A novel approach for server consolidation to balance energy consumption and SLA violations

Abstract

Cloud computing is an on demand resources based computing model. This requires large amount of physical resources to provide services to the users on the basis of pay-per-usage model. So immoderate demand of physical resources has been increase in computational power. As a result, a large 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 an essential role in reducing energy consumption. It provides technique that facilitate better use of resource is known as server consolidation. Server consolidation is used to reduce total number of servers by providing better use of the available data centre resources, but the excessively consolidation of virtual machines may cause to service-level agreement (SLAs) violations. SLA establishes the Quality of Service (QoS) agreed between service-based systems. Therefore, our goal is to reduce energy consumptions as well as SLA violations. It is very convenient to strike a tradeoff between energy efficient and minimization of SLA violations. In proposed work, a new approach is given, which provides balance between energy consumption and SLA violations. Simulation results using WorkflowSim show that our proposed approach can reduce SLA violation by approx 5.67% and energy consumption by approx 16.23% under the PlanetLab workload compared to MBFD algorithm.