[prev in list] [next in list] [prev in thread] [next in thread] 

List:       tru64-unix-managers
Subject:    Killing jobs
From:       <norton () nrlmry ! navy ! mil>
Date:       1996-09-30 21:41:13
[Download RAW message or body]


Hi -

I'm trying to debug some codes that use the DECthreads stuff (via KAP
F90) on 3.2D.  Every now and then a code will go into a state where it
is taking 100% of all CPUs it is running on and will not respond to
any attempts to kill it.

Killing the job from the controlling shell does not stop it.

Killing the controlling shell does not work.  The controlling shell
dies, but the process it is running doesn't (e.g. killing dbx when
running dbx a.out kills dbx but leaves a.out running).

Suspending it does not stop it, even though the controlling shell
thinks it is suspended.

Using kill -9 on the PID does not kill it.

I have been unable to kill of these rouge jobs short of rebooting the
system.  Has anyone else seen this and been able to figure out how to
kill this type of job without kicking everyone off the system and
rebooting?  I'm willing to do just about anything to avoid having to
reboot (even editing the running kernel's structures, if I had some
documentation on what to do).

thanks
-dave

[prev in list] [next in list] [prev in thread] [next in thread]