Abstract We study the problem of scheduling periodic and sporadic time-critical tasks on the nodes of a distributed system. A periodic task consists of an infinite number of requests, each of which has a prescribed deadline. Each node is a monoprocessor machine and we assume that the existence of enough processing power for all periodic tasks is verified at system initialization time. We develop an algorithm which determines on-line, whether or not an occurring sporadic task can be guaranteed to meet its timing requirements. Guarantee means that all the tasks, sporadic and periodic, supported by the node can be feasibly scheduled by the preemptive Earliest Deadline scheduling strategy. We show that this algorithm is optimal, e.g., guarantees a task whenever it is possible and hence provides an efficient acceptance test which runs in linear time.
[1]
Omri Serlin,et al.
Scheduling of time critical processes
,
1899,
AFIPS '72 (Spring).
[2]
J. Leung,et al.
A Note on Preemptive Scheduling of Periodic, Real-Time Tasks
,
1980,
Inf. Process. Lett..
[3]
Chung Laung Liu,et al.
Scheduling Algorithms for Multiprogramming in a Hard-Real-Time Environment
,
1989,
JACM.
[4]
W. A. Horn.
Some simple scheduling algorithms
,
1974
.
[5]
Charles U. Martel,et al.
Scheduling Periodically Occurring Tasks on Multiple Processors
,
1981,
Inf. Process. Lett..
[6]
Eduardo D. Sontag,et al.
Real Addition and the Polynomial Hierarchy
,
1985,
Inf. Process. Lett..