top argent, please advise system crash

[Norman_21]

Hi guys,

I would highly appreciate your prompt advice as the hp-ux K class k370 has just crashed and started rtebooting
by it self. This is happened before and HP had to replace the processor as the cause of this problem but I can see it is still happening. Any advice on how to find out the reason why this happen all the time.
Thanks in advance.

[Rajeev Shukla · Accepted solution]

the best is tell HP, send the crash dump to HP and let them have a look at it and they can tell what exactly has gone wrong.

Rajeev

[Steven E. Protter]

Search for documents on how to use the q4 utility. I have a document at work, but not home, you you need a guide on how to create a crash dump for HP to analyze.

Hardware or software can cause your problem.

[Steven E. Protter]

Search for documents on how to use the q4 utility. I have a document at work, but not home, if you need a guide on how to create a crash dump for HP to analyze.

Hardware or software can cause your problem.

[avsrini]

Hi AB,
Can you find any messages on syslog.log?.

Any how you can find the crash dumps in /var/adm/crash/crash.1 (
if this crash is not first, then crash.no).
there will be a core file,
some images.* file, copy of kernel and an INDEX file
with details of the node and core fils.
You can tar or fbackup this whole dir and send to hp.

Also pls update the link, once you solved the issue. So that it will help many people in future.

Regards
Srini.

[avsrini]

Hi AB,
You can download the HP WDB debugger program from software.hp.com.
This has an X- GUI also. It is very easy to work on.
You can give this a try, if you are interested. Also you can
use this debugger to find any core dump, created by programs in hpux. You have to
just select the core file and the program that caused the core dump. WDB will give you the possible cause for the core dump.

Hope you'll find this interesting
Srini.

[Michael Tully]

This problem is notorious on 'K' class servers and usually points to a backplane problem, if a CPU has already been replaced. Make sure that if you have replaced a CPU that the firmware may need to be updated.

[Norman_21]

Thanks alot guys,
very useful tips. However, the server is K class and Micheal is right, the CPU has been replaced by HP twice and the system still
crashed for the 3rd time and everyime the hp keep telling us
it is the CPU.

[Norman_21]

Micheal Tully,

you said"Make sure that if you have replaced a CPU that the firmware may need to be updated."
Could you please explain in more details on how to update the firmware.
Thanks in advance.

[Rajeev Shukla]

I still remember K class at one stage troubled a lot specially for CPU Problem. We got CPU replaced 5 times and then our system was stable. But i guess this thing should have been taken care by HP till now coz its a story 3 years old.
Ok now for the firmware version. When you interupt the boot process and go to PDC you can then type info to see the information and i dont remember but you can see the firmware version there.
The firmware upgrade requires HP engineers help. So you should insist him in keeping updated the firmware.

Rajeev