/md4: Device or resource busy >>> >>> >>>>Sounds like /dev/md4 already exists. What does >>>> cat /proc/mdstat >>>>show? >>> >>>>NeilBrown >>> >>> >>>cat /proc/mdstat >>>Personalities : >>>md2 : inactive hdc3[0] >>> 8723200 blocks >>>md3 : inactive hdg1[0] >>> 61440448 blocks >>>md4 : inactive hdg3[0] >>> 35214400 blocks >>>md1 : inactive hdc2[0] >>> 20482752 blocks >>>md0 : inactive hdc1[0] >>> 104320 blocks >>>unused devices: > > > So md4 (and others) is partially created. If you > mdadm --stop /dev/md4 > mdadm --create --verbose /dev/md4 --level=1 --raid-devices=2 /dev/hde3 /dev/hdg3 > > it should work. > > > >>>I got this only after doing the following. >>> >>>mdadm --manage --add /dev/md2 /dev/hdc3 >>> > > > I guess you are trying to add devices to non-existent arrays... Maybe > mdadm should complain about that... > > >>>for each md device. If I then go back and try to add /dev/hda1 to the md0 >>>I get this. >>> >>>mdadm --manage --add /dev/md0 /dev/hda1 >>>mdadm: add new device failed for /dev/hda1: Device or resource >>>busy > > > Maybe hda1 is currently in use? Maybe it is mounted as a filesystem? > > What exactly are you trying to achieve? > > NeilBrown > I live in California and this server is in Ohio ~3000 mi away. Any way I am rebuilding it remotely do some problems from a few months ago. Anyway I had someone in Ohio format the four drives. I was then left with a system on drives a and e. Now after the fact I'm trying to get the software raid working on a-c e-g. but I am unable to do so using mdadm. I'm trying to coordinate this with my Ohio counterpart when he is available. At this point I am thinking that I need to have him boot the box into rescue mode issue the mdadm commands, edit the fstab file and then reboot. Does this sound like a feasible approach? -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.6 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFD3Z3rv+6+qSFv3TURAkJ1AJ92JGA3HD3fozi0W+h5OsICi90cGQCgt0Ia WMOBvFOINP+HA0fUV85O0Ug= =P9vG -----END PGP SIGNATURE-----