In Proceedings of ICDCS’97: The IEEE International Conference on Distributed Computing Systems, Baltimore, Maryland, May 1997. Load Profiling A Methodology for Scheduling Real Time Tasks in a Distributed System

Abstract

Traditionally the goal of load management protocols for distributed systems has been to ensure that nodes are equally loaded In this paper we show that for real time systems load balancing is not desirable since it re sults in the available bandwidth being distributed equally amongst all nodes in e ect making all nodes in the system capable of o ering almost the same bandwidth e g in cycles per second to incoming tasks We show that this one size ts all practice leads to a higher rate of missed deadlines as incoming tasks may be de nied service because they require bandwidth that can not be granted at any single node while plenty of frag mented bandwidth is collectively available in the system We propose a new load pro ling strategy that allows the nodes of a distributed system to be unequally loaded so as to maximize the chances of nding a node that would satisfy the computational needs of incoming real time tasks The performance of the proposed protocol is eval uated via simulation and is contrasted to other dynamic scheduling protocols for real time distributed systems

Extracted Key Phrases

8 Figures and Tables

02040'96'98'00'02'04'06'08'10'12'14'16
Citations per Year

448 Citations

Semantic Scholar estimates that this publication has 448 citations based on the available data.

See our FAQ for additional information.

Cite this paper

@inproceedings{Bestavros1997InPO, title={In Proceedings of ICDCS’97: The IEEE International Conference on Distributed Computing Systems, Baltimore, Maryland, May 1997. Load Profiling A Methodology for Scheduling Real Time Tasks in a Distributed System}, author={Azer Bestavros}, year={1997} }