VM Consolidation based on Overload Detection and VM Selection Policy

Even though cloud computing has been a big boon to the ICT (Information and communication technology) industry, it faces high energy consumption and substantial CO2 emission. Due to the increase in demand for computational resources, it is now necessary and of utmost significance to improve the energy consumption of the cloud system. Virtual Machine (VM) consolidation is one of the powerful tools to improve energy efficiency as it reduces the number of VM migrations by managing VMs from overloaded/underloaded hosts. Implementation of VM consolidation techniques leads to a decrease in the amount of hardware consumption, energy consumption, and data footprints which leads to an increased Quality of Service (QoS). In this paper, an energy aware VM selection algorithm is proposed along with an overload detection algorithm. The proposed algorithm runs on the CloudSim toolkit environment and analyzes it based on different parameters like energy consumption, SLA violation, server shutdown, and the number of VM migrations to analyze energy efficiency improvement. This modified approach exhibited better performance on all the parameters as compared to the existing algorithms.