Indian Journal of Science and Technology

Abstract

Objectives: One of the essential requirements for improving Quality of Service of Cloud is to design load balancing algorithms that equally spread the load among the VMs such that neither of them is overloaded at any given point of time while ensuring that each of the VM is optimally utilised. The objective of this paper is to propose an efficient VM load balancing policy called Timestamp based Stateful Throttled VM load balancing algorithm. Methods/Statistical Analysis: The proposed algorithm deals with the space performance constraints of the existing Stateful Throttled VM Load balancing algorithm. It is based on the principle of locality of reference and deletes the old state information to reclaim space. The authors have carried out experimental analysis to compare the proposed algorithm with the existing algorithms. Findings: The comparative analysis shows that the algorithm behaves in the same manner as Stateful Throttled Algorithm while taking into consideration its space performance limitations. The authors have observed better performance in terms of response time and data center processing time for Timestamp based Stateful Throttled VM load balancing algorithm than Throttled VM load balancing algorithm in case of spatially distributed VMs. Application/Improvements: The salient features of the proposed algorithm are better response time and data center processing time in case of spatially distributed data centers in cloud.
Keywords: Cloud Computing, Data Center, Load Balancing, Virtualization, Virtual Machine