From mboxrd@z Thu Jan 1 00:00:00 1970 From: Phil Turmel Subject: Re: Brocken Raid & LUKS Date: Fri, 22 Feb 2013 21:22:52 -0500 Message-ID: <512827FC.7010403@turmel.org> References: <5123A1CC.2000003@heisl.org> <51263785.2010001@turmel.org> <51263D9D.1080002@heisl.org> <51263F7E.7040207@turmel.org> <5126421E.3040702@turmel.org> <51264C18.8000201@heisl.org> <51264E26.9050100@turmel.org> <51264EBF.9090000@heisl.org> <51264F7F.3020508@turmel.org> <512650A1.7070103@heisl.org> <51265132.7070706@turmel.org> <512656B5.4090505@heisl.org> <51265824.4030407@heisl.org> <51265B0B.9020108@turmel.org> <51265DA7.2030209@heisl.org> <512660B9.8090609@turmel.org> <5126629A.1090002@heisl.org> <51266360.9030402@turmel.org> <5126678D.9030101@heisl.org> <51266D73.5020700@turmel.org> <51267192.6090205@heisl.org> <51267467.9040603@turmel.org> <512675A6.1000801@heisl.org> <5126797C.8090105@heisl.org> <51269DE0.5070905@heisl.org> <512748FA.2000709@heisl.org> <51277876.30008@turmel. org> <51278793.80904@heisl.org> <512790AE.2080102@turmel.org> <5127B64A.3000808@heisl.org> <5127D857.9090204@heisl.org> <5127F2B8.4050601@turmel.org> <5127F45F.5020608@heisl.org> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <5127F45F.5020608@heisl.org> Sender: linux-raid-owner@vger.kernel.org To: Stone Cc: linux-raid List-Id: linux-raid.ids On 02/22/2013 05:42 PM, Stone wrote: > Am 22.02.2013 23:35, schrieb Phil Turmel: >> Please recreate the array with this combination, then show: >> >> mdadm -D /dev/md2 >> mdadm -E /dev/sd[bce]1 >> cat /proc/partitions >> for x in /dev/sd[bce] ; do fdisk -l $x ; done > i dont know what chunk do you what but i think you mean 512k Yes. > mdadm --create /dev/md2 --assume-clean --chunk=512 --verbose --level=5 > --raid-devices=4 /dev/sdb1 /dev/sdc1 missing /dev/sde1 > mdadm -E /dev/sd[bce]1 > /dev/sdb1: > Avail Dev Size : 3907027037 (1863.02 GiB 2000.40 GB) > Array Size : 11721071616 (5589.04 GiB 6001.19 GB) > Used Dev Size : 3907023872 (1863.01 GiB 2000.40 GB) > /dev/sdc1: > Avail Dev Size : 3907024896 (1863.01 GiB 2000.40 GB) > Array Size : 11721071616 (5589.04 GiB 6001.19 GB) > Used Dev Size : 3907023872 (1863.01 GiB 2000.40 GB) See the difference? > /dev/sde1: > Avail Dev Size : 3907027037 (1863.02 GiB 2000.40 GB) > Array Size : 11721071616 (5589.04 GiB 6001.19 GB) > Used Dev Size : 3907023872 (1863.01 GiB 2000.40 GB) > /proc/partitions > major minor #blocks name > > 8 0 244198584 sda > 8 1 248832 sda1 > 8 2 1 sda2 > 8 5 243947520 sda5 > 252 0 235589632 dm-0 > 252 1 8351744 dm-1 > 8 16 1953514584 sdb > 8 17 1953514542 sdb1 > 8 32 1953514584 sdc > 8 33 1953513472 sdc1 And here? > 8 48 1953514584 sdd > 8 49 1953514542 sdd1 > 8 64 1953514584 sde > 8 65 1953514542 sde1 > 9 2 5860535808 md2 > for x in /dev/sd[bce] ; do fdisk -l $x ; done > > WARNING: GPT (GUID Partition Table) detected on '/dev/sdb'! The util > fdisk doesn't support GPT. Use GNU Parted. Oops. But you get the point, I hope. /dev/sdc has a different partition table from /dev/sdb and /dev/sde. That short partition is causing mdadm to make the array too small for the filesystem in it. You need to fix the partitions on /dev/sdc to exactly match /dev/sdb and /dev/sde. Can you explain how it might have become different? And use "parted /dev/sdb print" to show the partitions instead of fdisk. Phil