Best Fit Sharing and Power Aware (BFSPA) Algorithm for VM placement in cloud environment

Abstract

The increasing need for computational powers in almost all spheres of society has resulted in the creation of big data centers. These large data centers exhaust large amounts of electrical power and hence have become a big issue for cloud service providers due to both environmental and financial concerns. These concers have led service providers to summon energy efficient techniques to save as much energy as possible while keeping in mind the quality of services to be delivered. This paper focuses on one such technique, i.e. Virtual Machine (VM) placement, where many VMs are deployed on a single physical machine. There have been different versions of VM placement to reach the optimum. In this paper, yet another version of VM placement is presented that uses the best fit technique to allocate VMs onto hosts, considering the effects on power consumption and memory sharing for various VM-host combinations while making the decision. The consideration of these two factors makes the presented algorithm more efficient than algorithms that are oblivious to these factors as well as those which consider either of them. The Best Fit Sharing and Power Aware Algorithm uses randomly generated data to check the efficiency.