64 bit Server 11.10 and MDADM problem

[Krupski]

Hi all,

I tried to upgrade my file server box running 11.04 64 bit server with 5 drives in a array to 11.10 server.

Now, my RAID array is /dev/md0 and composed of 5 drives with no partitions. That is, I created the array, formatted /dev/md0 with EXT4 and that's it. Worked fine.

Now MDADM for 11.10 complains that I have a "degraded array" and wants me to fix it or hit ctrl-D to continue (kinda difficult for a headless server box). It complains because it sees no "valid partition table".

If I do ctrl-D out of it and mount the array manually, it works fine. But the system won't boot without complaining first.

I see no reason why I should have to back up terabytes of data and rebuild my array only to add partitions to the drives so MDADM doesn't complain.

Is there any way around this problem? Or do I have to make my 11.04 my final version?

Thanks...

-- Roger

[rubylaser]

What's the output of these commands?

cat /proc/mdstat

cat /etc/mdadm/mdadm.conf

mdadm --detail /dev/md0

There's really no need for a partition on top of an mdadm array, so I'm betting it's not liking your mdadm.conf file.

[Krupski]

What's the output of these commands?

cat /proc/mdstat

cat /etc/mdadm/mdadm.conf

mdadm --detail /dev/md0

There's really no need for a partition on top of an mdadm array, so I'm betting it's not liking your mdadm.conf file.

mdstat:

root@storage:/# cat /proc/mdstat 
Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10] 
md0 : active raid6 sdd[3] sdb[1] sdc[2] sde[4] sda[0]
      2930284032 blocks super 1.2 level 6, 512k chunk, algorithm 2 [5/5] [UUUUU]
      
unused devices: <none>

mdadm.conf:

cat /etc/mdadm/mdadm.conf
# mdadm.conf
#
# Please refer to mdadm.conf(5) for information about this file.
#

# by default, scan all partitions (/proc/partitions) for MD superblocks.
# alternatively, specify devices to scan, using wildcards if desired.
DEVICE partitions

# auto-create devices with Debian standard permissions
CREATE owner=root group=disk mode=0660 auto=yes

# automatically tag new arrays as belonging to the local system
HOMEHOST <system>

# instruct the monitoring daemon where to send mail alerts
MAILADDR krupski@roadrunner.com

# definitions of existing MD arrays
ARRAY /dev/md0 UUID=311b8a28:5619dd16:d41b3874:59f9a3d7

mdadm --detail /dev/md0:

root@storage:/# mdadm --detail /dev/md0
/dev/md0:
        Version : 1.2
  Creation Time : Fri Dec 23 21:52:01 2011
     Raid Level : raid6
     Array Size : 2930284032 (2794.54 GiB 3000.61 GB)
  Used Dev Size : 976761344 (931.51 GiB 1000.20 GB)
   Raid Devices : 5
  Total Devices : 5
    Persistence : Superblock is persistent

    Update Time : Sun Feb  5 17:18:17 2012
          State : clean
 Active Devices : 5
Working Devices : 5
 Failed Devices : 0
  Spare Devices : 0

         Layout : left-symmetric
     Chunk Size : 512K

           Name : storage:0  (local to host storage)
           UUID : 311b8a28:5619dd16:d41b3874:59f9a3d7
         Events : 45

    Number   Major   Minor   RaidDevice State
       0       8        0        0      active sync   /dev/sda
       1       8       16        1      active sync   /dev/sdb
       2       8       32        2      active sync   /dev/sdc
       3       8       48        3      active sync   /dev/sdd
       4       8       64        4      active sync   /dev/sde

All looks normal to me. Note that this is running on my old 11.04 system. I yanked 11.10 because of the need to hit ctrl-D to boot.

(oh and please nobody yell at me for using the root account... I disabled UAC in Windows 7 too... and the world is still turning).

(p.s.: The boot drive is /dev/sdf - a 40GB SSD card which is not part of the array).

[rubylaser]

This looks correct, but I'm betting that you needed to recreate your /etc/mdadm/mdadm.conf file on 11.10 to get it to startup correct. Or, you could have run dpkg-reconfigure mdadm, and selected the all option for arrays to startup automatically.

[Krupski]

This looks correct, but I'm betting that you needed to recreate your /etc/mdadm/mdadm.conf file on 11.10 to get it to startup correct. Or, you could have run dpkg-reconfigure mdadm, and selected the all option for arrays to startup automatically.

Well, what worries me is in the past, when I've upgraded Ubuntu versions, I just did a fresh re-install (to the separate SSD drive which is not part of the RAID array), then installed MDADM and it found my existing array, generated the mdadm.conf file with the proper data and it all just worked.

Now, doing the same thing (moving from 11.04 to 11.10), upon installing MDADM I get the message that the array is "degraded" because "no valid partition table" is found.

Since what USED to always work now suddenly does NOT work, I have to assume that the problem lies elsewhere than my "upgrade technique".

It seems to me intuitively wrong that I now need to change some configuration to make it work when in the past I never had to.

If it had something to do with a fundamental change in the 2.6 -> 3.0 kernel design, then I could accept it.

Is it possible that an array without any partition table was "always wrong but ignored" in the past and now "being enforced as necessary"?

Thanks for all your input so far!

-- Roger

[rubylaser]

You don't need to change your configuration, but you do need to ensure that it exists otherwise mdadm won't know how to assemble the array. Finally, there's no need for a partition on top of mdadm unless, you're using LVM to create smaller volumes on top of mdadm, but you haven't mentioned that, so that's not the case.

On 11.04 what's the output of these?

mount

fdisk -l

cat /etc/fstab | grep md0