Killing a process

Once in a rare while your computer will die and when you log back on
to continue your work, it will report that you can't run the program
because another copy of it is already running.  Typically that is the copy
that was running when your connection died.  The mainframe let it continue,
but won't let you back in it.

At this point, you may tear off your clothes and run naked and screaming
out of the lab; or if you prefer, you may try this solution:  

Enter 
        ps -aux |grep lcrew     (Replace lcrew with your login name)

The program will respond with a repport like this:

lcrew    26943 15.8  0.4  280  576 p2 R    12:42   0:00 ps -aux
lcrew    26944  0.0  0.2   40  248 p2 S    12:42   0:00 grep lcrew
lcrew     9675  0.0  0.3   48  432 p2 S    07:56   0:00 -ksh TERM network HOME=/
lcrew    25794  0.0  0.9  952 1448 p2 T    12:22   0:04 nn


From that list, I pick the program (far right) that I want to kill.  Then
I kill it by entering its number (second column) within this syntax:


        kill -9 25794

In this case, I killed the copy of 'nn' that was running.  That allowed
me to start a new copy of nn.


I use this solution often enough to be a nuisance; yet I refuse to memorize
the fix:  instead, I save it as a file named "kill" in my directory.