HPMC errors on newly acquired rp7410

[Charbel Bou-eid]

Hello all,

We have acquired a new HP rp7410 with 2x PA8700+ (875MHz) CPUs and 2GB of RAM. During the first power-up, we got the Fault & Attention lights flashing and the server rebooted. After checking, we found that it is always haging just after it boots from the primary disk and it starts displaying the memory address and the offset for the kernel, and after that it dumps to the HDD and reboots again.

We went to MP, and we checked the SL. I have attached the output of the SL.

Can somebody help please? Do you think it is a hard or soft error?

Thanks in advance for your help.
Charbel

[Charbel Bou-eid]

Hello guys,

I need your help please.

New develpoment.
1- I decided to install new OS on the second disk, everyhting went OK, until I installed the GoldBase11i Patches (June 04), the server hanged on the next reboot. I rebooted on vmunix.prev: OK, then I installed MirrorDisk/UX, and on the reboot, it hanged again. Again rebooting on vmunix.prev:OK. After that, it hanged again on the next provoqued reboot.

2- I decided to re-install and discovered that server hanged this time at the Kernel loading from the OS Install CD. I'm doubting now an intermittent H/W Fault on the Cell Board level.

P.S. Do you think that having 2x power cables instead of 4 can lead to such fault?

I have attached the HPMC I got from the "BCH>ser> pim hpmc" & "BCH>ser> el" to be used with the "HPMC Analyser Tool" if anubody has it. By the way, does anybody knows from where I can download it?

Thanks in advance for your urgent help.

Cheers,
Charbel

[Jeff Schussele · Accepted solution]

Hi Charbel,

This is my *biggest* complaint about these new cell-based systems.
Without the "Diag/Maint" manuals how are we "mortals" supposed to interpret these msgs?
They won't release these manuals *until* they quit selling them.
IMO it's probably a cell problem or possibly an I/O cage issue.
Bottom-line - IF it's a new system then it's under warranty which means it's on HP's $ to come out to rectify it.
Use your power & call them on it & make them perform.

My 2 cents,
Jeff

[Joerg-Udo Werner]

Hello Charbel,
to analyse your problem more data is needed.
1. Logon to the MP and dump the complete Error logs and Activity logs to file.
2. Read the complete HPMC information at the BCH prompt with the following commands:
"pim hpmc" (you did that already) and "el" (Errorlogs).

If you only get a hang condition without a system panic or HPMC crash you someone needs to analyse the dump after you did a "tc" from the MP:CM>.

Regards,
Udo

[Charbel Bou-eid]

Hi Jeff,
You're right. It was actually a PCI Card Cage problem. We have replaced the cage with a new one and it seems the server is working fine. Anyway, we are still testing it to ensure that this was the fault.
Thanks,
Charbel