Non-Evolutionary Algorithms for Scheduling Dependent Tasks in Distributed Heterogeneous Computing Environments

Journal Article · Thu Sep 01 04:00:00 EDT 2005 · Journal of Parallel and Distributed Computing

DOI:https://doi.org/10.1016/j.jpdc.2005.04.017· OSTI ID:912188

Boyer, Wayne F; Hura, Gurdeep S

The Problem of obtaining an optimal matching and scheduling of interdependent tasks in distributed heterogeneous computing (DHC) environments is well known to be an NP-hard problem. In a DHC system, task execution time is dependent on the machine to which it is assigned and task precedence constraints are represented by a directed acyclic graph. Recent research in evolutionary techniques has shown that genetic algorithms usually obtain more efficient schedules that other known algorithms. We propose a non-evolutionary random scheduling (RS) algorithm for efficient matching and scheduling of inter-dependent tasks in a DHC system. RS is a succession of randomized task orderings and a heuristic mapping from task order to schedule. Randomized task ordering is effectively a topological sort where the outcome may be any possible task order for which the task precedent constraints are maintained. A detailed comparison to existing evolutionary techniques (GA and PSGA) shows the proposed algorithm is less complex than evolutionary techniques, computes schedules in less time, requires less memory and fewer tuning parameters. Simulation results show that the average schedules produced by RS are approximately as efficient as PSGA schedules for all cases studied and clearly more efficient than PSGA for certain cases. The standard formulation for the scheduling problem addressed in this paper is Rm|prec|Cmax.,

• Research Organization: Idaho National Laboratory (INL)
• Sponsoring Organization: USDOE
• DOE Contract Number: AC07-99ID13727
• OSTI ID: 912188
• Report Number(s): INEEL/JOU-03-00503
• Journal Information: Journal of Parallel and Distributed Computing, Journal Name: Journal of Parallel and Distributed Computing Journal Issue: 9 Vol. 65
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS
ALGORITHMS
Dependent task scheduling
Distributed heterogeneous computing
GENETICS
Heterogeneous processing
Matching and scheduling
SCHEDULES
SIMULATION
TUNING