ABSTRACT

Cloud computing is gaining more importance gradually in the field of computing services with the support of data centers across the world. A large number of enterprises and individuals are opting for cloud computing services for resource requirements. The number of requests for services is raising in cloud computing, which leads to increase in power consumption of data centers with high pace. This caused the rise in ownership cost of service providers and harmful carbon footprints into the environment. Therefore, it is imperative to optimize the power requirements in data centers to mitigate the cost of ownership and to make it environment-friendly. In today’s era, virtualization plays a significant role to minimize power consumption during virtual machine live migration in data centers. This paper presents the hybrid genetic algorithm that provisions various virtual machines to hosts in such a way to optimize power requirements of cloud services during virtual machine live migrations. Simulation experiments have been carried out with a variety of characteristics as input to Power Optimizing Genetic Algorithm with different allied parameters of migration. Results have shown that proposed genetic algorithm optimize power consumption and migration overhead with defined test problems as compared to recent virtual machine placement method. The statistical approaches have been applied to validate the reliability of the simulation results.