Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!wasatch!cs.utexas.edu!uunet!bartek1!tapa!larry
From: larry@tapa.uucp (Larry Pajakowski)
Newsgroups: comp.unix.xenix
Subject: init's death solved (maybe)
Message-ID: <1989Jul5.154102.703@tapa.uucp>
Date: 5 Jul 89 15:41:02 GMT
Organization: Pata Consultants
Lines: 41

First thank's to all those who took the time to reply to my problems with init
exiting.  Most suggestions were that init was being signaled by a process with
root permissions.  Ed Hew also suggested looking at the process accounting log
for hints.

Here is what I think was going on.

Once an hour I run a script to disable and enable the tty port attached to the
TB+ to take care of the port/modem going comotose.  Concurrent and unknown to
me a getty was being re-spawned on tty11 every minute.  This is very similar
to problems we and others have had with smart modems not configured correctly.

Now the Compaq this runs on only has 10 function keys so tty11 didn't have
anything to talk to.  I suspect tty11 being enabled was a relic of the 2.2.6
version running on the machine previously.  I seem to remember that the tty
ports were renamed as part of the 2.3.1. release to support F11 and F12.

The log of failures indicated that they occured right after the
disable/enable script was run.  Looking at the process accounting log revealed
that when the disable/enable script executed at the same time (one second
resolution here) as the re-spawning of the tty11 getty init would exit.

I suspect that init had received the kill signal say from the disable/enable
when the kill signal from the getty exiting happened.  This then caused init
to get confused and exit.

The solution then was to disable tty11 which wasn't supposed to be enabled in
the first place. 

Most likely if this is the problem there may be other situations like this
where init is busy processing signals at about the same time from exiting
processes, enable and disable.  Beware.

SCO I like you product and I hope you can fix this.  The last 6 months have
been rather frustrating to me and our users.  Again thank's to all who
responded.

Larry Pajakowski
Abbott Labs.
larry@abtcser
1-312-937-1153