Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

i had an mdadm raid6 array with 13x1TB drives. Within 10 minutes 3 of these drives fell out of the array... we assume bad cable to the controller card and replaced, however now we need to get the drives back into a working array.

because md0 was marked as failed we removed the mdadm array and made a new md0 with the original 13 drives. 1 failed again during rebuild so we now have a degraded md0. The problem is that lvm does not see the array that exists within mdadm. Is there anything we can do to get our data back?

$pvscan

PV /dev/sda5   VG nasbox   lvm2 [29.57 GiB / 0    free]
  Total: 1 [29.57 GiB] / in use: 1 [29.57 GiB] / in no VG: 0 [0   ]

$cat /proc/mdstat

Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]
md0 : active raid6 sdc1[1] sdg1[5] sdb1[0] sdf1[4] sde1[3] sdd1[2] sdi1[7] sdl1[10] dm1[11] sdh1[6] sdj1[8] sdn1[12]
      10744336064 blocks super 1.2 level 6, 64k chunk, algorithm 2 [13/12] [UUUUUUUUU_UUU]

unused devices: <none>

what i think we need to do is get lvm to detect the mdadm array so that we can mount it, but if i create a new volume group in LVM, it'll wipe all the data from the array.

So to put it simply, How do we get our data from md0...

UPDATE: one of our sysadmins was able to restore an LVM config backup so it shows up in LVM now, however we are still unable to mount the drive for viewing of the data. Maybe bad partition table?

$pvscan

PV /dev/sda5   VG nasbox   lvm2 [29.57 GiB / 0    free]
PV /dev/md0    VG zeus     lvm2 [10.01 TiB / 4.28 TiB free]
Total: 2 [10.04 TiB] / in use: 2 [10.04 TiB] / in no VG: 0 [0   ]

$mount /dev/md0

mount: /dev/mapper/zeus-data already mounted or /mnt/zeus busy
share|improve this question
    
You're sure it's using LVM, right? What do you get from pvdisplay /dev/md0? –  Shane Madden Jul 29 '12 at 18:30

1 Answer 1

Information is taken from this article, see if it is of any help: http://www.linuxjournal.com/article/8874

pvscan shows that there are LVM physical volumes present, but LVM volume group is not necessarily active. See the output of vgdisplay, should be something like

--- Volume group ---
VG Name              zeus
System ID
Format               lvm2
[...]

If the zeus volume group does not show, you might need to make it available with vgchange zeus -a y. Then, check thet output of lvdisplay. This should output each logical volume in the system:

--- Logical volume ---
LV Name                /dev/zeus/volume_name
VG Name                zeus
LV Status              available
[etc...]

And you should be able to mount the desired volume with mount /dev/zeus/volume_name /mnt/target.

share|improve this answer
    
Unfortunetly the volumes were active, but the filesystem was wacked out. We think the partition tables were corrupt and out of sync as a result of the simultaneous drive drops. Couldn't get them back into sync... we've given up after 6 days at this. The downtime was beginning to outweigh the loss of data. The unfortunately tough lesson in all this: backups are still necessary. –  pironic Aug 2 '12 at 23:00

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.