How to remove a PV from ODM

Roman Kanala

unread,

Nov 25, 2000, 3:00:00 AM11/25/00

to

I have got an interesting problem: a Seagate Medalist disk stopped.
It was hdisk1 and it is still visible when doing lspv, but does belong
to any volume group. Also it does respond to any command, including
varyoff, rmlv, getlvodm etc. Diag says that the motor stopped.
It was in a proper volume group user2vg that I have removed.
VG rootvg is there and intact so that the machine is up and running.

My first idea was to remove hdisk1 from ODM and
use unallocated space on hdisk0 to re-create some vital filesystems
from backups, but I cannot remove mount point of filesystems
left on the crashed disk :
0516-306 getlvodm: Unable to find user2vg in the Device
Configuration Database
0516-912 rmlv. Unable to remove logical volume user2a.

I also have problems changing the status of the filesystems and
of paging space located on the crashed disk to have it forgotten
on the next startup:
0516-010 lqueryvg: Volume group must beb varied on; use varyonvg
command.
0516-010 lquerylv: Volume group must beb varied on; use varyonvg
command.
0516-704 chlv: Unable to change logical volume user2a.

Did anyone encounter such a situation ? What to do when
I cannot define filesystems with the same names on other
disks, cannot remove them nor making them forget on next
reboot ?

Thanks in advance

Roman Kanala
kan...@sc2a.unige.ch

Jan Muench

unread,

Nov 28, 2000, 3:00:00 AM11/28/00

to

Roman Kanala wrote:

hi roman,

at the first look this looks like that the vg was removed from the odm
before the pv.

to fix it you should take a look at this paper from IBM.

regards Jan

repairvgodm

Norman Levin

unread,

Nov 30, 2000, 3:00:00 AM11/30/00

to

Try
odmget -q 'name=user2vg' CuAt
if a stanza shows and you are convinced the vg is really gone:
odmdelete -q 'name=user2vg' -o CuAt
use at your own risk.
Norm

--
Norman Levin

Roman Kanala

unread,

Nov 30, 2000, 3:00:00 AM11/30/00

to

Jan Muench wrote in response to my previous query:

>at the first look this looks like that the vg was removed from the odm
>before the pv.

Yes exactly.

> to fix it you should take a look at this paper from IBM.

> Rebuilding a Volume Group's Customized Device Database

...

PV=/dev/ipldevice
VG=rootvg
lqueryvg -Lp $PV | awk '{ print $2 }' | while read LVname; do
odmdelete -q "name = $LVname" -o CuAt
odmdelete -q "name = $LVname" -o CuDv
odmdelete -q "value3 = $LVname" -o CuDvDr
odmdelete -q "dependency = $LVname" -o CuDep
done
odmdelete -q "name = $VG" -o CuAt
odmdelete -q "parent = $VG" -o CuDv
odmdelete -q "name = $VG" -o CuDv
odmdelete -q "name = $VG" -o CuDep
odmdelete -q "dependency = $VG" -o CuDep
if [ "$VG" = rootvg ]
then
odmdelete -q "value1 = 10" -o CuDvDr
else
odmdelete -q "value1 = $VG" -o CuDvDr
fi
odmdelete -q "value3 = $VG" -o CuDvDr
importvg -y $VG $PV # ignore lvaryoffvg errors
varyonvg $VG
synclvodm -v $VG
savebase

> [ TechDocs Ref: 90605223414650 Publish Date: Feb. 01, 2000
> 4FAX Ref: 2418 ]

Thank you very much, it worked. Without this it would be impossible
to get rid of lvg's in hdisk1 ODM records because it was dead.
Now I have a new hdisk1 and everything is running.

Roman Kanala