Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!usc!apple!agate!darkstar!ucscb.UCSC.EDU!sequoia
From: sequoia@ucscb.ucsc.edu (Gary M. Lin)
Newsgroups: comp.unix.admin
Subject: Re: idle kill program needed
Message-ID: <8418@darkstar.ucsc.edu>
Date: 1 Nov 90 16:55:24 GMT
References: <1990Oct29.180340.3166@d.cs.okstate.edu>
Sender: usenet@darkstar.ucsc.edu
Followup-To: quantum@slugmail.ucsc.edu
Distribution: na
Organization: University of California, Santa Cruz; Open Access Computing
Lines: 29


In article <1990Oct29.180340.3166@d.cs.okstate.edu> klarich@d.cs.okstate.edu (KLARICH TERRY JAME) writes:
>
>I know there was just a big discussion about killing off idle users.
>However, I did not pay any attention because I never thaught I would need
>to worry about this.  Well, my boss asked me to get something running which
>would get rid of idle users after a certain time.  I ftped a program like
>this from simtel; but, it could not handle people coming in from telnet.
>Also, this program would just pay attention to the last time a character
>was read and then start counting.  So, if the program the user was running
>did not require a keyboard read within kill_time minutes, they were logged off
>even though they were not idle.


	The CATS programmers on campus once installed an idle-user killer
	on our open-access system, to free up ports on a heavily loaded
	IS68K running 4.2BSD.  It would sit in background waiting for your
	tty to be silent for say, 15 minutes, before logging you out.

	Needless to say, students didn't appreciate this 'feature', as
	logging back in entails waiting on a 10 minute queue.  So people
	ran 'yes' and left their terminals for hours on end logged in.
	So far, the programmers haven't caught on :)

--------
Gary M. Lin	Cowell College, University of California, Santa Cruz

INTERNET:	sequoia@ucscb.ucsc.edu, quantum@slugmail.ucsc.edu
UUCP:	..ucbvax!ucscc!ucscb!sequoia