Can't Display vg00

[Geetha_1]

A C8000 workstation running HP-UX 11i, had boot problems after user manually powered down the server with all processes running. The kernel and vmunix.prev got corrupted. I had no luck trying to do an alternate boot from mirror. I reloaded the system from a boot image. The system is now up and running but I can't display vg00. It says

vgdisplay vg00
vgdisplay: Couldn't query volume group "vg00".
Possible error in the Volume Group minor number; Please check and make sure the group minor number is unique.
vgdisplay: Cannot display volume group "vg00".

I saved the old lvmtab file and did a vgscan -v but it didn't help. Any ideas, how I fix this? Thanks in advance

Sorry, I mistakenly opened the thread in Linux sys administration earlier.
Geetha.

[Coolmar]

Can you attach/post the following:

# strings /etc/lvmtab
# ioscan -funC disk
# vgdisplay -v /dev/vg00

[spex]

Hi Geetha,

Per the error message, look for VGs with duplicate minor device numbers:

# ll /dev/vg*/group | sort -k6

PCS

[Peter Godron]

Hi,
See page 33 in :
http://www4.itrc.hp.com/service/james/dispDoc.do?docURL=http%3A%2F%2Fsearch.hp.com%2Fredirect.html%3Furl%3Dhttp%253A%2F%2Fdocs.hp.com%2Fen%2F5991-1236%2FWhen_Good_Disks_Go_Bad.pdf%26qt%3D%252Bwhen%2B%252Bgood%2B%252Bdisks%26hit%3D1&aid=SEARCH_MANUAL&pil=8&serStr=when+good+disks

Other threads:
http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=966072

[Uday_S_Ankolekar]

usually moving lvmtab to a different place and run vgscan -v to create new lvmtab
If this did not work for you try to reboot in lvm maintenance mode, export vg00 and import the disks for vg00.
Before booting into maintenance mode write down the device files for the disks in vg00 and note their minor number for the group file.
Reboot the system and stop the boot process
boot on the pri disk and answer yes to interacting with isl at the isl prompt type hpux -lm

Once the unix prompt is reached.

vgexport vg00

*Note use a map file if special names are used for the lvols in vg00

mkdir /dev/vg00

mknod /dev/vg00/group c 64 0x000000

vgimport vg00 /dev/dsk/c?t?d? /dev/dsk/c?t?d?


Reboot the system.
Good luck..
-USA..

[Coolmar · Accepted solution]

This thread should help you:

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=962083

[Geetha_1]

Peter,
I tried the steps in pg 33 of the document. Excellent document but it didn't help.
Uday - I will try getting in LVM maint mode later this evening. Currently users are using the server and I can't work on it.
Thx.

[Sandman!]

You should follow Uday's advice in order to rebuild vg00 in LVM maintenance mode. Check out the tkb doc below too:

http://www1.itrc.hp.com/service/cki/docDisplay.do?docLocale=en_US&docId=200000068499361

[Geetha_1]

That seems to be a good document. I will keep you posted after I try it tonight - when there are no users logged on. Thanks for your suggestions.
Geetha.

[Geoff Wild]

Your answer is in the error:

"Possible error in the Volume Group minor number; Please check and make sure the group minor number is unique."

ll /dev/vg*/group

There is prbably more then 1 with the same minor number as vg00!

Change the duplicate.

Rgds...Geoff

[Geetha_1]

I don't have any duplicate minor #s. Here's the output...

ll /dev/*/group
crw-r----- 1 root sys 64 0x000000 Aug 18 13:37 /dev/vg00/group
crw-r--r-- 1 root sys 64 0x010000 Aug 21 11:35 /dev/vg_apps/group
crw-r--r-- 1 root sys 64 0x020000 Dec 4 10:58 /dev/vg_cargo/group

[Geetha_1]

Here's the output requested...

