We have an RS/6000 model 7025 F40 and it won't boot.

It is reporting the error F8E 2BA00022  on the LCD window.
 

1) I can't find anywhere what F8E refers to ... Do you know the answer?
I know that 2BA00022 means: 
>Service processor reports bad CRC 
>error-CMOS/NVRAM

I have replaced the power supply, the processor, memory, and the system board.  
I'm stuck at this point.
 
I have even disconnected the SCSI chain completely ...
 
Thanks for your help in advance
 
Larry McGhaw
CONNX Solutions
 
lm

---

Request for Question Clarification by missy-ga on 04 Jun 2003 22:24 PDT

Hello Larry,

I've searched for several hours for a solution to your problem, with
little success.

Searching on the error message you've provided yielded several posts
on Usenet looking for help with a problem similar to yours:

"I have an old RS/600 model 7025 F40 that is failing to boot,
reporting
an error 2BA00022 F8E. [...]"

"RS/6000 7025 F8E failing with 2BA00022" - comp.unix.aix
http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=utf-8&th=737b69be81e68d71&rnum=1

"I have a 7025-F40 system with a F8E error on the LCD pannel. Also has
error 2BA00022[...]"

"Help! F8E error on F40" -  fj.sys.rs6000
http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&oe=utf-8&th=dc9138bf46f901c7&rnum=2

In both instances, it was suggested that the problem is one of memory
failure.  Pierre Wes-Nicolett in fj.sys.rs6000 suggests booting the
machine with only one memory card.

Dan in comp.unix.aix suggests draining the NVRAM first.

Further information is quite scarce.  I have been unable to find
instructions for draining the NVRAM, and the IBM manuals for the
RS/600:

RS/6000 7025 F40 Series IBM User's Guide
http://www-1.ibm.com/servers/eserver/pseries/library/hardware_docs/sa38/380514.pdf

RS/6000 7025 F40 Series IBM Service Guide
http://www-1.ibm.com/servers/eserver/pseries/library/hardware_docs/sa38/380515.pdf

...have been of little assistance.

You may want to call in the pros for further assistance:

Donohue's RS/600® Help Service Page   
http://www.s6000.com/help.html

I'm sorry I couldn't be of more help to you.

--Missy

---

Clarification of Question by larrymcghaw-ga on 05 Jun 2003 03:56 PDT

Thanks Missy for trying.

I was aware of the prior USENET posts, but I was also unable to find a
resolution to the problem.  I have already tried draining the NVRAM
(on both the original, and the replacement system board), and I have
also purchased a pair of replacement memory DIMMs, but the results are
the same.

---

Clarification of Question by larrymcghaw-ga on 25 Jun 2003 11:58 PDT

I gave up, and had an IBM Support Rep come out to figure out the
problem.
He was able to fix it in about 2 hours.

For any poor soul who may encounter this problem again ... here is the
problem, and the solution:

As it turns out, The original problem was a bad power supply.

However, the bad power supply damaged the system board, so now two
components had failed.

So we replaced ... in order, 
the memory --- no effect
the system board --- since the power supply was defective, as soon as
we put in a new system board, the bad power supply caused the new
system board to fail as well, ... so we now had two damaged system
boards.

Then we replaced the power supply --- however, we now had two damanged
system boards, so the system still did not boot.

The solution was to replace the power supply, and the system board
simultaneously.  The system booted perfectly!!!

Thanks for everyone help along the way.

Free advice is worth what you pay for it, but "former RS6K admin, not
a ga researcher, speaking only for myself and strictly on an as is
basis":

F8E means "bad value in EPOW register." Very unhelpful, huh?

"F8E 2BA00022," despite the NVRAM indications, _usually_ means "Bad
power supply." Yes, the nvram contents got corrupted, but that's a
symptom, not the cause. If all else was right it should clear, though
it might require a second reboot. When you say you've replaced the
power supply, I assume you're dead certain that you replaced it with a
good one, and not one out of a spare parts pile, and not just one of
two.

When it's not power supply, it's usually sysplanar. Of course, you say
you've replaced that as well.

That puts you into "weird" territory.

Because it's simple, I'd first check connector J17 to make sure you
have no pushed in/faulty pins.

If not, then best guess is: Memory, processor and NVRAM are not the
issue. I'd first try replacing the OPS panel if you didn't do that
when you replaced the sysplanar. Then, more of a long shot, the SCSI
backplane.

rickt-ga

A question for you  ... Would it be possible for us to boot without
the ops panel connected? (using the service processor and an ASCII
terminal).

Or is the ops panel required for boot.

In other words, can we eliminate the ops panel as the problem without
getting a replacement.


Additionally, if we don't have a SCSI cable plugged into the
sysplanar, we should still be able to boot without this error. 
Correct?

Thanks

lm

I wish I could be more useful -- this only happened to me once, and I
only remember it because it dragged for long enough to go into my
notes to self file. (We'd gotten a replacement power supply that was
bad from stock, the only tinme that ever happened to me.)

So, although I recall very well that the third component that got
swapped out was the OPS panel, it was under service at the time, and
we never had to think about trying to boot it without the panel. I've
never done that, and really have no clue what would happen if it were
tried. I was the OS/network admin, not the hardware guy.

I took the SCSI backplane out of mine because the CE wanted it out.
Sorry that I don't recall if he was really suspecting an electrical
problem or if it was just because of his support people questioning
again and again "do you have it to a true minimum config?"