[aixpremkumar]

i have one vg karvg in two pv's (hdisk0,hdisk2)

i have the following output

# lspv hdisk0
0516-072 : Bad-block directory is corrupted. The physical
volume data has been permanently damaged. Run diagnostics.
PHYSICAL VOLUME: hdisk0 VOLUME GROUP: karvg
PV IDENTIFIER: 000744476c2b6f37 VG IDENTIFIER 0007444700004c00000001156
c329e6d
PV STATE: ???????
STALE PARTITIONS: ??????? ALLOCATABLE: ???????
PP SIZE: ??????? LOGICAL VOLUMES: ???????
TOTAL PPs: ??????? VG DESCRIPTORS: ???????
FREE PPs: ??????? HOT SPARE: ???????
USED PPs: ??????? MAX REQUEST: 256 kilobytes
FREE DISTRIBUTION: ???????
USED DISTRIBUTION: ???????

# lsvg -l karvg
0516-072 : Bad-block directory is corrupted. The physical
volume data has been permanently damaged. Run diagnostics.

i thinck VGDA is corrupted.

how can i run diagnostic.

[Jorke]

did you try the command diag ?
Silly question, but do you have a backup ?
what does errpt say ?

It seems to me your disk is experiencing problems .. if this is a hardware error you should replace the disk.
if your disk is dead, but not defective and you have mirrored your volumegroup over two disks you need to reducevg karvg hdisk0. then you can rmdev -dl your disk, cfgmgr , format disk, readd disk to volumegroup and voila.
then you should be able to view your vg again.

If you have not mirrored your vg, don't reducevg your vg .. then you loose your data (probably)

succes.

JK

[cccompton]

Just to get more info, try checking one of the logical volumes on the disk:

# fsck /dev/fslv00

If you get output like following it doesn't look good:

****************
The current volume is: /dev/fslv00
fsck: 0507-018 Superblock is corrupt and cannot be repaired.
Both primary and secondary copies are corrupt. Cannot continue.
fsck: 0506-042 Execute module "/sbin/helpers/jfs2/fsck64" failed.

You can also check each superblock as follows:

# lquerypv -h /dev/fslv00 8000 100 # try primary superblock
# lquerypv -h /dev/fslv00 F000 100 # try secondary superblock

A healthy disk will have output something like the following:

00008000 4A324653 00000001 00000000 43067023 |J2FS........C.p#|
00008010 00000000 00000000 00000000 00006461 |..............da|
00008020 74613237 00000000 00000000 00000000 |ta27............|
00008030 00000000 00000000 00000000 017FFB68 |...............h|
00008040 00001000 0000000C 00000200 00000009 |................|
00008050 00000003 00008000 00020100 00000001 |................|

If there is no data the disk is corrupted and you will need to recover using your well-documented and tested recovery procedures.

You should probably get IBM support to walk you through this if you value the data on the system.

[nomad]

is this the same steps for a rootvg's hdisk that is dead ?