Energy Efficient Resource Provisioning through Energy and SLA Aware(ESA) Algorithm in Cloud Computing

Abstract

Cloud Computing is an Internet-based computing in which services provided by the Cloud Service Providers (CSPs) are in the same way as utilities i.e, pay-per-use model. It enables the clients and endeavors to store and process their information in a cloud infrastructure to give them a chance to concentrate better on their business needs. With the rapid increase in popularity of Cloud Computing growth of data centers also increases to ac- complish the routine services. As a result, a great amount of energy is consumed by data centers even when they remain underutilized. Proper management of this energy consumption becomes a key issue. Virtualization plays a vital role in reducing energy consumption. The process of migra- tion of Virtual Machine (VM) from one server to another is known as Virtual Machine Migration (VMM). Fault tolerance, load balancing, improved performance, and energy management are some benefits of VMM. But it leads to degradation of performance and SLA violations that cannot be ignored. In our work, we propose an algorithm namely Energy and SLA Aware (ESA) VM place- ment algorithm that reduces the overall energy consumption based on 1) After allocation, host utilization 2) VM creation history and 3) the difference between power consumption, SLA violation by the host before and after allocation. It also considers energy efficiency which is a function of power and SLA. The implementation of the proposed algorithm is done in CloudSim. The results show that our method provides a considerable amount of reduction in energy consumption and average SLA violations.