Summary: Fair Scheduling of Dynamic Task Systems on
Multiprocessors #
Anand Srinivasan and James H. Anderson
Department of Computer Science, University of North Carolina at Chapel Hill
Abstract
In dynamic real­time task systems, tasks that are sub­
ject to deadlines are allowed to join and leave the sys­
tem. In previous work, Stoica et al. and Baruah et al.
presented conditions under which such joins and leaves
may occur in fair­scheduled uniprocessor systems with­
out causing missed deadlines. In this paper, we extend
their work by considering fair­scheduled multiproces­
sors. We show that their conditions are su#cient on M
processors, under all known (dynamic­priority) Pfair
scheduling algorithms, if the utilization of every subset
of M - 1 tasks is at most one. Further, for the gen­
eral case in which task utilizations are not restricted in
this way, we derive su#cient join/leave conditions for
the PD 2 Pfair algorithm. We also show that, in gen­
eral, these conditions cannot be improved upon without

Source: Anderson, James - Department of Computer Science, University of North Carolina at Chapel Hill

Collections: Computer Technologies and Information Sciences