From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.5-pre1 (2020-06-20) on
	ip-172-31-74-118.ec2.internal
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 required=3.0 tests=BAYES_00 autolearn=ham
	autolearn_force=no version=3.4.5-pre1
Date: 2 Nov 92 16:22:21 GMT
From: agate!usenet.ins.cwru.edu!magnus.acs.ohio-state.edu!cis.ohio-state.edu!ne
ws.sei.cmu.edu!jbg@ucbvax.Berkeley.EDU  (John Goodenough)
Subject: Re: Real Time Scheduling Methods (Deadline Scheduling)
Message-ID: <1992Nov2.162221.22507@sei.cmu.edu>
List-Id: <comp.lang.ada>

One disadvantage of deadline scheduling is that when a system is overloaded,
it is unpredictable which task will miss its deadline.  The early popularizers
of rate monotonic analysis (RMA), John Lehocsky and Lui Sha, investigated
deadline scheduling before turning to rate monotonic analysis.  They disliked
the unpredictability of deadline scheduling under system overload, and so
turned to seeing if RMA could be usefully applied to real systems.  (Needless
to say, they found it could.)  This is not to say that deadline scheduling is
"bad" or should not be used, but one should be aware of the tradeoffs involved
when choosing between techniques.	

[Under RMA, when a system is overloaded, it is predictable which tasks will
make their deadlines (namely all tasks above a certain priority).  If a
critical task is not in this set of tasks, there is a standard technique for
ensuring it will meet its deadline while some other task potentially will miss
a deadline.]

John B. Goodenough					Goodenough@sei.cmu.edu
Software Engineering Institute				412-268-6391