vg00 missing from /etc/lvmtab

[Thierry D]

Hi all,

Got a problem so far with no production issue but it is bugging me ...

The lvmtab does not contain anymore vg00 and the corresponding disk.

What I already did:
* checked via vgdisplay but it does not know vg00

vgdisplay: Volume group "/dev/vg00" does not exist in the "/etc/lvmtab" file.
vgdisplay: Cannot display volume group "vg00".


* move lvmtab to lvmtab.old and vgscan -a -v
--> still no vg00 and disks

* checked boot device which corresponds to the correct disk
vmunix: Boot device's HP-UX HW path is: 0/0/4/1/0.10.0
setboot:
Primary bootpath : 0/0/4/1/0.10.0
HA Alternate bootpath :
Alternate bootpath : 0/0/4/1/0.8.0

* tried vgimport with the mapfile and it did not work because vg00 even though it is not in the lvmtab is active ...

Beginning the import process on Volume Group "vg00".
Volume group "/dev/vg00" is still active.

* so tried to deactivate vg00 but since it is not in the lvmtab it cant deactivate it ...

vgchange: Volume group "/dev/vg00" does not exist in the "/etc/lvmtab" file.

* checked the disk via diskinfo and ioscan with no problem there ... size OK and claimed and special files present too

What I am not sure and would need more info

** when using the lvm11 command, I can see there is some lvm header and info but I dont know how to "decode" it and to definitively know its for vg00 and the current system ...

** also I thought of using the vgcfgrestore to restore the header, but would like to be sure I wont be screwed the system more than it is, because so far all is still working just LVMTAB which is complaining ....

** Any one as some ideas ?

Thanks

Thierry

[Steven E. Protter]

Shalom,

Mysterious.

/usr/sbin/lvlnboot -v

Please post that output.

Looks to me as if vg00 has been exported with vxexport

You should be able to use vgimport to import it.

If you have used standard naming of logical volumes, that should not present any issues.

Please also post.

bdf

Select and post the stuff related to vg00 if any exists.

SEP

[點燃]

saludos Thierry,

Happened to see this link?

http://h30499.www3.hp.com/t5/LVM-and-VxVM/vg00-missing-from-lvmtab-vgdisplay-and-lvlnboot/m-p/3175706#M18949

Would it be possible to boot the machine to lvm maintenance mode?
Do you have an ignite backup handy?
Please post the bdf output here.

mejor de las suertes.

[Uday_S_Ankolekar]

do you have file called lvmtab_p

Check with strings /etc/lvmtab_p

V.2.0 of 'vgcreate doesnot update lvmtab

-USA..

[Thierry D]

Hello,

thanks for the different tips and here some more stuff about it.

First, I can't reboot in LVM maintenance mode, server in production and there is no way I will get a time window to do that. Regarding backup via ignite yes there are, but I am not sure exactly when they were made, should check it out ...

Then:

* there is no lvmtab_p file in /etc
* vgimport from vg00 mapfile does not work
#vgimport -m vg00.mapfile -v vg00
Beginning the import process on Volume Group "vg00".
Volume group "/dev/vg00" is still active.
#vgchange -a n vg00
vgchange: Volume group "/dev/vg00" does not exist in the "/etc/lvmtab" file.

* bdf looks fine
#bdf | egrep "Filesystem|vg00"
Filesystem kbytes used avail %used Mounted on
/dev/vg00/lvol3 1015808 460296 551232 46% /
/dev/vg00/lvol1 1835008 211584 1610776 12% /stand
/dev/vg00/lvol8 8912896 3360544 5509416 38% /var
/dev/vg00/lvol9 20480000 63099 19140916 0% /var/adm/crash
/dev/vg00/lvol7 7831552 2813320 4979048 36% /usr
/dev/vg00/lvol4 524288 144792 379496 28% /tmp
/dev/vg00/lvol6 9338880 4995800 4309400 54% /opt
/dev/vg00/lvol5 131072 16056 114184 12% /home

