1833625 Members
3423 Online
110062 Solutions
New Discussion

vgimport

 
Darren Murray_1
Frequent Advisor

vgimport

A system has had its disk array moved from one machine to another but the same os loaded

The hardware paths for the disks have changed so /etc/lvmtab still contains the old hardware paths. ie /dev/dsk/c6t0d0, /dev/dsk/c6t0d1, /dev/dsk/c6t0d1

vgexport -m /tmp/vg01.map /dev/vg01 was run


How when they have tried to do an import by running
mkdir /dev/vg01
mknod /dev/vg01/group c 64 0x010000

vgimport -m /tmp/vg01.map /dev/vg01
/dev/dsk/c8t0d0 /dev/dsk/c8t0d1 /dev/dsk/c8t0d2

is being run to import the disks with the new path the following error occurs

vgimport: Physical volume "/dev/dsk/c8t0d0" is not a block special file.
vgimport: Physical volume "/dev/dsk/c8t0d1" is not a block special file.
vgimport: Physical volume "/dev/dsk/c8t1d2" is not a block special file.
6 REPLIES 6
TwoProc
Honored Contributor

Re: vgimport

You're going to need to run vgscan to get more out of this... see man vgscan
We are the people our parents warned us about --Jimmy Buffett
Jim Mallett
Honored Contributor

Re: vgimport

Check your /dev/dsk folder, are the devices actually listed there? (c8t0d0, etc) You would get that error if the devices didn't exist.

It's probably too late but if you had done a vgexport on the old machine with the -s option, finding the new disk would be much easier.

Run ioscan -fnC disk to see if you can see the disk. If not, try an insf -e

The problem is the mapping may not end up being as clean as c6t0d0 --> c8t0d0 on the new machine.

Jim
Hindsight is 20/20
TwoProc
Honored Contributor

Re: vgimport

Oh, and after vgscan'ing it back in , you're going to need to run vgchange.
We are the people our parents warned us about --Jimmy Buffett
HGN
Honored Contributor
harry d brown jr
Honored Contributor

Re: vgimport


ioscan -fn -C disk

shows what?

live free or die
harry d brown jr
Live Free or Die
A. Clay Stephenson
Acclaimed Contributor

Re: vgimport

First make absolutely certain that your new /dev/vg01/group device node has a unique minor device number. 0x010000 might already be in use by another VG and chaos will result. You probably need to do a new ioscan -fn and then an insf so that the disk device nodes get created. What does ls -l /dev/dsk/c8* display?
If it ain't broke, I can fix that.