A comparison study of different algorithms for energy-aware placement of virtual machines

Abstract

Cloud Computing services are essential to modern society. The increasing number of people and organisations using these types of services results in a higher demand in datacenters, which in turn, is raising energy consumption and carbon footprint. Reducing energy consumption has become a subject of interest to many researchers, who approach the problem with different optimisation processes and scheduling algorithms. This article shows an extensive vision of the steps followed by a datacenter, upon the arrival of a task or application by showing how it traverses along the processing time-line, and focusing on the energy-aware point of view of the datacenter. A crucial role is played by placement process of Virtual Machines (VM). Simulations using the CloudSim simulator were performed and results are reported to show a performance comparison of several selected algorithms, which focus in the VM placement problem, and considering two scenarios: empty and loaded datacenter. The results are evaluated in terms of energy consumption, quality of service and resource memory efficiency, among others