NetFinity 5000 POST problem

[pmcwga]

System: IBM Netfinity dual 450's raid 512 ram OS win 2k adv server

Overview: system has been working flawless until something goes bump in the middle of the night...!

Problem: Noticed erro led (Exclamation mark one) nothing else out of the ordinary. Rebooted system, started off with several POST errors, 062, three consecutive boot fails, 162 confg has changed and 164 memory.

I pulled the cover off and looked at the system board leds, only one on was NMI error. reseated the memory, may have cleared the 164 out.

Reviewed the IBM error codes etc and tried various corrective measures. replaced battery, pulled out cards etc but to no avail. Attempted BIOS recovery via switch 5...
pulled processors etc
During POST, system boots (no beep) PS lights are ok check memory ok, continiously receive error 162 config has changed, then goes to a NMI error ocurred at, if left alone, the system will reboot and repeat the errors. indicated see system log etc..

I'm at a total loss as to where to go/look at now as can't get into the setup via F1 or anthing. Somewhere I remeber that one can access the logs via null modem but?? don't think bios cmos can be cleared out on this system? Would love to get into the system log as well to help pinpoint..

Would appreciate any direction on what to try...

thanks

 

[catorze]

To clear the bang <!> light on the front panel, and since you had 3 consecutive boot failures, you should enter CMOS <F1> and Reload Default Settings (to re-enable CPU caches - the system automatically disabled them after 3 boot failures), clear the System Error Log, then shutdown the system. Disconnect power for 30 seconds, then reconnect power (this resets the built in Advanced Systems Management Processor).

 

[pmcwga]

Thanks for getting back so quickly

Still no go... On bootup, memory countsdown ok etc, press F1 here's what comes up

F1 has been pressed. Booting to setup
162 configuration error
Searching for Vital Product Data.... Done

(Screen comes up indicating raid...)nothing else then

see the system error log. Unexpected NMI at: F000:76AD

just can't open the setup... to clear this/look at errors etc.

where do I look now?

 

[rclarke250]

Will F2 allow you to go into Diags? Can't remember if it is on this machine type or not, If it does you can view and clear error logs from there. But it sounds like the system board is hosed.

I shall use google before asking stupid questions!

 

[pmcwga]

F2 is on this for diagnostics, it will not let one into this as well... even changed out keyboards to try but to no avail. hope sys board not gone, but??

thanks for the insight... will keep trying...

 

[margudmju]

My guess is that your system board is dead. NMI means "Non Maskable Interupt" and usually occurs either if:

1. You just made changes to the hardware configuration and they are very wrong (ex: wrong CPU)
2. The system board is dead.

 

[catorze]

Non-Maskable Interrupts usually point to memory errors. If the installed RAM is not IBM, or is not Registered memory, remove it and try again. If it is, confirm in the Hardware Maintenance Manual the recommended configuration, then run diagnostics <F2> on the memory (this will take a while, so plan accordingly).

 

[tbwus]

If your using raid, and have a scsi card in the machine for a tape backup, try removing the scsi card. I had a similiar situation with a ibm 5500.

 

[catorze]

If you're using a RAID card, it should be installed in Slot 5 (so POST hits the RAID card before the onboard SCSI), and the SCSI BIOS should be disabled (BIOS is only required if booting to the chipset) via Ctrl+A > Advanced settings. What's the story on the memory?