Resurrecting old disk

[Kevin Eady_1]

Hi,
I've had to re-install HP-UX 11.11, as the OS disk crashed.
Unfortunately, all my backups are corrupt as the tape drive was also faulty.
My problem is that I have a second disk that I need to re-mount, but I can't seem to do it using sam as I'm going around in circles. i.e. everytime I seem to be getting close, it keeps saying that there are old LVM structures on the disk and all your data will be lost if you proceed.
Is there any easy way that I can just mount this disk without using sam ?

Any help would be appreciated,

Kevin Eady

[Victor Fridyev]

Hi,

Try to do the following:
mkdir /dev/vgrest
mknod /dev/vgrest c 64 0x070000
vgimport /dev/vgrest /dev/dsk/cxxxx
vgchange -a y vgrest
And after that try to mount all file systems.
Please use a unique minor number in the mknod command and replace cxxxx with real disk address

HTH

[Kevin Eady_1]

Thanks for you speedy reply.
Just one question though, how do I find out the real disk address ?
I tried to get the "device information" in sam, but it wouldn't let me have the info as I hadn't configured it, and when I try to configure it, it says the bit about losing all my information.
I also don't know what the original disk address was as I don't have a copy of the "fstab".

Kevin

[Victor Fridyev]

In order to find a disk device address or name, run
ioscan -fnCdisk

If the disk is internal, near it's slot you can find it's H/W address and, according to it from output of ioscam you'll see its device name

[YAQUB_1]

Hi,

>>> how do I find out the real disk address ?
==> ioscan -fnC disk
==> lvlnboot -v
==> setboot -v

[Kevin Eady_1]

Thanks, have got the disk address now, but tried to do as you suggest and get these errors now ....
# mkdir /dev/vgrest
# mknod /dev/vgrest c 64 0x070000
mknod: File exists
# vgimport /dev/vgrest /dev/dsk/c3t15d0
vgimport: "/dev/vgrest/group": not a character device.

any thoughts ?

[Andres_13]

Is because de minor number is not unique, try as follows:

1. # rm -r /dev/vgrest
2. # mkdir /dev/vgrest
3. # mknod /dev/vgrest c 64 0x0f0000
4. # vgimport /dev/vgrest /dev/dsk/c3t15d0

If you get any output messages from mknod command you must repeat steps 1-3 changing the minor number (the "0f" in 0x0f0000) by other not currently in use.

Regards!

[Kevin Eady_1]

Have tried loads of different minor numbers now and still get "file exists". Maybe something else is wrong ?
Have done a "vgscan" and it says that Physical Volume "/dev/dsk/c3t15d0" is not part of a Volume Group.
Can this be part of the problem ?

[Pete Randall]

What does 'll /dev/vgrest' show?


Pete

[Kevin Eady_1]

Just says "total 0"

[Torsten.]

I'm not really sure about your current status.

What is the result of

# vgscan -v -p

(this is preview mode)

[TTr]

There is a typo

> # mknod /dev/vgrest c 64 0x070000

> 3. # mknod /dev/vgrest c 64 0x0f0000


You left out the group file

# mknod /dev/vgrest/group c 64 0x070000

[Kevin Eady_1]

Yes, it give a few errors ...
{hp11} root> # vgscan -v -p
Error getting path name of volume group with group file minor number 0x10000.
Minor number of /dev/vg10/group is not unique. /dev/vg02/group has the same mino
r number.
vgscan: Couldn't access the list of physical volumes for volume group "/dev/vg01
".
vgscan: Couldn't access the list of physical volumes for volume group "/dev/evg0
5".
vgscan: Couldn't access the list of physical volumes for volume group "/dev/evg2
0".
vgscan: Couldn't access the list of physical volumes for volume group "/dev/evg5
0".
vgscan: Couldn't access the list of physical volumes for volume group "/dev/evg7
0".
vgscan: Couldn't access the list of physical volumes for volume group "/dev/evg6
0".
vgscan: The physical volume "/dev/dsk/c1t15d0" is already recorded in the "/etc/
lvmtab" file.
Physical Volume "/dev/dsk/c3t15d0" is not part of a Volume Group
/dev/vg00
/dev/dsk/c1t15d0


The Volume Group /dev/vg01 was not matched with any Physical Volumes.
The Volume Group /dev/evg05 was not matched with any Physical Volumes.
The Volume Group /dev/evg20 was not matched with any Physical Volumes.
The Volume Group /dev/evg50 was not matched with any Physical Volumes.
The Volume Group /dev/evg70 was not matched with any Physical Volumes.
The Volume Group /dev/evg60 was not matched with any Physical Volumes.

