[SOLVED] mdadm: only give one device per ARRAY line: /dev/md/:raid and array

laughing_man77 · 03-09-2012, 03:15 PM

Hi all,

My base system is Ubuntu 11.10:

I recently fixed redirecting root mail to my account in thunderbird, and realised I was has receiving daily email reports from anacron (cron.daily), re mdadm.

Code:

/etc/cron.daily/mdadm:
mdadm: only give one device per ARRAY line: /dev/md/:raid and array

I looked up the particular cron job, and tried the command manually, and received the same error message

Code:

$ sudo mdadm --monitor --scan --oneshot
mdadm: only give one device per ARRAY line: /dev/md/:raid and  array

I checked out mdadm.conf, but could not see anything obvious, other than a reference to my old array (presumably from when I imported the backup after a botched remount after a reinstall), which I commented out. I tried referring to the device by its /dev path rather than UUID, but made no difference.
fstab:

Code:

proc            /proc           proc    nodev,noexec,nosuid 0       0
/dev/sda1       /               ext4    errors=remount-ro 0       1
# swap was on /dev/sda5 during installation
UUID=3f458052-e09e-4bc7-a54b-0fb9d7d114c6 none            swap    sw              0       0
#UUID=86dc953d-4ebc-4a6e-a4e5-8fbede80d552    /media/raid    ext4    defaults    0    2
/dev/md127    /media/raid    ext4    defaults    0    2

mdadm.conf

Code:

DEVICE partitions
ARRAY /dev/md/:raid metadata=1.2 name=:raid array UUID=f9099a38:9bd89ac8:a955705d:3a3244ad
#ARRAY /dev/md/:raid metadata=1.2 name=:raid array /dev/md127

# 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 root

# definitions of existing MD arrays
#ARRAY /dev/md/0 metadata=1.0 UUID=25c0930f:743a64e7:69503f88:5de3a7a4  name=linux-ynag:0

The array contains 4 disks in RAID5, and is mounted perfectly, and I can access the data with no problem. It worries me a little, because it may mean that there is something wrong in the configuration and/or I will not receive any disk fail alerts.

Thanks in advance

catkin · 03-11-2012, 12:03 AM

It could be the "array" in "name=:raid array" that is triggering the problem.

laughing_man77 · 03-11-2012, 12:57 AM

Quote:

Originally Posted by catkin

It could be the "array" in "name=:raid array" that is triggering the problem.

Yes, that fixed it. Editing mdadm.conf to

Code:

ARRAY /dev/md/:raid metadata=1.2 name=:raid UUID=f9099a38:9bd89ac8:a955705d:3a3244ad

Fixed it.

Thanks catkin.

catkin · 03-11-2012, 08:50 PM

Maybe not so simple unless the superblocks do not support names. From the mdadm.conf man page: "name= The value should be a simple textual name as was given to mdadm when the array was created. This must match the name stored in the superblock on a device for that device to be included in the array. Not all superblock formats support names."

laughing_man77 · 03-23-2012, 04:05 PM

Quote:

Originally Posted by catkin

Maybe not so simple unless the superblocks do not support names. From the mdadm.conf man page: "name= The value should be a simple textual name as was given to mdadm when the array was created. This must match the name stored in the superblock on a device for that device to be included in the array. Not all superblock formats support names."

Sorry for the late reply. I've been a bit hectic at work, and no time to test the concern.

I have to say that the warning emails have stopped, and when I reboot, the raid is still mounted and accessible. Is there possibly still something not right here?

Cheers