How to remove the vg when there is not disk exists.

[Silver_1]

All,

I have a vg in the /etc/lvmtab and the corresponding disk doen't exists.

Also i get the vgdisplay with errors.

The vg has one logical volume and the bdf shows that it is mounted. I cannot unmount it.

I am getting errors related to the disk in the syslog.

I have tried re-naming the lvmtab and do vgscan etc....still i get errors

I also know that the reboot will fix the whole problem, but any idea how can i do without a reboot ?

Tx.

[Tim Nelson]

One way is to vgexport the offending empty non responsive volume group.

vgexport /dev/vgxyz

[RAC_1]

What are the errors?? No sidk in vgs?? Then how lvol under is there and mounted?? what does lvdisplay on that lvol say?

[Rick Garland]

Use the vgexport command.
This will remove the vg instance name from the system.

/usr/bin/vgexport

[Pedro Cirne]

Hi,

Post here:

#strings /etc/lvmtab
#vgdisplay -v

and the errors you're getting on syslog.

Let us know!

Pedro

[Marvin Strong]

after you moved the lvmtab, and ran vgscan, did you do a vgchange and activate the vg again?

[Silver_1]

All,

Here are the errors in the syslog. HP confirms me that these are related to the

Aug 22 09:52:14 prd01 vmunix: SCSI: 28 bp: 0000000071086060
Aug 22 09:52:14 prd01 vmunix: 00 dev: 1f1c5300 00 0
Aug 22 09:52:14 prd01 vmunix: 0 00 1c0db: 00 28 00 0200 00 00
Aug 22 09:52:14 prd01 vmunix: 00 00 residual: 400
Aug 22 09:52:14 prd01 vmunix: 10 00 status: (200) None -- Select Timeout
Aug 22 09:52:14 prd01 vmunix: 00 02 00
Aug 22 09:52:14 prd01 vmunix: residual: 400
Aug 22 09:52:14 prd01 vmunix: status: (200) None -- Select Timeout
Aug 22 09:52:19 prd01 vmunix:
Aug 22 09:52:19 prd01 vmunix: SCSI: bp: 000000007109a060
Aug 22 09:52:19 prd01 vmunix: dev: 1f085300
Aug 22 09:52:19 prd01 vmunix: cdb: 28 00 00 00 00 10 00 00 02 00
Aug 22 09:52:19 prd01 vmunix: residual: 400
Aug 22 09:52:19 prd01 vmunix: status: (200) None -- Select Timeout
Aug 22 09:52:19 prd01 vmunix:
Aug 22 09:52:19 prd01 vmunix: SCSI: bp: 0000000071086060
Aug 22 09:52:19 prd01 vmunix: dev: 1f1c5300
Aug 22 09:52:19 prd01 vmunix: cdb: 28 00 00 00 00 10 00 00 02 00
Aug 22 09:52:19 prd01 vmunix: residual: 400
Aug 22 09:52:19 prd01 vmunix: status: (200) None -- Select Timeout
Aug 22 09:52:24 prd01 vmunix:
Aug 22 09:52:24 prd01 vmunix: SCSI: bp: 000000007109a060
Aug 22 09:52:24 prd01 vmunix: SCSI: dev: 1f085300 bp: 0000000071086060
Aug 22 09:52:24 prd01 vmunix: dev:
Aug 22 09:52:24 prd01 vmunix: 1f1c 53c0d0b:
Aug 22 09:52:24 prd01 vmunix:
Aug 22 09:52:24 prd01 vmunix: 28 cdb: 00 28 00 00 00 00 00 00 10 00 10 00 00 0000 0 202 00 00
Aug 22 09:52:24 prd01 vmunix:
Aug 22 09:52:24 prd01 vmunix: residual: 400
Aug 22 09:52:24 prd01 vmunix: status: (200) None -- Select Timeout

Tx

[RAC_1]

ll -R /dev | grep -i 085300

What is this device. If disk ,check if you can see the disk in ioscan.
ioscan -fnCdisk
If in ioscan it is claimed, run dd, diskinfo on it to know what it is.

diskinfo -v /dev/rdsk/cxtxdx
dd if=/dev/rmt/cxtxdx of=/dev/null bs=8k

Anil

[Silver_1]

When i export it shows that it cannot query the associated disks.
When i try to vgchange the same thing.
If i try vgscan after deleting the lvmtab. The entry goes from the lvmtab but the above errors still comes on the system's syslog.

/dev/vg01/lvol1 14221312 4628256 9518176 33% /an01

[Silver_1]

diskinfo /dev/dsk/c8t5d3
diskinfo: can't open /dev/dsk/c8t5d3: No such file or directory

[saju_2]

Hi

Pls refer the below link...

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=885949

It seems that the system unable to find the disks but the device file for the disks in the VG is still there or in some unclaimed state.

Can u post ioscan -fnC disk, strings /etc/lvmtab , bdf output...

Regards
CS

[OldSchool]

greetings....


Take a look at this document from the knowledge base. I've used that procedure successfully in the past

http://www1.itrc.hp.com/service/cki/docDisplay.do?docLocale=en_US&docId=200000080049938

[Silver_1]

Rebooted the server and was able to export the VG.