I can't find my MDADM array to kill it

jbruyet · 04-01-2013, 05:34 PM

Hey all, I created an MDADM RAID array and apparently I didn't do it right this time. When I check /proc/mdstat this is what I see:

Code:

oot@Backup:/# cat /proc/mdstat
Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]
md10 : active (auto-read-only) raid10 sdd1[3] sdc1[2] sdb1[1] sda1[0]
      1953520640 blocks super 1.2 512K chunks 2 near-copies [4/4] [UUUU]
        resync=PENDING

I took a look at fdisk -l but the array wasn't there like it was in the past. I tried to look it from fdisk:

Code:

root@Backup:/dev# fdisk /dev/md10

Unable to open /dev/md10

For what it's worth I was using the instructions listed on this web page:

Code:

http://kromey.us/2009/08/raid-10-with-mdadm-65.html

I'm guessing that I'll have to manually destroy the array and start over but I'd prefer to find out what I did wrong and fix it rather than do a kludgy, brute-force repair. Any help here would be greatly appreciated.

Thanks,

Joe B

syg00 · 04-01-2013, 08:43 PM

Never learn if you don't take the opportunity to play - try "mdadm --readwrite /dev/md10".

Read the manpage, and take note of my sigline.

jbruyet · 04-02-2013, 10:58 AM

syg00, I guess I should have given that information too:

Code:

root@Backup:/dev# mdadm --readwrite /dev/md10
mdadm: error opening /dev/md10: No such file or directory
root@Backup:/dev#

I have all important stuff safely placed on another server. I just can't figure out why I can't get this working when I had it working a while back (I was going retask this machine so I eliminated the array).

Thanks,

Joe B

jbruyet · 04-08-2013, 06:30 PM

Ok, no more responses means it's time to Nuke & Repopulate. Thanks syg00.

Thanks,

Jobee

jbruyet · 04-09-2013, 04:47 PM

WELL, my array showed up in a different location. I just went to the server room to do my Nuke & Repopulate but before that I did a quick df -h and there it was under /media/dededfb1-85c4-40c1-a647-701b9025d7dd:

Code:

jobee@Backup:~$ df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/sde1              36G   14G   20G  41% /
tmpfs                 469M     0  469M   0% /lib/init/rw
udev                  464M  188K  464M   1% /dev
tmpfs                 469M     0  469M   0% /dev/shm
/dev/md10             1.8T  881G  860G  51% /media/dededfb1-85c4-40c1-a647-701b9025d7dd
jobee@Backup:~$

I checked and found all of the files intact. SO, how did this array move from /mnt/MD10 to its current location??? On top of that the results of some commands has changed:

Code:

root@Backup:/dev# cat /proc/mdstat
Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]
md10 : active raid10 sda1[0] sdd1[3] sdc1[2] sdb1[1]
      1953520640 blocks super 1.2 512K chunks 2 near-copies [4/4] [UUUU]
      [======>..............]  resync = 34.7% (679125120/1953520640) finish=106.2min speed=199963K/sec

unused devices: <none>
root@Backup:/dev#

and

Code:

root@Backup:/dev# fdisk /dev/md10
Device contains neither a valid DOS partition table, nor Sun, SGI or OSF disklabel
Building a new DOS disklabel with disk identifier 0x39098d5d.
Changes will remain in memory only, until you decide to write them.
After that, of course, the previous content won't be recoverable.

Warning: invalid flag 0x0000 of partition table 4 will be corrected by w(rite)

The device presents a logical sector size that is smaller than
the physical sector size. Aligning to a physical sector (or optimal
I/O) size boundary is recommended, or performance may be impacted.

WARNING: DOS-compatible mode is deprecated. It's strongly recommended to
         switch off the mode (command 'c') and change display units to
         sectors (command 'u').

Command (m for help):

and lastly

Code:

root@Backup:/dev# mdadm --readwrite /dev/md10
mdadm: failed to set writable for /dev/md10: Device or resource busy
root@Backup:/dev#

I was told earlier that we had a power outage last night. This server is not on a UPS at this time but it is set to boot up automatically when the power comes back on. Anyone have any ideas? I'm not new to Linux but this MDADM stuff is pretty new to me.

Thanks,

Joe B

chrism01 · 04-09-2013, 08:07 PM

You can't fdisk a SW RAID; fdisk works on physical disks.
You fdisk the component disks and set type to 'fd', then use mdadm to construct the RAID.
Note that for HW RAID, it is one disk from the OS' point of view (ie the real disks are 'invisible').
Try Chaps 4 & 6 here http://www.linuxtopia.org/online_boo...ion/index.html and/or Chap 13 http://www.linuxtopia.org/online_boo...storage_admin/