A Power-Utilization Based Virtual Machine Allocation Algorithm

Cloud computing becomes a buzzword now a day because of its simplicity, reliability, portability and security in maintaining and manipulating services from remote locations in the world. As the uses of cloud computing increase, more and more data centers are deploying throughout the world for providing high quality services to the consumers. The incremental performance of these data centers deserves a vast amount of energy to keep breathing and to response on demand immediately. Hence, development of efficient solutions is necessary that can reduce energy consumption without violating the performance much. Utilizing the physical resources at their best can be an answer to this problem. In this work, we have used the Simulated Annealing algorithm where it uses both of the power and the CPU utilization to select the best power saving host for allocating a VM. We have developed the algorithm in such way that as much workloads as possible can be served using least number of servers in a data center. We also cared about the SLA violation, performance degradation, VM migration etc. which is clearly clarified in the outcomes. We have performed exhaustive simulation study using our algorithm and the outcome of simulation results confirm the efficient resource utilization of our algorithm.