OSR5.0.2 won't boot after CPU replacement

OSR5.0.2 won't boot after CPU replacement

Post by Steven L. Ste » Sun, 08 Feb 1998 04:00:00



Yesterday afternoon, we attempted to replace the 2nd CPU (which was
suspected to be causing some recent PANICs) in our Compaq ProLiant 5000.
After a normal Unix "shutdown -g0 -y" and power-off, the old CPU2 chip was
removed and the new one installed.  When attempting to reboot, the message
"unsupported processor" appeared and the system halted.  We then found
that the CPU kit contained a new ROMPaq diskette (Ver. 3.14, 09/25/1997)
and instructions.  We cold-booted the system with the ROMPaq diskette in
drive A and allowed the firmware upgrade to complete.

When we then rebooted, we got the following message:

%tape ....
G
WARNING: hd : no root disk controller
H iinit
PANIC: srmountfun - Error 19 mounting rootdev (1/42)
Unable to freeze processor 2, proceeding.
Error 19 opening dumpdev (1/41)
Dump not completed
** Safe to Power Off **
          or
** Press any key to Reboot **

Looking at SCO website, I found a couple of articles pertaining to this
problem, and the recommended solution was to run the EISA Configuration
Utility.  So, we did this, but every time we attempt to reboot, we get the
same error.

I called Compaq Tech Support and spoke with two technicians.  The first
one had me clear non-volatile RAM (using Compaq Configuration Utility),
save configuration, and then reboot, run Compaq Configuration Utility
again, and configure (manually).  They had me change some IRQ settings.
Rebooting again resulted in same errors. At his suggestion, I ran Compaq
Diagnostics on CPU, RAM, and Drive Array.  No problems were detected.

They then had me remove 128mb RAM (which had been installed a week ago
without problem) and try to reboot, run EISA Config, and boot SCO.  Same
result.  We also tried removing the 2nd CPU and repeating reboot, EISA
Config, boot SCO, with the same result.  And, we tried with only CPU1,
only CPU2 (in the CPU1 position), and only CPU1 with the new processor.
All with the same result.

The second technician told me the next step is to replace the System I/O
Board.  So, we're ordering one.  But, I think the problem is within SCO or
my EISA configuration.

I've tried reloading a different ROMPaq, in case the diskette that came
with the CPU was causing the problem (making diskette from Compaq CD),
running EISA Config, and rebooting SCO.  Same result.

I'm desperately seeking help.  We'll be reseating all the boards this
morning.  If that doesn't solve problem, I'll attempt to boot from
Emergency Boot/Root Diskettes.  But, I suspect one of you has a clue to
help me resolve this problem.

Our system:

uname -X follows:

System = SCO_SV
Node = cbscmpq2
Release = 3.2v5.0.2
KernelID = 96/01/23
Machine = Pentium
BusType = EISA
Serial = sco009767
Users = 512-user
OEM# = 0
Origin# = 1
NumCPU = 2

Hardware is Compaq Proliant 5000, 2-200mhz pentium processors (w/512k
cache), 256mb RAM,
Smart Array (7 4gb SCSI drives, 6 of which are mirrored, 7th is hot
spare.), Wangdat 4mm DAT, cd-rom, 3.5" 1.44mb floppy

--
Steven L. Stein
Cornell Business Services

 
 
 

OSR5.0.2 won't boot after CPU replacement

Post by Bela Lubki » Fri, 13 Feb 1998 04:00:00



> Yesterday afternoon, we attempted to replace the 2nd CPU (which was
> suspected to be causing some recent PANICs) in our Compaq ProLiant 5000.
> After a normal Unix "shutdown -g0 -y" and power-off, the old CPU2 chip was
> removed and the new one installed.  When attempting to reboot, the message
> "unsupported processor" appeared and the system halted.  We then found
> that the CPU kit contained a new ROMPaq diskette (Ver. 3.14, 09/25/1997)
> and instructions.  We cold-booted the system with the ROMPaq diskette in
> drive A and allowed the firmware upgrade to complete.

> When we then rebooted, we got the following message:

> %tape ....
> G
> WARNING: hd : no root disk controller

That's crucial.  It isn't recognizing your host adapter.

Quote:> Hardware is Compaq Proliant 5000, 2-200mhz pentium processors (w/512k
> cache), 256mb RAM,
> Smart Array (7 4gb SCSI drives, 6 of which are mirrored, 7th is hot
> spare.), Wangdat 4mm DAT, cd-rom, 3.5" 1.44mb floppy

"Smart Array" is a peripheral device, right?  What's the host adapter?

Concentrate on getting the host adapter and its driver to the same
settings.  You might get some milage out of an
"adapter=fooha(addr,irq,dma)" bootstring.

Were any of the boards moved around?  Some slots are more equal than
others -- you might have put the host adapter into a non-bus-mastering
slot or something like that.  *Order* of boards in slots can also make a
difference in some cases.

If there are any expendable ISA boards in the machine, remove them.
They can cause IRQ conflicts that may not be visible, but cause
problems.  Put them back in later.

Quote:>Bela<

--
Back a bit early from our world travels.  Vietnam in a few months (we hope).
Travelogue http://www.armory.com/~alexia/trip/trip.html still being updated.