Path: utzoo!utgpu!watserv1!watmath!att!mcdchg!laidbak!ism.isc.com!uunet!mcsun!hp4nl!svin02!wsinpdb
From: wsinpdb@svin02.info.win.tue.nl (Paul de Bra)
Newsgroups: comp.unix.admin
Subject: Re: Program to renice jobs
Message-ID: <1500@svin02.info.win.tue.nl>
Date: 19 Oct 90 16:08:25 GMT
References: <1990Oct19.094659@granada.mit.edu>
Organization: Eindhoven University of Technology, The Netherlands
Lines: 18

In article <1990Oct19.094659@granada.mit.edu> andrew@granada.mit.edu (Andrew Gunstensen) writes:
>Is there a good method of renicing jobs automatically once they
>have exceeded some CPU time limit?

Funny, while i worked at the univ of Antwerp, our Gould system
with Utx/32, which was based on BSD4.2, had this built into the kernel.
All jobs would automatically be niced after using some amount of
cpu-time.

I presume other manufacturers took that code out because more and more
programs legitimately use a lot of cpu-time (nfs deamons, X-server,
GNU-emacs, etc)

Maybe it's still in the source code, albeit commented out.
Then you can go look for it and put it in again. But as far as i know
there is no code to allow for exceptions, unless owned by root.

Paul.
(debra@research.att.com)