N-Class aborting boot

[Ralph Grothe]

Good Morning Mates,

the idea of Michael is probably the most promissing that will get me quickly to a bare OS.
When I get it installed and can boot this
I can care for the various firmware upgrades and installation of drivers for the temporarily removed I/O cards.
I only wonder how the swagentd will work without the availability of a network.
I can dimly remember from former installs that there was a catch.
But maybe I will find the answer in the knowledge base.
I also think to remember that the boot up was painstakingly slow with an absent network because of late timeouts of most networked services.
But I think one can always bail out by simply pressing break in init.
I am now retreating to the server room and will give it a try.
I'll come back later to report and assign points.
Er, by the way, where does the PDC release version show up?
It may well appear somewhere during POST,
but I don't know all the boot screens by heart.

[Torsten. · Accepted solution]

You will see the current installed firmware in the very first bootup screen:

example:

Firmware Version 43.43

Duplex Console IO Dependent Code (IODC) revision 1

------------------------------------------------------------------------------
(c) Copyright 1995-1998, Hewlett-Packard Company, All rights reserved
------------------------------------------------------------------------------

Processor Speed State CoProcessor State Cache Size
Number State Inst Data
--------- -------- --------------------- ----------------- ------------
0 440 MHz Active Functional 512 KB 1 MB
1 440 MHz Idle Functional 512 KB 1 MB
2 440 MHz Idle Functional 512 KB 1 MB
3 440 MHz Idle Functional 512 KB 1 MB
4 440 MHz Idle Functional 512 KB 1 MB
5 440 MHz Idle Functional 512 KB 1 MB
6 440 MHz Idle Functional 512 KB 1 MB
7 440 MHz Idle Functional 512 KB 1 MB


You can set up the network on the core I/O card and use /etc/hosts for resolution only or simply touch this file:

touch /var/adm/sw/standalone

(Don't forget to remove it later)

[Ralph Grothe]

Torsten,

funny asside,
I downloaded the firmware from your link
where they claim it to be the latest release as of May 2006.
But look, they have given it a timestamp from 2003

$ gzip -dc PF_CPIW4343.tar.gz|tar tvf -
-rw-rw-rw- 0 3 3365888 Dec 10 18:35:04 2003 PF_CPIW4343.frm

Unfortunately there is no Readme in the tarball.
So is this frm file supposed to be written to tape to later boot from?

[Ralph Grothe]

Ah, that was it.
Thanks for the reminder of the standalone semaphore file /var/adm/sw/standalone

[Ralph Grothe]

Sorry,
found the installation instruction one click away from your link in the readme.

[Torsten.]

Ralph, follow the "Versionshinweise" link, this will give you the instructions ;-)

Most important:

dd if=PF_CPIW4343.frm of=/dev/rmt/0m bs=2k
^^^^^^^^^^^
(device path dependent)

NOTE: Blocksize (bs) must be 2k


The files were moved from ITRC to the Business Support Center some time ago, this is the reason for the different time stamps.

[Ralph Grothe]

Torsten,

maybe you also have an answer to this strange phenomenon I encounter with the box'es LAN console.
I configured and enabled it
but whenever I try to login to it I get kicked out immediately (bogus IP)

$ TERM=vt100 telnet 1.2.3.4
Trying...
Connected to 1.2.3.4.
Escape character is '^]'.
Connection closed.


I know that our network admins are segmenting their LANs by VLANs.
Could there lie a possible cause?
I think to have read somewhere that especially older hardware sometimes has trouble with the 4 byte wider ethernet frames caused by VLAN tagging.

[Torsten.]

Not really sure about this.
Remember, there is this setting in EL:

enable/disable telnet (web connection only).

Did you try the web connection?

I would connect a notebook using a cross over cable to ensure the telnet connection is working first. If it does, the problem is probably somewhere in the network.

[Ralph Grothe]

Hello again,

I managed to set up a PC in the server room with access to our WWW proxy.
Now it is easier for me to log my advances here.
From the BHC IN->FV command I could verify that the current firmware version is 42.06.
I also verified that I have no problem with my replaced RAM DIMMs as I managed to boot from a boot disk of another to be scrapped N-Class that still happened to contain an 11.00 OE.
Of course couldn't it resolve quorum issues of non present (external) disks of its lvmtab.
But I was glad I could boot into the init sequence were I bailed out.
This is a big releive as I was fearing to have damaged (or not properly inserted) some DIMM.
After I had created the firmware upgrade tape and tried to boot from it I ran into an IPL error of "bad LIF magic" in the IODC.
Hm, but I had correct cksums.
But revisiting the firmware's intstallation section in the readme I saw what probably has caused this.
I used the non-rewindable tape device when I created the tape on another box.
But in the instructions it says of=/dev/rmt/0m.
So I'll try again with a new tape...

About my trouble with the LAN console port.
This turned out as a classical demonstration effect when I ran through the whole resetting to defaults, resetting the GSP, and rerunning LC while our network admin (who happened to be present in the server room because of some patch work there).
When I wanted to demonstrate the error to him suddenly I was able to login.
I swear I haven't done it differently the times before.
Must have been the presence of our Holy Networkness?
I'll come back after hopefully having sucessfully booted from the second tape...

[Ralph Grothe]

Outch, how embarrassing,
I tried to flash the gziped tarball!!!
Nevermind, the flashing from 42.06 to 43.43 went smoothly after I had written the actual frm file to tape, and booted from it.