SUMMARY: how to stop serial "doorstopping" [stair-stepping] on sunfire v240?

From: Duncan C White <dcw_at_doc.ic.ac.uk>
Date: Thu Jun 14 2007 - 14:39:42 EDT
Hello all,

On Thu, 14 Jun 2007, I <dcw@doc.ic.ac.uk> wrote:

> I've setup serial access to a Sunfire v240.. [but] once I boot Solaris 8
> all the "output" messages start "doorstepping" ...
>
> SunOS Release 5.8 Version Generic_117000-05 64-bit
> Copyright 1983-2003 Sun Microsystems, Inc.  All rights reserved.
> Hardware watchdog enabled
> configuring IPv4 interfaces: bge0.
>                                   Hostname: pipit.doc.ic.ac.uk

Many thanks to all who responded:

Adam Kirby <A.Kirby@surrey.ac.uk>
John O'Reilly <jro4747@bjc.org>
Mike's List <mikelist@sky.net>
Crist Clark <Crist.Clark@globalstar.com>
Edward Scown <eascown3@yahoo.com>

First off, my apologies for switching my brain off while posting
this.  Several people pointed out very politely that I meant
"stair-stepping" not "door-st[oe]pping".. Sigh. I'm sure that's
what I typed all along, isn't it?:-)

Adam and John both correctly pointed out that Sunsolve document
40646 discussed the "stair-step" phenonemon during booting, this
is caused by corrupted "serial" entries in /etc/path_to_inst.  I'm
not quite sure why the entries were corrupted - given that the
server was freshly jumpstarted, the automated install had only just
created /etc/path_to_inst! - but this was undeniably the cause.

The sunsolve document suggests two remedies:

1. get solaris to rebuild /etc/path_to_inst by moving it sideways
    and rebooting with "boot -a" or "boot -ar".

2. hand editing the /etc/path_to_inst file, fixing the broken entries.

I tried (1), ran into some further problems with it failing to look
up the /dev and /devices entries for the root disk and not regenerating
/etc/path_to_inst, backed out, and then successfully tried (2), fixing
the problem completely.  In particular, the initial "serial" entries
(grep serial /etc/path_to_inst) were:

"/pci@1f,0/isa@7/serial@0,3f8" 0 "su"
"/pci@1f,0/isa@7/serial@0,2e8" 3 "su"
"/pci@1f,0/pci@1,1/isa@7/serial@0,3f8" 1 "su"
"/pci@1f,0/pci@1,1/isa@7/serial@0,2e8" 2 "su"
"/pci@1e,600000/isa@7/serial@0,3f8" 4 "su"
"/pci@1e,600000/isa@7/serial@0,2e8" 5 "su"

and I replaced them with:

"/pci@1f,0/isa@7/serial@0,3f8" 0 "su"
"/pci@1e,600000/isa@7/serial@0,3f8" 1 "su"
"/pci@1e,600000/isa@7/serial@0,2e8" 2 "su"

[i.e. kept the first, lost the 2nd-4th, and edited the "4","5" column to
read "1","2" respectively..]

and rebooted, all traces of stair-stepping are now gone, and tcsh
doesn't even whine about "no job control" either:-)

So, once more sun-managers rides to the rescue!  Thanks very much
everyone!  [Perhaps if I'd managed to type "stair-stepping" into
google/sunsolve I might have got somewhere without asking the list:-)]

 	cheers,
 	duncan

---------------------------------------------------------------------------
Duncan C. White, Unix Systems Programmer, Dept of Computing,
                  Imperial College, London, SW7 2BZ, UK.
---------------------------------------------------------------------------
 	           ALL THE REAVERS IN THE 'VERSE: GRAAAAAAAAAAAAAAAAAAAAAR!
 	               - Serenity in 2000 words or less, Ron Swartzendruber
---------------------------------------------------------------------------
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
Received on Thu Jun 14 14:40:06 2007

This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:44:06 EST