N class won't boot / failure

[Dan Alexander]

Have an N4000 with the following System Alert:

SYSTEM NAME: test
DATE: 01/01/1970 TIME: 13:31:11
ALERT LEVEL: 15 = Fatal hardware or configuration problem prevents operation

REASON FOR ALERT
SOURCE: 1 = processor
SOURCE DETAIL: 1 = processor general SOURCE ID: 3
PROBLEM DETAIL: 0 = no problem detail

LEDs: RUN ATTENTION FAULT REMOTE POWER
FLASH FLASH OFF ON ON
LED State: Running non-OS code. Non-critical error detected.
Check Chassis and Console Logs for error messages.

0x000030F0110316B0 080000FF FF861400 - type 0 = Data Field Unused
0x580038F0110316B0 00004600 010D1F0B - type 11 = Timestamp 01/01/1970 13:31:11


===========

does anyone know what this really is?

[S.K. Chan]

Alert level 15, the highest there is and usually indicating some sort of hardware problem, in your case it sounds like processor to me. Do this .. (I'm not sure if you got as far as the BCH's main menu)..

Main Menu: Enter command or menu > in
...
Information Menu: Enter command > pr
..

What does it says ? Basically "pr" will print processor information. What about in GSP mode , have you double checked the error logs ?

GSP > sl

Enter "e" for error logs. The think is your timestamp is outdated, so that makes me wonder.

[Eugeny Brychkov]

Dan,
it's hard to tell you what it exactly means. Are there any other GSP log messages displayed before or after this one? Anything in /var/adm/syslog/syslog.log? What does server do when displays it - reboots, hangs, or continues to work?
Looking to LEDs status displayed by GSP server should be in PDC or ISL (really 'running non-OS code')
Eugeny

[Eugeny Brychkov]

Two more very useful GSP commands in addition to SK's:
SS - system state
PS - power subsystem state
There should be everything configured (SS) and nothing failed (PS)
Eugeny

[Dan Alexander]

Thanks for the replies.

To answer both of you, this is the FIRST message that comes up after a reset. After about ~30 self tests...

So I cannot even get to the BCH Main Menu. It just spits out a couple more of these alerts, and that's it! I suspected a processor problem too.

It's an 8 way box, and in fact it spits out 8 seperate alerts before it hangs indefinitely. This is a fairly new box, so I really doubt ALL 8 cpus are bad! I just wish there were a way to decode this to an exact problem, so I could fix and move on!

so if anyone can decode this, please do!

[S.K. Chan]

I need more .. after ..
PROBLEM DETAIL: 0 = no problem detail
.. there should be something like ..
REPORTING ENTITY TYPE : ..
REPORTING ENTITY ID : ..

Can you post that as well. Did you call HP to decode these ? They can easily gave you the answer as the chassis codes are highly dependent on your PDC revision.

[Dan Alexander]

Eugeney,

Both SS and PS show all OK.
Still stuck though!

Any other suggestions?

[Dan Alexander]

S.K.

Here is another entry in full :

SYSTEM NAME: testn2
DATE: 01/01/1970 TIME: 13:31:11
ALERT LEVEL: 15 = Fatal hardware or configuration problem prevents operation

SOURCE: 1 = processor
SOURCE DETAIL: 1 = processor general SOURCE ID: 5
PROBLEM DETAIL: 0 = no problem detail

CALLER ACTIVITY: 1 = test STATUS: 0
CALLER SUBACTIVITY: 6B = implementation dependent
REPORTING ENTITY TYPE: 0 = system firmware REPORTING ENTITY ID: 05

0x000050F0110516B0 080000FF FF861400 type 0 = Data Field Unused
0x580058F0110516B0 00004600 010D1F0B type 11 = Timestamp 01/01/1970 13:31:11
Type CR for next entry, - CR for previous entry, Q CR to quit.

[Bill Douglass]

I don't think there is much you can do with this problem. The fastest solution is to get HP onsite and do some hardware swapping.

[S.K. Chan]

I was hoping for a "obvious" description of the error code translation so that I can make an intelligent guess but it's not comprehensive to me. Anyway from the reporting type entity 0 (ie firmware) chassis codes table that I got (based on PDC rev 40.49) taking the last 4 digit of the hex code "16B0" and mathching it with the description I got ..
"CC_BOOT_UNEXPECTED_INTERRUPT". I do not know what that means exactly and this is as far as I can help. I do not have the documentation on how to proceed with that error. It's best you call HP (if it's a new box,it got warranty). I would like to know what's the outcome of this though. Thanks ..

[Jeff Schussele]

Hi Dan,

Last thing I would do before logging a HW call is completely pull all power from the PDUs to the system.
It's possible the GSP has becomed hosed by a faulty power monitor board. There have been a run of bad L & N-class systems that got out with bad PM boards that can cause all manner of false errors.
If this is the case you still need to get HP to replace it, but this may get you running & verify the true culprit.

Rgds,
Jeff

[Dan Alexander]

Again, thanks for the replies. I will try and pull all power, and reset everything. I already tried once, but did so very quickly. I will also reset the gsp. I will try and keep this updated as I continue to troubleshoot this.


Thanks all!

[Marlene Smith]

I highly recommend calling HP - assuming that this is a supported box. I would also recommend putting in and running the diagnostic cd - or wait until the CE gets there? so he can see what hardware bails, etc. best of luck!!

[Christian Morais]

Hi Dan,

One information more:

There is a bug with some N Class servers with 8 proc. Sometimes we need to add a new board called a Trim Card that corrects some problems in the system clock. Ask HP to know if your servers can be concerned with this problem. If yes, HP CE just have to add this card in a special slot on the system board and all will gonna be all right

Regards

Christian

[Dan Alexander]

Turned out that the system board was bad. It was replaced, and all is fine now..