[FAIL] IBM x3650 7979 Booting the Controller Kernel

width=580

I’ve been experiencing a weird issue with a decommissioned server. I have an IBM x3650 7979 server with a ServeRaid 8k-l controller and during boot it just hangs at Booting the Controller Kernel and will eventually error with unable to start controller.  This post will document things I’ve done to try to resolve this issue.

I have replaced the SR8k-l controller, removed all peripherals and internal unnecessary components as well as pulled the drives from the backplane. I actually swapped backplanes as well, to no avail.

No errors indicated on lightpath diagnostics.

width=580

Looking at the connector on the backplane for SAS, it seemed warped.  I rigged up a different Lenovo RS210 server and used that backplane for temporary testing, to no avail either.

Next step is to try Zero Mode flash:  ftp://ftp.software.ibm.com/systems/support/bp_system_x/x3650_zero-mode.pdf (Local PDF Copy: x3650_zero-mode)

width=580

I ended up downloading the ISO ibm_fw_aacraid_5.2.0-16002_anyos_32-64.iso and within it contains the Floppy Diskette images.  However, I’m extracting the ISO and copying to a USB drive to boot from that.  I do have some floppy disks and a USB floppy disk, but I want to try this method first.  It’s shitty because this server takes a solid 20 minutes to get through controller kernel boot.  I’ll try the jumper after I try booting from CD-ROM.

Floppy what?

Here are the diskette images for ServeRaid 8k, 8k-l, 8i, 8s.  I had to use Rawwrite .7 from http://www.chrysocome.net/downloads/rawwritewin-0.7.zip as the version supplied on the ISO under diskette/tools/ did not work on Windows 10:

I think it’s been a good solid 5 years since I’ve had to touch a floppy…

width=580

Logic Bomb

The cpfiles.bat script kept looping me when prompting for Disk 3 to hit enter, I would insert Disk 3 and it would then ask for Disk 2 because a file check was looking for a specific file on A: which was only on Disk 2… So I modified the cpfiles.bat to eliminate checking:

@ECHO OFF
cls
echo .
echo This program will update the firmware on all
echo IBM ServeRAID-8k controllers in the system
echo to firmware build 16002.
echo .
choice Do you want to continue.
if errorlevel 2 goto exit
if errorlevel 1 goto load
:load
echo .
echo Copying Image Files.
copy A:afu.exe > NUL
copy A:*.ufi > NUL
echo .
echo Insert diskette #2
pause
:retry2
cls
REM if not exist A:AS481002.ufi goto DiskError2
echo Copying Image Files.
copy A:*.ufi > NUL
cls
echo .
echo Insert diskette #3
pause
:retry3
cls
REM if not exist A:AS481003.ufi goto DiskError3
echo Copying Image Files.
copy A:*.ufi > NUL
cls
echo .
echo Insert diskette #4
pause
:retry4
cls
REM if not exist A:AS481004.ufi goto DiskError4
echo Copying Image Files.
copy A:*.ufi > NUL
cls
AFU UPDATE AS4810 /CALL
goto AllDone
:DiskError2
echo Image file A:AS481002.ufi missing!
echo Insert diskette #2
echo .
choice Do you want to retry.
if errorlevel 2 goto exit
if errorlevel 1 goto retry2
:DiskError3
echo Image file A:AS481003.ufi missing!
echo Insert diskette #3
echo .
choice Do you want to retry.
if errorlevel 2 goto exit
if errorlevel 1 goto retry3
:DiskError4
echo Image file A:AS481004.ufi missing!
echo Insert diskette #4
echo .
choice Do you want to retry.
if errorlevel 2 goto exit
if errorlevel 1 goto retry4
:exit
echo .
echo Program terminated by user, controllers have not been updated.
:AllDone

Finally some progress…

width=580

It’s dead, Jim.

Well, after flashing and attempting a boot again, it still doesn’t boot.  I’m throwing in the towel and will use this guy for parts now.

width=580

But wait, there’s more!

Follow up post on me replacing the system board

Published by

Rich

Just another IT guy.

Leave a Reply

Your email address will not be published. Required fields are marked *