server went down?

[Shar Hunter]

Sorry, hit wrong button.


Our hp server went down this weekend. I came in this morning and it had the following message:

FLT CBFD
TRAPS cpu 012

I powered it off then back on, and it booted up all right.

The syslog started over. I am not sure if an old syslog is still in there or not.

Any ideas? Does anyone know what that error code is?

Thanks.

[James R. Ferguson]

Hi Troy:

You should have the prior 'syslog' as '/var/adm/syslog/OLDsyslog.log'.

Regards!

...JRF...

[Shar Hunter]

OK, thanks. I did not even look. Duh!

The old log is clean and just stops prior to the system crash. No evidence there.

But then it is only the syslog.

Anyone know about that error code?

FLT CBFD

it was on the LED display.

[Ken Hubnik_2]

Also check the /etc/shutdownlog and /var/adm/tombstones for a ts** file. Also, see if the attention light is on and if so read to system logs.

[MANOJ SRIVASTAVA]

Also , you can check for the shutdownlog under /etc , the same will ahve some reason . Then you may like to see the OLDsyslog under /var/adm/syslog to know what went wrong.


It looks some cpu's failing.


Manoj Srivastava

[Oleg Zieaev_1 · Accepted solution]

Ken.

You may consider checking /var/adm/crash directory for crash dump like crash.1.
Inside crash.1 check INDEX file. The reason for crash should be in the file.

From what you have described looks like CPU fault. You can decode crash files using Q4. Or ask HP (in case you have support contract) to do it. This will give you the answer.

Hope this helps.
0leg

[Shar Hunter]

OK, I will call HP and put in call.

It does look like a CPU is on the fritz.

Thanks for the help.

[John Bolene]

Any CBF FLT is a nested HPMC and needs HP to look at it

You should have done a TOC at the time to get a dump

Term list follows

HPMC - high priority machine check