ROOT@LSOW-2DB: /apps/ug_tmp # ioscan -fnC disk
Class I H/W Path Driver S/W State H/W Type Description
=========================================================================
disk 0 0/2/1/0.3.0 sdisk CLAIMED DEVICE HP 73.4GST373307LW
/dev/dsk/c0t3d0 /dev/rdsk/c0t3d0
disk 1 0/2/1/0.4.0 sdisk CLAIMED DEVICE HP 73.4GST373307LW
/dev/dsk/c0t4d0 /dev/rdsk/c0t4d0
disk 2 0/2/1/0.5.0 sdisk CLAIMED DEVICE HP 73.4GST373453LW
/dev/dsk/c0t5d0 /dev/rdsk/c0t5d0
disk 3 0/2/1/0.6.0 sdisk CLAIMED DEVICE HP 73.4GST373453LW
/dev/dsk/c0t6d0 /dev/rdsk/c0t6d0
disk 4 0/3/2/0.1.0.0 sdisk CLAIMED DEVICE HL-DT-STDVD-ROM GDR8163B
/dev/dsk/c3t0d0 /dev/rdsk/c3t0d0
ROOT@LSOW-2DB: /apps/ug_tmp # strings /etc/lvmtab
/dev/vg00
/dev/dsk/c0t6d0
/dev/dsk/c0t5d0
/dev/vg_apps
/dev/dsk/c0t3d0
/dev/vg_cargo
/dev/dsk/c0t4d0

[Geoff Wild]

Okay - I have another suggestion.

Can you do a lvdisplay /dev/vg00/lvol1

If yes, is Mirror copies = 1 ?

If yes, try reducing the mirror on all lvols in vg00:

for vg in `ls /dev/vg00/lv*`
do
lvreduce -m 0 $vg
done

Then try vgdisplay /dev/vg00


If that works, then recreate your root mirror.

Rgds...Geoff

[Geetha_1]

Geoff,
I seem to have the mirrored disk in lvmtab but the lvdisplay shows no mirrored copies of the lvs. I tried lvreducing the lvols but am having no success.

vgreduce /dev/vg00 /dev/dsk/c0t5d0
vgreduce: Couldn't query volume group "/dev/vg00".
Possible error in the Volume Group minor number; Please check and make sure the group minor number is unique.

PS: c0t6d0 is boot disk and c0t5d0 is the mirror.

[Geoff Wild]

Can you try:

vgreduce /dev/vg00 /dev/dsk/c0t5d0


If yes, then you can do a pvcreate on it, and import it back into the vg.

What is happening, is both disks have that minor number - yet it really isn't in the vg - as you re-imaged it.

Rgds...Geoff

[Geetha_1]

Same problem...

ROOT@LSOW-2DB: /apps/ug_tmp # vgreduce /dev/vg00 /dev/dsk/c0t5d0
vgreduce: Couldn't query volume group "/dev/vg00".
Possible error in the Volume Group minor number; Please check and make sure the group minor number is unique.
Thx.

[Geoff Wild]

Doh - ignore my last post - sorry - just realized you tried it.

If you are absolutely sure that cot5d0 is the mirror - try pvcreate -f /dev/rdsk/c0t5d0

That may be risky - you may have to go to lvm maintenance mode as Uday suggested.

Rgds...Geoff

[Geetha_1]

Geoff,
Yes, I tried that too as I was sure it was the mirror.
vcreate -f /dev/rdsk/c0t5d0
pvcreate: The physical volume "/dev/dsk/c0t5d0" is already recorded in the "/etc/lvmtab" file.
but, it didn't work. I'll have to go to the lvm maintenance mode today when the users are not using the server.
Thx.

[Geoff Wild]

What about mv /etc/lvmtab /etc/lvmtab.bad

DON'T vgscan.

Then pvcreate?

Rgds...Geoff

[Geetha_1]

That worked, but when I do a vgscan, it doesn't show any vg00.

[Geoff Wild]

try vgscan -av

Rgds...Geoff

[Geetha_1]

Geoff - here's the output. Thx.
#vgscan -av

vgscan: has no correspoding valid raw device file under /dev/rdsk.
Verification of unique LVM disk id on each disk in the volume group
/dev/vg00 failed.

Scan of Physical Volumes Complete.

strings /etc/lvmtab
/dev/vg_apps
/dev/dsk/c0t3d0
/dev/vg_cargo
/dev/dsk/c0t4d0

[Geoff Wild]

Okay, this is very strange...

You could try a vgcfgrestore to the primary.

vgcfgrestore -n /dev/vg00 /dev/dsk/c0t6d0

I don't know if that will work or not - as the system is running - yet it doesn't think vg00 is there?

I think you sould try moving the lvmtab.bad to lvmtab and rerunning vgscan -av first.

Rgds...Geoff

[Calandrello]

Friend in this schemes has
mcserviceguard?

[Geetha_1]

No. It doesn't have Serviceguard. Users are still using the machine hence I can't try booting in LVM mode. Will keep you posted.
Geetha.