/etc/lvmtab is out of date with the running kernel

[Janne_5]

I get this when vgcfgbackup is run:

vgcfgbackup: /etc/lvmtab is out of date with the running kernel:Kernel indicates 7 disks for "/dev/vg01"; /etc/lvmtab has 6 disks

The number of current and active physical volumes, printed by vgdisplay as Cur PV and Act PV, are not the same. Cur PV is 4 and Act is 3

Did:
vgreduce -f vg01
Not all extents are free. i.e. Out of 868 PEs, only 0 are free.
You must free all PEs using lvreduce/lvremove before the PV can be removed.
Example: lvreduce -A n -m 0 /dev/vg01/lvol1.
lvremove -A n /dev/vg01/lvol1
skip alternate link /dev/dsk/c6t0d0
skip alternate link /dev/dsk/c6t0d2
skip alternate link /dev/dsk/c4t0d1
Here's the map of used PEs

--- Logical extents ---
LE LV PE Status 1
0000 lvol57 0000 ???
0001 lvol57 0001 ???
0002 lvol57 0002 ???

0397 lvol97 0866 ???
0398 lvol97 0867 ???

vgdisplay and lvdisplay don't display lvol57 and lvol97 so they can´t be lvremoved or lvreduced.

Recreating lvmtab with vgscan doesn't make any difference as it's kernel and not the lvmtab that is screwed up. Cur PV 4 and Act 3 remains unchanged.

Any advice?

[AwadheshPandey]

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=18493&admit=-682735245+1188462889506+28353475
http://docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad.pdf

[Janne_5]

Awadhesh,

When_Good_Disks_Go_Bad.pdf advises to locate any missing disks. Those disks are deleted from the diskarray so that's not an option anymore.

Tried;
vgreduce -f vg01
rm /etc/lvmtab
vgscan -v -a
don't help.

I'm going to try vgexport, vgimport trick as soon I get a service window.

[Steven E. Protter]

Shalom,

A disk has gone bad.

Identify the disk.

http://docs.hp.com/en/5991-1236/When_Good_Disks_Go_Bad.pdf

/mv /etc/lvmtab /etc/lvmtab.save

vgscan -a

SEP

[whiteknight]

Hi Jane,


Check this documents, I agreed with SEP, possible of the disk is bad.

Cur PV and Act PV do not agree in vgdisplay
--------------------------------------------
http://www1.itrc.hp.com/service/cki/docDisplay.do?docLocale=en&docId=emr_na-c01002819-2


WK
please assign points

[Torsten.]

"Kernel indicates 7 disks for "/dev/vg01"; /etc/lvmtab has 6 disks"

What do you expect? 6 disks?

This is most of the times a result of a wrong replacement procedure.

What steps have you done?

[Janne_5]

SEP, the disk is identified as gone. It's was removed by purpose.

WK, don't have access to the american site. Can't read.

Torsten, I expect 6.
3 disks + alt path = 6

Steps I have taken,
vgreduce -f vg01
mv /etc/lvmtab /etc/lvmtab.bak
vgscan -v -a
Tried above both in in lvm maintenance mode and multiuser.

reboot
vgexport -m /tmp/vg01.map /dev/vg01
mkdir /dev/vg01
mknod /dev/vg01/group c 64 0x01000
vgimport -m /tmp/vg01.map -f /tmp/vg01_disk_file /dev/vg01

None of it makes any diffrence:

n12901:/>vgcfgbackup vg01
vgcfgbackup: /etc/lvmtab is out of date with the running kernel:Kernel indicates 7 disks for "/dev/vg01"; /etc/lvmtab has 6 disks.
Cannot proceed with backup.

[Torsten.]

"Steps I have taken,
vgreduce -f vg01
mv /etc/lvmtab /etc/lvmtab.bak
vgscan -v -a"

What steps before this?
(did you replace the disk already?)

Most common mistake:
vgextend the new disk ... or similar.

[Janne_5]

Torsten,

Things got messed up when we migrated data from one diskarray to an other. We did this with lvm mirroring (lvextend -m 1). When the data was mirrored to new disk the old disk was deleted.
I think that we forgot to to remove mirroring (lvreduce -m 0) from the old disk and forgot removing disk from vg before the lun was deleted. But I'm not sure, as far as I know we never had lvol57 and lvol97 on the system.

It was a long time ago and my colleague whos on vacation now was doing the migration...

I'm going to log a support call to HP now to see if they can assist.

[Robert-Jan Goossens_1 · Accepted solution]

Janne, have a look at this doc from the Europen Tech database.

http://www8.itrc.hp.com/service/cki/docDisplay.do?docLocale=en&docId=emr_na-c01002594-2

Submitted Date: Wed Apr 18 07:00:00 CEST 2007
Title: lvmtab is out of date with running kernel
Document ID: emr_na-c01002594-2

Regards,
Robert-Jan