Energy Saving in Cloud Computing

Abstract

Cloud computing is becoming more necessary since numerous internet services are required in daily life. Enhancing the energy efficiency of data centers is essential since they consume a large amount of electricity. In this study, different server assignment policies, with and without jockeying, are simulated. Queueing theory is applied to analyze these policies. Processor sharing servers will be arranged in several ways under these policies. In addition, the effect of idle power is concerned in this study. To find out the most appropriate policy, different situations are simulated by Python programs. After getting these results, graphs are plotted to compare different policies. In this study, it is shown that if the jockeying costs and idle energy consumption rates are negligible, most energy-efficiency server first (MEESF) would perform better than Most energy-efficiency available server first Accounting for Idle Power (MAIP). However, MEESF would be worse than MAIP if the above costs and rates are significant. In addition, E* policy (derived from MEESF) would perform better than MEESF if the number of job arrivals is small. Otherwise, two policy would have the same performance.