SUMMARY E4000_Board_Failure

Hi,

Original question:
My E4000 with 3 boards (6 Cpus 250MHZ 1 Gig memory)  running Solaris 8
has suddenly rebooted and disabled a board without many messages to find
that's wrong. Only:
Date Hostname unix: [ID 796976 kern.notice] System booting after fatal error FATAL
...
prtdiag shows that the board is disabled with messages:
Analysis for Board 3
--------------------
AC: UPA Port A Dtag Parity Error
        The error could be caused by:
                Data Tags for UPA Port A
                Address Controller
Question: 
How can I found what is wrong in this board ( cpu , memory , bus ,... )?

Thanks to:
Steven Haywood 
Dagobert Michelsen
krenzischek@Encompasserve.org
Luigi Marra 
Pablo Panchig
Tony Walsh 

Most people suggested to boot the machine in diag mode 
(setenv diag-switch? true)  or to put the key on diag mode  ie:
(turn the key 2 clicks clockwise)
Nothing more.
Some said the board is really out because the FATAL error means the system was
unable to determine the exact cause of failure.
Tony Walsh  suggests that my   KJP (Kernel Jumbo Patch) level is to downlevel and 
suggest to apply  108528-29 
After loaded 108528-29  and reboot nothing more.

So no really solutions. Now i'm trying to remove some parts (memory, .. ) and
reboot to see if some changes will occur, also  trying to put the board in another
location. This will take sometimes because the machine runs as NIS and license 
server so it's difficult to stop it.
Anyway thanks to all peole.
By
Rene Occelli
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers