Sorry, I'm on the other side of the globe (= very different time zone) and actually on my summer vacation, so I won't answer to any work-like phone calls nor chat requests...
I'll answer what I can, whenever I can, but I won't spoil my long-awaited vacation over this.
Here's what I noticed from the startup output:
---------
/dev/vg01/oracle_9i :
vxfs fsck: sanity check: /dev/vg01/oracle_9i needs checking
vxfs fsck: fsck read failure bno = 4583504, off = 0, len = 2048
full file system check requir
/dev/vg01/lvolopt :
vxfs fsck: sanity check failed: read of super-block on /dev/vg01/lvolopt failed:
I/O error
cannot open device /dev/vg01/lvolopt
file system check failure, aborting ...
[...]
/dev/vg01/lvolusr :
vxfs fsck: sanity check failed: read of super-block on /dev/vg01/lvolusr failed:
I/O error
cannot open device /dev/vg01/lvolusr
file system check failure, aborting ...
----------
Your vg01 has apparently taken some major damage. Unfortunately, your /usr and /opt filesystems seem to be on the damaged portions of vg01, so this damage has seriously crippled the OS.
If /usr is not available, that explains the "INIT: Command is respawning too rapidly" error messages. The console login prompt is produced by /usr/sbin/getty, and that obviously cannot run if /usr is not there. That also would explain why most of the startup scripts are failing: they require /usr/bin/ksh or some other tools from /usr to execute successfully.
Basically, you have two ways to access the machine now: either get it to boot to single-user mode or start it up using the recovery mode of the installation CD. Both of these require a working console. Your vg00 seems to be fine, so the single-user boot should still provide you with a shell access.
You probably must TOC the server again to get it to boot to single-user mode: the current state seems like a dead-end. As there is no getty running on the console, the system is incapable of receiving any meaningful input from the console.
Exactly how did you get the console messages (in the message you posted at Aug 29, 2008 19:35:32 GMT)?
Does the server have a display and/or PS/2 keyboard connected? If it has, try using them instead of/in addition to Hyperterminal & serial cable: the system might prefer them over a serial console.
If that does not help, disconnect any PS/2 mouse/keyboard from the server, make sure the laptop is connected and Hyperterminal is running, and TOC the server again. Most PA-RISC machines with integrated keyboard connector (like the D-class) will automatically switch to accepting input from the serial console port if there is no keyboard directly attached to the machine at the time it's doing self tests.
After successfully getting to the single-user mode, your next problems will be:
- How to get replacement disks, suitable for your old D-class (to replace the failed disk(s) of vg01)
- If the LVs of your vg01 are mirrored, you may be able to recover some of all of the data by activating the vg01 volume group with "vgchange -q n -a y vg01". But you should prepare to restore from backups. (You *do* have backups, right?)
MK
MK
