Restoring vg configuration..

[Severus]

Hi, we are having problem in booting up the server. Most probably it had panicked and shows the following error while trying to boot-up:

Boot
: disk(0/1/1/0.1.0.0.0.0.0;0)/stand/vmunix
12644352 + 2932736 + 5682088 start 0x1f67e8

alloc_pdc_pages: Relocating PDC from 0xfffffff0f0c00000 to 0x3f900000.
gate64: sysvec_vaddr = 0xc0002000 for 2 pages
NOTICE: autofs_link(): File system was registered at index 3.
NOTICE: cachefs_link(): File system was registered at index 5.
NOTICE: nfs3_link(): File system was registered at index 6.
td: claimed Tachyon XL2 Fibre Channel Mass Storage card at 0/2/1/0
td: claimed Tachyon XL2 Fibre Channel Mass Storage card at 0/5/2/0

System Console is on the Built-In Serial Interface
igelan0: INITIALIZING HP A6825-60101 PCI 1000Base-T Adapter at hardware path 0/1/2/0
igelan1: INITIALIZING HP A6825-60101 PCI 1000Base-T Adapter at hardware path 0/6/1/0
Logical volume 64, 0x4 configured as ROOT
DIAGNOSTIC SYSTEM WARNING:
The diagnostic logging facility has started receiving excessive
errors from the I/O subsystem. I/O error entries will be lost
until the cause of the excessive I/O logging is corrected.
If the diaglogd daemon is not active, use the Daemon Startup command
in stm to start it.
If the diaglogd daemon is active, use the logtool utility in stm
to determine which I/O subsystem is logging excessive errors.

Logical volume 64, 0x2 configured as SWAP
Logical volume 64, 0x2 configured as DUMP
Logical volume 64, 0x3 configured as DUMP
Swap device table: (start & size given in 512-byte blocks)
entry 0 - major is 64, minor is 0x2; start = 0, size = 8388608
Starting the STREAMS daemons-phase 1
Checking root file system.

We tried several ways to boot (hpux –is, hpux –lm, boot alt, boot /stand/vmunix.prev) but all had the same problem. HP hardware support team has confirmed that there is no hw level problem and suggested that the problem is with OS.

We have an ignite backup of several days back. Currently the system has some changes in VG and LV level of some of the VGs belongs to SAN storage (vg00 unchanged), from that backup point. We have copies of vgcfgbackup of all the VGs in home directory (which is again in the vg00).

OS version: HPUX 11i ver 1

How can we restore VG configuration? as the ignite backup does not have the current lvmtab file.

Is there any option to mount the lvols of vg00 (before restoring any data) after booting from ignite backup?

Thanks.

[F Verschuren · Accepted solution]

reading the Error I do not think you can boot from ignite, it realy sounts like a hartware problem,

Last time I saw this error the v-par software was to old and one CPU was not able to do I/O

is this a v/n par? mayby you have simular problems (HP can not always determin if it is hartware)

ps you vcan boot a recovery from Ignite, and you have offcause always the mirror disk that ignite will not overwrite.

If you have a dvd player try to boot from that...

[Khairy]

hi severus,

looking at the log above, the server boot /stand/vmunix and that usually indicate you're booting up a non-vpar kernel.

It seems the server managed to load the kernel and found the disk devices, unfortunately there were errors with the IO subsystem. It may indicate a bad cabling or vg00 disk is failing. But i prefer to see more output b4 i could come to that conclusion.

If not, then it seems u hv to recover from ignite backup.

Ignite will detect the changes and abort automatic recover and go to interactive mode. Use interactive mode and proceed to recover the system.

When it finish, boot to single user mode and check for fstab and lvmtab entries. Backup them if you require.

Perform vgscan to detect new disks from the san storage (as you explained earlier this ignite backup does not contain latest vg info).

It will tell you if there are changes with the vgs. If its ok, proceed to vgimport the san storage disk devices.

Let us know the progress.

Rgds

[likid0]

How can we restore VG configuration?

When restoring from the ignite image, by default there is an option to import the vgs,so by default it will try to import your Vgs, so you should have no problem, if it doesn't do it correclty, you will need to have an vgexport map with the -s , or a cfg2html ouyput, or you need to know which vg had each disk.

so if you have a vg map with the -s, it has the vgid on the map(you should have one in your server /etc/lvmconf, called yourvg.mapfile), and it will import the vg with all the disks that have that id, you should do:


vgimport -s -v -m /tmp/your-vgmap-with-s /dev/yourvg
vgchange -a y /dev/yourvg


if you dont have a -s export a cfg2html, in the lvm-tab section, you have the vgid of your server:

vg_id:B8EF71ED3F7168EE vg_state:0000 vg_incno:00000000 pv_cnt:00000002

[Leonid Mishkind]

Hi.
When you boot from ignite the vg00 lvols are been rebuilt.
After you boot from ignite you can try and run vgscan .
Do you have a backup of /etc/lvmconf directory?
If so , you can use vgcfgrestore to recreate the volumes.

[Torsten.]

IMHO this is either a rp4440, rp3440 or an older L-class.
All of them may have more than 1 internal disk on the same bus. The I/O errors could be due to multiple and very frequent scsi resets cause by a bad disk - why not?
In this case it does no matter if you boot from the primary or alternate disk - because the bus is the same.
From this information I would not eliminate a hardware problem as a root cause.
You can try to boot from the offline diagnostics CD or the install CD to see if you get the ODE or kernel loaded or remove 1 disk and boot from the remaining.

[Severus]

Hello,

Thanks to all who helped me. Our problem was solved. We did "hpux -is -lq" from ISL and it booted in the single user mode. After that when we go to run level 2 it did a fsck and booted successfully and the problem was gone.

A bit fishy!

Thanks.

[Severus]

from ISL prompt "hpux -is -lq" let it boot in the single user mode.