On a Debian 64AMD server, I created two raids. /dev/md0 and /dev/md1.
On top of /dev/md1 I created a partition table and a volumegroup with 2 lv's to be used with lvm for a VTL.
After a reboot to enable the modules for the VTL software, the Lv's with the volumegroup and the partition table are gone.

What's going on here? Does anyone knows what could happen?
I tried to re-assemble, without result.

mdadm --examine /dev/md1:
mdadm: No md superblock detected on /dev/md1.

cat /proctab/mdstat:
Personalities : [raid1]
md1 : active raid1 sdb[0] sdc[1]
976631360 blocks super 1.2 [2/2] [UU]

md0 : active raid1 sda1[0]
54653824 blocks super 1.2 [2/1] [U_]

Anyone?

You're not supposed to run mdadm --examine on an md device; it's a command used to examine individual devices/partitions that may or may not be part of an md array. The error message is therefore perfectly normal; an md device should not have an md superblock, unless you've created a nested md device.

The problem may be that you've created a partition table on the md devices before creating logical volumes. What sort of partition table did you create? Can you still view it with the appropriate tool? If so, do the md partitions appear as devices in /dev (and if not, does running partprobe help)?

Does pvdisplay show any physical volumes at all?

1 members found this post helpful.