Human Opinion Dynamics Based Job Scheduling

1831 Words Jun 29th, 2016 8 Pages
A Novel Human Opinion Dynamics based Job Scheduling in Multi-cluster Environment

Aakanksha

Department of computer engineering & technology, Guru Nanak Dev University, Amritsar, India aakanksha_murgai21@yahoo.com

Abstract—Scheduling parallel jobs is a vital issue inside multi-cluster environment as it plays crucial role in exploiting resource usage along with minimizing response time. The way scheduler schedules the job depends on considering various factors such as number of jobs, availability of processors, arrival time, execution time. Various scheduling strategies have been introduced over the decades to distribute processors among clusters. This paper introduces the detailed method for scheduling jobs across clusters and gives an attempt to implement the same for scheduling jobs and allocation of resources to optimize the performance in multi-cluster heterogeneous environments by applying a novel Human Opinion Dynamics based Optimization technique. The multi-objective function is being included in this technique to optimize the makespan and flowtime, thus solving the problem of optimal allocation of jobs.

Keywords—Multi-cluster environment; scheduling; heterogeneous clusters; Human Opinion Dynamics Algorithm (HOD).

I. INTRODUCTION

Accurate job scheduling has a great impact on overall system performance; depending upon various job specifications. Allocation of system resources on demand to different jobs by scheduler is known as job scheduling. But, scheduling…
Open Document