Is there a general problem known for NW on IBM HS21-XM or with the EMulex 1105-BCv

we have one NW6.5.8 server running on a blade server with identical hw configuration.

This is a BladeCenter H with currently 7 identical blades for different tasks,
running sles 10, Win2k3R2 and up to now NW6.5.8 on *ONE* of the blades.

HS21-XM (FW 1.12 and 1.15 tried)
2x Quad Xeon
16 Gig RAM
Emulex 1105-BCv FC-HBA, Boot Code BB3.02.a6, Firmware BS2.70A5.

The 1st NW6.5.8 OS installation passed smoothly without problems.
With the reboot the server hangs just after loading ACPIDRV.PSM. (not just the
keyboard gone as posted in another thread. For that trouble with another blade, same
blade center the hdetect -MIR did help)

So I "hard powered off" the server to start another installation from the scratch
with deleting DOS and NetWare partition. This passes ACPIDRV, SCSIHD, BX2, then also

This time I was able to switch to the server's console screen prior the hang (so the
keyboard wasn't "gone", that is noticable from the status LEDs locally connected to
the Blade Center, they "run", if the blade has no keyboard connection):

There I can see 6 times this message, telling exactly the same time:

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~
9-21-2009 2:42:00 pm: SERVER-5.70-266 [nmID=4001A]
A scheduled "Work To Do" took over one minute to be run.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~

Debugger presents:

Break at 0010da97 because of Keyboard request
NOTICE: Executing in a real mode interrupt context.
Enter command "gp" to <g>o until again in a <p>protected mode context.
Current Focus Processor: 00
EAX=00000000 EBX=0000DEB8 ECX=00000001 EDX=00000000
ESI=0050A080 EDI=00000003 EBP=00000001 ESP=80021978
EIP=0010DA97 FLAGS=00000097 (CF PF AF SF)
0010DA97 8AC7 MOV AL, BH
# ?
Address in LOADER.NLM at code start +00008157h
Previous: -000000C5 0010d9d2 LOADER.NLM|RMHWIntPMEntryPoint
Current: 00000000 0010da97
Next: +00000025 0010dabc LOADER.NLM|TurnOnMachineCheckForP0


-> Server Console: the last loaded modules are

-> File Copy Status Screen:
source path: z:\startup
Dst: c:\nwserver

final files copied: nwconfig.msg, nwprv.msg, nwsnut.msg, reboot.bmp, treams.msg,
upgrade.bmp, vrepair.msg

after a <g> or <gp> I have to press ALT+CTRLL+CTRLR+ESC many times before the OS
jumped to the debugger again.

The values debugger presents for "?" are identical as before.

I think there still is a problem for the multi-pathing:
If multi-path support enabled = off, NSSMU (during the setup) "is seeing" the boot
disk twice. with ON, just once, but then "list failover devices" does *NOT* present
a failover device: Maybe build-in failover feature in the LPFC.HAM, so that the OS
doesn't see it? But then I would have expected, that the OS' multi-path support
would not have to be switched to on, so that NSSMU does *NOT* see the disk(s) twice?

Maybe I missed something essential for the configuration of the Emulex 1105-BCv HBA?

Any suggestions appreciated,

Regards, Rudi.