SUMMARY: Help deciphering Solaris reboot messages

From: Mark Martinez CIC-2 (090632@mantaray.LANL.GOV)
Date: Thu Apr 18 1996 - 15:00:59 CDT


My question was:

>
> Would someone please provide some guidance or pointers to documentation
> which can help me decipher messages relating to a spontaneous reboot?
>
> panic reboot messages deleted for the sake of brevity
>

What I did:

Updated all recommended patches and I plan on investing in the book
recommended below.

Summary of responses:

"tl_do_proto" and "tl_wsrv" might be a clue. Do any of
the applications that run on this SS2000 use the letters
"TL" in there name ? If so try to match the times of the
crash up with whatever the application is doing. This
might lead you somewhere.

+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Bill Holzapfel billh@dcvast.com
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

While any kernel panic is a major problem, this one stands out because
it was 'sched' that caused the panic. sched in the Unix scheduler, the
process that is ALWAYS at the top of a 'ps -ef', ALWAYS process ID 0.

Are you doing anything out of the ordinary in terms of real time
processes or the like? I would make sure that the Kernel Jumbo
Patch is loaded. If it is, you might try the latest version.

=================================================================
Adam Nevins E-mail: adam_nevins@ltx.com
=================================================================

I could be wrong, but this looks like your kernel called a system call that
didn't exist. I think I recognise this as being fixed by 101945-37, and see
you're running -32, so try that first.

The best way to learn about all this is to buy "Panic! Unix System Crash Dump
Analysis" by Chris Drake and Kimberley Brown. ISBN: 0-13-149386-8. My copy is
signed by Kimberley, and it comes with a CD with useful utilities on, including
a kludge to let you adb crashdumps from different kernel architectures and OS
levels than you're running on. Brilliant. Usually running adb -k on the
coredump and getting the stack dump and messages out is enough to match a known
bug, o rforward to your service provider for them to analyse.

======================================================================
Mark R. Bowyer mbowyer@uk.cv.com
======================================================================

Jeez !! Looks like you need some *big* patches....!!! I would not know
which ones - they could be application specific. What special
applications are you running on your 2000....???

======================================================================
Rahul Roy Email: roy@bluestone.com
======================================================================

thanks

---
Mark L.B. Martinez (mlbm@lanl.gov)
Research Library, MS P362, P.O. Box 7113, Los Alamos, NM 87544-7113
Phone +1 505.667.1223   Fax +1 505.665.6452



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:10:58 CDT