[Torsten.]

ok and now

# ll /dev/vg*/group
# strings /etc/lvmtab

[Kevin Eady_1]

OK, can now do the "mknod" when I include group.
But now get this ...
{hp11} root> # vgimport /dev/vgrest /dev/dsk/c3t15d0
Physical Volume "/dev/dsk/c3t15d0" is not part of a Volume Group
{hp11} root> # ll /dev/vg*/group
crw-r----- 1 root sys 64 0x000000 Apr 23 17:07 /dev/vg00/group
crw-r--r-- 1 root sys 64 0x030000 Feb 12 2005 /dev/vg01/group
cr--r--r-- 1 root sys 64 0x010000 Apr 30 16:45 /dev/vg02/group
crw-r--r-- 1 root sys 64 0x010000 Dec 15 14:24 /dev/vg10/group
crw-r--r-- 1 root sys 64 0x0f0000 May 8 16:31 /dev/vgrest/group
{hp11} root> # strings /etc/lvmtab
/dev/vg00
/dev/dsk/c1t15d0

[Torsten.]

there is a problem with vg02 and vg10.

because of your naming please do this too

# ll /dev/evg*/group

[Torsten.]

the reason is we need to find an unused group number first ... these numbers must be unique.

[Kevin Eady_1]

Yes, I think that that particular problem was caused because I had to use an ignite tape from another system to install the OS.

/dev/evg05:
total 0
crw-r--r-- 1 root sys 64 0x040000 Nov 7 2006 group
brw-r----- 1 root sys 64 0x040001 Nov 7 2006 lv_u05_oracledb
brw-r----- 1 root sys 64 0x040002 Nov 7 2006 lv_u05_oradata
crw-r----- 1 root sys 64 0x040001 Nov 7 2006 rlv_u05_oracledb
crw-r----- 1 root sys 64 0x040002 Nov 7 2006 rlv_u05_oradata

/dev/evg20:
total 0
crw-r--r-- 1 root sys 64 0x060000 Feb 12 2005 group
brw-r----- 1 root sys 64 0x060001 Feb 12 2005 lv_u20
crw-r----- 1 root sys 64 0x060001 Feb 12 2005 rlv_u20

/dev/evg50:
total 0
crw-r--r-- 1 root sys 64 0x070000 Feb 12 2005 group
brw-r----- 1 root sys 64 0x070001 Feb 12 2005 lv_u50
crw-r----- 1 root sys 64 0x070001 Feb 12 2005 rlv_u50

/dev/evg60:
total 0
crw-r--r-- 1 root sys 64 0x0a0000 Dec 15 14:26 group
brw-r----- 1 root sys 64 0x0a0001 Dec 15 14:26 lv_u60
crw-r----- 1 root sys 64 0x0a0001 Dec 15 14:26 rlv_u60

/dev/evg70:
total 0
crw-r--r-- 1 root sys 64 0x090000 Feb 12 2005 group
brw-r----- 1 root sys 64 0x090001 Feb 12 2005 lv_u70
crw-r----- 1 root sys 64 0x090001 Feb 12 2005 rlv_u70

[TTr]

It looks like all these volume groups are not in use. They came in with the ignite tape. Based on "strings /etc/lvmtab" all these /dev/vg* can be deleted. Please confirm this.

It looks like you only have two disks on this server, c1t15d0 is the boot disk and c3t15d0 is the disk you are trying to import.

This can be verified with
ioscan -fnkC disk
lvlnboot -v

> {hp11} root> # vgimport /dev/vgrest /dev/dsk/c3t15d0
Physical Volume "/dev/dsk/c3t15d0" is not part of a Volume Group

This should have worked. Unless the conflict of group minor nimbers prevented it. Did you possibly initialize (pvcreate) the second disk from within SAM?

Verify theat all these volume groups do not exist in this server and delete the /dev/vg??? directories except /dev/vg00 and /dev/vgrest and try the vgimport again.

[Kevin Eady_1]

Thanks for all your help, but I've decided that I've spent too much time on this and I'm just going to re-initialise the disk and start from scratch.

Kevin