Abstract—Fuzzy Systems have shown their utility for solving a wide range of problems in different application domains. The use of Genetic Algorithms for designing fuzzy systems allows us to introduce the learning and adaptation capabilities. This has attracted considerable attention in parallel job scheduling. In this paper, we present a methodology for automatically generating online scheduling strategies for different types of process granularity in parallel job scheduling. The scheduling problem includes all the synchronization granularity of parallel jobs.In order to allow a wide range of objective functions, we use a rule based scheduling strategy. The rule system classifies all possible scheduling strategies for the process grains and assigns an appropriate scheduling strategy based on the process grains. The rule bases are developed with the help of a genetic fuzzy system that uses workloads from Logs of real parallel workloads from production systems. http://www.cs.hiji.ac.il/labs/parallel/workload/logs.html. We have already developed a new scheduling algorithm called Agile Algorithm which schedules the jobs according to the synchronization granularity and this paper focuses on the good optimized results for the Agile Algorithm using Genetic Fuzzy System.
Index Terms—Genetic Algorithm, Fuzzy Systems, Parallel Jobs, Performance metrics.
S. V. Sudha, working as Assistant Professor in the Department of Information Technology, Kalignar Karunanidhi Institute of Technology, Coimbatore 641402, Tamil Nadu, India (e-mail: email@example.com)
K. Thanushkodi, Principal of Akshaya College of Engineering and Technology, Coimbatore -642 109, Tamil Nadu, India
Cite: S. V. Sudha and K. Thanushkodi, "Process Grain Sized Based Scheduling of Parallel Jobs using Genetic Fuzzy Systems," International Journal of Computer Theory and Engineering vol. 2, no. 5, pp. 768-772, 2010.