Recent research issues in cloud computing

Abstract

Presently, the energy consumed by the Data Centers has reached an alarming level. One of the efficient ways to conserve energy in Data Centers is by server consolidation. By going for server consolidation, the utilization of the hosts are increased, and as a result of this increased utilization of the hosts, the total number of hosts used for running the applications can be reduced. In this thesis, a combinational strategy of Five-Circle algorithm and Tabu Search is employed to conserve energy through efficient server consolidation. Five-Circle Algorithm is used to separate the Virtual Machines (VMs) based on their characteristics. The output of the Five-Circle Algorithm gives the regular VMs. These regular VMs are used for configuration generation to increase the utilization of individual Physical Machines (PMs). Tabu Search algorithm is used for configuration generation. Irregular Virtual Machines are handled by dynamically allocating reserved hosts. A comparison of performance parameters is carried out for Tabu Search and Simulated Annealing. And based on the experimental results, it can be concluded that the proposed algorithm (Tabu Search) performed better than Simulated Annealing in saving energy. Tabu Search provides a maximum energy saving of 74%. The conclusion presents the amendments that can be implemented to enhance the performance of the proposed method.