Problem with LVM2 and device-mapper

Hello,

I am using CentOS 4 and just did an upgrade from 4.1 to 4.2 on my x86_64 home server. After a reboot, one of my vol groups stopped working. It's a volume group spread across 2 disks and when I try to activate that volgroup I get the following:

<snip>
[root@hetfield mapper]# vgchange -ay VolGroup02
device-mapper ioctl cmd 9 failed: Invalid argument Couldn't load device 'VolGroup02-DVD'.
1 logical volume(s) in volume group "VolGroup02" now active
<snip>

and the relevant portion of the /var/log/messages has:

<snip>

Oct 17 08:45:07 hetfield kernel: device-mapper: dm-linear: Device lookup
failed
Oct 17 08:45:07 hetfield kernel: device-mapper: error adding target to

table
<snip>

My other two volgroups are working fine. It's just this third one that has problems.

I did quite a bit of googling yesterday, but I didn't come up with much although I have to confess that I know jack about the device mapper...

Is there anything I can provide to diagnose the problem further? Since CentOS tracks RHEL4 very closely, I suspect this affects RHEL4 as well.

Regards,

Pat

After a failed vgchange -ay, what does 'dmsetup ls' report?

I thought I've seen something like this in the past, but it always had to do with a partial load of a logical volume.

brassow

Hello,

<snip>
[root@hetfield ~]# vgchange -ay VolGroup02
device-mapper ioctl cmd 9 failed: Invalid argument Couldn't load device 'VolGroup02-DVD'.
1 logical volume(s) in volume group "VolGroup02" now active
[root@hetfield ~]# dmsetup ls
VolGroup00-Home (253, 1)
pdc_jahacaib (253, 6)
VolGroup00-Swap (253, 5)
VolGroup02-DVD (253, 10)
VolGroup00-Root (253, 0)
VolGroup01-Media (253, 9)
pdc_jahacaib1 (253, 7)
VolGroup00-Usr (253, 2)
VolGroup01-Data (253, 8)
VolGroup00-Var (253, 4)
VolGroup00-Local (253, 3)
[root@hetfield ~]# ls -lF /dev/mapper
total 0

crw-------  1 root root  10, 63 Oct 15 10:00 control
brw-rw----  1 root disk 253,  6 Oct 15 17:00 pdc_jahacaib
brw-rw----  1 root disk 253,  7 Oct 15 17:00 pdc_jahacaib1
brw-rw----  1 root disk 253,  1 Oct 15 10:00 VolGroup00-Home
brw-rw----  1 root disk 253,  3 Oct 15 10:00 VolGroup00-Local
brw-rw----  1 root disk 253,  0 Oct 15 10:00 VolGroup00-Root
brw-rw----  1 root disk 253,  5 Oct 15 10:00 VolGroup00-Swap
brw-rw----  1 root disk 253,  2 Oct 15 10:00 VolGroup00-Usr
brw-rw----  1 root disk 253,  4 Oct 15 10:00 VolGroup00-Var
brw-rw----  1 root disk 253,  8 Oct 15 17:00 VolGroup01-Data
brw-rw----  1 root disk 253,  9 Oct 15 17:00 VolGroup01-Media
brw-rw----  1 root disk 253, 10 Oct 18 14:32 VolGroup02-DVD

<snip>

The node is indeed created in /dev/mapper (which I assume 'dmsetup ls' is querying.) If I deactivate that volgroup the entry is removed.

Hope this helps.

Regards,

Pat

Patrick,

I am using Scientific Linux CERN SLC release 4.1. got the some problem after upgrade too.

Got the some problem after upgrade too.
For me it was "/sbin/dmraid -i -a y" in /etc/rc.d/rc.sysinit which grab and lock somehow PV on PDC (hmmm. RAID card). I don't know how it happend in details but looks like "rpm -e dmraid" should help.

Best regards,
 Valery Mitsyn