Title: Modified MapReduce for efficient data management: a task scheduling technique

Authors: D. Rajeswari; V. Jawahar Senthilkumar; M. Prakash; S. Ramamoorthy

Addresses: Department of Computer Science and Engineering, Faculty of Engineering and Technology, SRM Institute of Science and Technology, Kattankulathur, Tamil Nadu, India ' Department of Electronics and Communication Engineering, College of Engineering, Anna University, Chennai, Tamil Nadu, India ' Department of Computer Science and Engineering, Faculty of Engineering and Technology, SRM Institute of Science and Technology, Kattankulathur, Tamil Nadu, India ' Department of Computer Science and Engineering, Faculty of Engineering and Technology, SRM Institute of Science and Technology, Kattankulathur, Tamil Nadu, India

Abstract: High performance computing efficiency depends on distributed systems. Task scheduling is an important key factor in distributed environment to reduce the amount of time used to obtain the optimal schedule. Genetic algorithm (GA) is the efficient technique to schedule task for distributed resources. MapReduce is used to develop distributed applications and it provides the efficient way to implement genetic algorithm scheduler for heterogeneous environment. Due to some specific characteristics, some application cannot be efficient in map reduce, because it follows a two-phase pattern. GA is an iterative process. To fit GA in MapReduce, parallel GA (PGA) is used. This paper presents modified MapReduce model for genetic algorithm to schedule an independent job which reduces makespan and flowtime. The outcome shows the efficacy of proposed method in the observation with existing techniques on benchmark instances.

Keywords: genetic algorithm; heterogeneous environment; modified MapReduce; multi-objective; task scheduling.

DOI: 10.1504/IJPSPM.2024.138765

International Journal of Public Sector Performance Management, 2024 Vol.13 No.4, pp.491 - 503

Received: 03 Jul 2020
Accepted: 11 Oct 2020
Published online: 31 May 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article