Authors: Anurag Jain; Rajneesh Kumar
Addresses: Department of Computer Science and Engineering, Chandigarh Engineering College, Landran, Mohali, Punjab, India ' Department of Computer Science and Engineering, Maharishi Markandeshwar Engineering College, Maharishi Markandeshwar University, Mullana, Ambala, India
Abstract: Cloud computing is the result of growth in networking and computer technologies in last few years. It focuses on providing scalable and reliable infrastructure for implementation of internet-based services. Its future depends on effective installation of infrastructure, efficient utilisation of resources and dynamic transformation of load. This will help in achieving better fault tolerance ratio and higher customer satisfaction. To distribute the workload dynamically across multiple nodes so that no server is either underutilised or overwhelmed, there is a need of efficient and effective load balancing strategy which should change itself dynamically according to environment and task. In this paper, authors have proposed a hybrid load balancing approach named JIMC by combining best features of some of the existing approaches. JIMC approach has been tested using cloud analyst simulator on various scenarios. It has been observed that JIMC approach is efficient in comparison to existing approaches and also prevent the system failure by giving warning of systems overloading in advance.
Keywords: cloud computing; cloud analyst; load balancing; task scheduling; join idle queue; join shortest queue; minimum completion time; minimum execution time; round robin; equally spread; throttled.
International Journal of Communication Networks and Distributed Systems, 2017 Vol.18 No.3/4, pp.264 - 286
Received: 15 Dec 2015
Accepted: 28 Jun 2016
Published online: 18 Apr 2017 *