Application of Viral System Algorithm in Load Balancing of Cloud Environment

Damodar Tiwari,Sanjeev Sharma,Shailendra Singh

doi:10.3844/jcssp.2018.908.918

Abstract

As the cloud computing technology is gaining popularity with time, more and more users and applications are shifting towards it. This is why clouds are experiencing high load, which demands for load balancing of user tasks submitted to cloud for execution. This makes load balancing of non-preemptive tasks a key issue in cloud computing. Superior task scheduling leads to balanced loads among cloud nodes, which results in faster execution of tasks. Task scheduling in cloud environment is an instance of NP-hard optimization problem. When few nodes in a cloud are overloaded whereas other nodes are under loaded then in such situation the performance of overloaded VMs is diminished. It demands a task scheduling so that the incoming tasks can be distributed uniformly across virtual machines (VMs) for proper utilization of available resources. In this study, we propose a novel load balancing algorithm named Viral System Based Load Balancing (VSB-LB) algorithm, which is based on bio-inspired viral system algorithm that distributes the tasks uniformly among VMs. The proposed algorithm is compared with basic load balancing algorithms such as First Come First Serve (FCFS), Weighted Round Robin (WRR) as well as newer bio-inspired Load balance Aware Genetic Algorithm (LAGA) to show its effectiveness. Simulation results proved that VSBLB outperforms FCFS and WRR and LAGA for performing load balancing.

Highlights

Now-a-days one of the fastest emerging fields in information technology is cloud computing, referred to as “the cloud,” which delivers on-demand computing resources including everything from platform and applications, to data centers over internet on a payfor-use basis
NIST has defined three service models that a cloud service provider can provide to consumers which are Software as a Service (SaaS), Platform as a Service (PaaS) and Infrastructure as a Service (IaaS)
Since effective load balancing can reduce the makespan considerably performance of VSBLB is compared with First Come First Serve (FCFS), Weighted Round Robin (WRR) and Load balance Aware Genetic Algorithm (LAGA) in terms of makespan i.e., the overall task completion time (Dhinesh Babu and Venkata Krishna, 2013)

Summary

Introduction

Now-a-days one of the fastest emerging fields in information technology is cloud computing, referred to as “the cloud,” which delivers on-demand computing resources including everything from platform and applications, to data centers over internet on a payfor-use basis It is entirely an internet-based approach where all the resources are placed on a cloud consisting number of high speed interlinked computers for serving the incoming requests from connected clients. Under this technology, clients can use computational power, software services and platform offered by cloud service providers while paying only for duration those resources have been accessed.

Objectives

Results

Conclusion