* there is no minor number duplicates for the different VGs
#ls -l /dev/*/group
crw-r----- 1 root sys 64 0x150000 Sep 1 2010 /dev/ipbvg03/group
crw-r----- 1 root sys 64 0x110000 Apr 8 2008 /dev/ippvg01/group
crw-r----- 1 root sys 64 0x100000 Apr 8 2008 /dev/omevg01/group
crw-r----- 1 root sys 64 0x160000 Dec 9 14:16 /dev/omevg02/group
crw-r----- 1 root sys 64 0x000000 Mar 10 2008 /dev/vg00/group
crw-r----- 1 root sys 64 0x130000 Jun 4 2008 /dev/vgappli01/group
crw-r----- 1 root sys 64 0x050000 Mar 26 2008 /dev/vglogiciel/group
crw-r----- 1 root sys 64 0x010000 Mar 26 2008 /dev/vgswap/group

* vg00 is using the conventional naming scheme defined automatically "lvolXX"
#ls -la /dev/vg00/*
crw-r----- 1 root sys 64 0x000000 Mar 10 2008 /dev/vg00/group
brw-r----- 1 root sys 64 0x000001 Mar 10 2008 /dev/vg00/lvol1
brw-r----- 1 root sys 64 0x000002 Mar 10 2008 /dev/vg00/lvol2
brw-r----- 1 root sys 64 0x000003 Mar 10 2008 /dev/vg00/lvol3
brw-r----- 1 root sys 64 0x000004 Mar 10 2008 /dev/vg00/lvol4
...

Now probably the interesting parts :

*lvlnboot is empty ... only shows "alternate link"
#lvlnboot -v
Current path "/dev/dsk/c20t0d0" is an alternate link, skip.
Current path "/dev/dsk/c20t0d1" is an
...
Current path "/dev/dsk/c20t2d5" is an alternate link, skip.
Current path "/dev/dsk/c20t2d6" is an alternate link, skip.

* here is the former lvmtab which should be the correct one
#strings lvmtab.red| more
/dev/vg00
,cNIG
/dev/dsk/c0t10d0s2
/dev/dsk/c0t8d0s2
/dev/vgswap
,cNIG
/dev/dsk/c2t2d7
/dev/dsk/c4t2d7
...

(OK I know I could use this one but there were newly added VGs, so this one is useless)

* here the vg00 mapfile and config file
#strings vg00.conf
CONFIG01
/dev/vg00
/dev/rdsk/c0t10d0s2
/dev/rdsk/c0t8d0s2
LVMREC01,cNIG
,cNIG
HPLVMBDRMF
,cNIG
LVMREC01,cNIG
,cNIG
HPLVMBDRMF
,cNIG
,cNIG
,cNIG
,cNIG
VGDA0001
VGSA0001M

#strings vg00.mapfile
VGID 2c634e4947d50ed7
1 lvol1
2 lvol2
3 lvol3
4 lvol4
5 lvol5
6 lvol6
7 lvol7
8 lvol8
9 lvol9


Any more thoughts ?

Thierry

[Steven E. Protter · Accepted solution]

Shalom,

You data pretty much proves my guess.

vg00 was exported. It needs to be imported with vgimport. You have the necessary files.

SEP

[Thierry D]

Hi Steven,

okay sounds a plan, but then one question:

how can I vgimport vg00 if the system is active, is there a way to do it "online" ?

if not I am guessing I should either boot in single or lvm maintenance, correct ?

Thierry

[rdiaz]

If you have a backup of the file /etc/lvmtab, (eg Data Protector), take that file and then simply place it in /etc.

It is important that before doing this, backup the current file /etc/lvmtab, eg:

# mv /etc/lvmtab /etc/lvmtab.old

Then do a vgscan to find any difference...

Yesterday I had this same situation after a ignite, I do not sincerely to be this situation,anyway, if the vgexport clear the structure of the "logical volume" is also necessary to restore the structure relates to vg00 in /dev/vg00

Finally do the following:

Verify everything looks good with a bdf , vgdisplay and lvdisplay

All volume groups are back and everything looks correct.

[UVK]

have you tried running vgscan ? as far as i know its not intrusive but will recreate your /etc/lvmtab

[Steven E. Protter]

Shalom,

Please post results of:

vgscan -a

SEP