/etc/lvmtab is out of date with the running kernel:

[Sana Chaarani]

Hi,

I have the following message whenever i want
to backup my volume group vgEP1 :

vgcfgbackup: Warning: couldn't query physical volume "/dev/dsk/c20t0d0":
The specified path does not correspond to physical volume attached to
this volume group
vgcfgbackup: Warning: couldn't query all of the physical volumes.
vgcfgbackup: Couldn't query physical volume "/dev/dsk/c20t0d0":
The specified path does not correspond to physical volume attached to
this volume group
vgcfgbackup: /etc/lvmtab is out of date with the running kernel:Kernel indicates 9 disks for "/dev/vgEP1"; /etc/lvmtab has 10 disks.
Cannot proceed with backup.
---------------------------

I tried vgreduce -f. It still gives me the following message :

vgreduce: Couldn't query physical volume "/dev/dsk/c20t0d0":
The specified path does not correspond to physical volume attached to
this volume group
-----------------------

Then I did : mv /etc/lvmtab /etc/lvmtab.old
then vgscan -v to recreate it.
The disk "/dev/dsk/c20t0d0" is removed from
my vgEP1, but for some reason, the new
/etc/lvmtab couldn't find my two boot disks
giving me the following message :

Following Physical Volumes belong to one Volume Group.
Unable to match these Physical Volumes to a Volume Group.
Use the vgimport command to complete the process.
/dev/dsk/c1t2d0
/dev/dsk/c2t2d0
----------------------

Can I do a vgimport on vg00 while the system
is running ?
What seems to be the problem ?

Thanks,

Sana

[SS_6]

Hi Sana,please provide some more clarity on the problem. You can vgexport volume with preview mode to have map file
#vgexport -p -m /tmp/xx.map /dev/vgxx
and than reimport it after exporting fully by unmounting fs, deactivate vg. Take backup before you start doing any change. you can not vgexport vg00 rather you can only use preview ooption.

[Sana Chaarani]

THanks for your reply.

My problem is , why the vgscan that I ran
wasn't able to recreate the entries
for vg00 in the new lvmtab.
It says :
Use the vgimport command to complete the process

How can I vgimport vg00 while the system
is up and running.

[Steven E. Protter]

Interestingly enough, you can vgexport the volume group vg00 and possibly vgimport it with the system running. The filesystems will remain mounted.

Best practices though would dictate you bring the system down since there are likely to be obstacles to getting this done.

You may even need to boot the system into lvm maintenance mode.

boot from the console

interupt at the 10 second prompt.

bo

Y on intervention.

hpux -lm


SEP

[Todd McDaniel_1]

I dont think I would reboot if your system cant find its boot disks as you say.

Does c20t0d0 exist, what do you know about it. pvdisplay, diskinfo???

If you want to remove it do this... identify the pvkey...

Display all the available information about a logical volume, but
display pvkey instead of PV Name in the status and distribution map.

lvdisplay -v -k /dev/vgxx/lvolX

Then lvreduce -k


This option should be used only in the
special instance when you want to reduce a
mirrored logical volume on a physical volume
that is missing or has failed.

Sorry cant help more, busy with work but that should get you started. To removing the bad disk.

[Sana Chaarani]

Thanks for all how replied.
I ended up placing a call. The problem
of vg00, was because the server is
in a cluster environement. For vgscan to be
able to recognise vg00, we have to remove
the file /dev/slvmvg : this file has the same minor number as vg00 and is recreated ever time the cluster starts. Then re-run vgscan and everything was fine : the entry of vg00
was there, and my trouble disk in vgEP1
was gone.
One last command : vgchange -a y vgEP1
and then vgcfgbackup.

Thanks to all,

Sana