boot failure hpux 10.20

[Aleksandar Vujevic]

After a reboot I get the following error message while loading hpux (isl gets loaded normally):

sysnc'ing disks

(some hardeware configuration display)

system panic
panic: (display==0xb800, flags==0x0) all VFS_MOUNTPOINTs failed: NEED DRIVERS???

As this looks like a corrupted file system, I tried to boot the core os cd without success (it says file system full).

What can I do now?

Aleks

[Steven E. Protter]

Looks like you lost a disk.

Most likely the root disk.

The failed disk needs to be replaced and then you should be able to restore your make_tape_recovery or make_net_recovery Ignite archive.

You should however be able to boot off the Core OS cd and then clear a full filesystem. The problem is that if you boot off core OS the root filesystem will not be the same as it would be if you booted off local disk.

I would make sure your disk connections and cables and power connections are solid, because its possible the problem was triggered there.

SEP

[Mel Burslan]

when booting from core os CD to a recovery shell, and this is when you are getting Disk is Full message, it is actually not the disk. Recovery shell tries to load itself into memory and it may be failing because it is not able to do so.
I do not want to be the bearer of the bad news but in case you are trying to run the recovery shell, it may be time to scratch everything out and start a reinstall of the OS. If you are getting this message (disk is full) again, I'd call HP hardware support and have the system examined.

[Suraj Singh_1]

Hope you have the backups with you.

If fsck fails, then you might have to recreate the filesystem on that particular LV, and later might want to extend it.

HTH,

Suraj

[Aleksandar Vujevic]

So I'm using "fsck -b 16" on the prompt for fsck.

Then the following message is displayed:

Alternate super block location: 16
/dev/rdsk/c0t6d0s1lvm
Phase 1 ...
...
Phase 5

everything ok, so far. Now the strange part:

Removing /sbin/fs/vxfs/fsck
Error stat(2;no such file or direcoty)
Cannot stat /sbin/fs/vxfs/fsck
/sbin/fs/hfs/fsck -y /dev/rdsk/c0t6d0s2lvm
BAD SUPERBLOCK ... USE -b OPTION
/sbin/fs/hfs/fsck -y /dev/rdsk/c0t6d0s2lvm failed

Why is vxfs/fsck being removed?
Why isn't the -b option passed along to the second volume?

Also trying a

fsck -b 16 /dev/rdsk/c0t6d0s2lvm

in the shell is not successful. It gives still a bad superblock.

Any hints are very appreciated!
Aleks

[Aleksandar Vujevic]

Sorry, got the copy & paste wrong: This should go before:

Ok, I got another recovery disk which will boot.

Runnig the recover shell I'm trying to do a

r - recover unbootable HPUX
b -don't rebuild bootlif, install files for boot & recovery of root fs

The device has logical volumes. Both volumes have a bad superblock.
So I'm using "fsck -b 16" on the prompt for fsck.

From now on see message above.

[Jino.P.V]

Hi
I think your root hard disk is having physical damage. Better you replace the Hard disk restore everything from ignite backup. I hope you have backup. Even you could fix it temporarily, it may create problem in future.

regards,

Jino

[Aleksandar Vujevic]

I could not fix the file system but thanks for your input anyway.

The recovery from a backup tape worked without problems. Good to have one...

Aleks

[Basheer_2]

Aleksandar,

I had a similar problem. This is what fixed

1) Shutdown the Server
2) Push all the scsi cabels properly to the disks
3) Then system came back