Sorry for the delay in posting a summary, but I wanted to wait a week and make sure
the machine wouldn't crash again. Thanks to all who replied.
The one I chose first was to diff my login program with the distribution login. The
two programs were indeed different, and since replacing login the machine hasn't crashed
again. (thanks firstname.lastname@example.org.AU (Brett Lymn))
I also received some helpful advice about getting crash dumps, so I have
that set up now if my machine reboots again......
email@example.com (Jim Davis) writes:
You can turn on crash dumping by adding (or uncommenting) the lines
mkdir -p /var/crash/`hostname`
echo -n 'checking for crash dump...'
intr savecore /var/crash/`hostname`
in /etc/rc.local. Crash dumps can be pretty big, so you'll want to make
sure /var/crash has enough room. (Or simply choose some other place with
more room for dumps; the name is arbitrary.)
Once you've captured a dump, you can analyze it with adb or /etc/crash. The
latter program is easier to use, and the man page has a pretty good summary
of how to use it. Getting a traceback is perhaps the most useful thing --
it will try to point to the particular place where the crash occurred. You
can also get a good idea of the overall state of the machine at the time --
what else was running, overall system load, process's state, etc.
The next thing I will try if that doesn't help will be a suggestion I received
from firstname.lastname@example.org.EDU (Robert Montjoy). He makes reference to a Sun Patch 100359-04
Keywords: panic assertion failure, machine freeze, console printf hangs, jumbo
Synopsis: SunOS 4.1;4.1.1;4.1.2: streams jumbo patch
SunOS release: 4.1 4.1.1 4.1.2
Topic: Streams Jumbo Patch
BugId's fixed with this patch: 1068158 1044100 1044888 1047355 1058097 1061174
Architectures for which this patch is available: Sun3 Sun3x Sun4 Sun4c Sun4m
I also received a paper "Building and Debugging SunOS Kernels" by
Hal Stern (email@example.com) at Sun Microsystems, Northeast Area Consulting Group
which should prove helpful in the future.
I even received one response from Perry_Hutchison.Portland@xerox.com who suggested it
may be a hardware problem.
Thanks again for all your help,
Computer Sciences Corporation
Environmental Research Laboratory
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:50 CDT