نوع مقاله : علمی-پژوهشی
نویسنده
گروه علوم کامپیوتر، پردیس ریاضی و کامپیوتر خوانسار، دانشگاه اصفهان
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسنده [English]
In the present study, Energy Valley Optimizer (EVO), as an emerging meta-heuristic optimization algorithm, is employed to solve the problem of virtual to physical machine placement (VMP) in a cloud data center. Further, in order to accommodate the algorithm for the placement problem, some necessary changes were applied to the algorithm. EVO algorithm equipped with a set of properly-designed operators was then employed to solve the VMP problem. Minimizing the power consumption was considered as the optimization goal. The VMP problem was formulated as a constraint optimzation problem. In the evaluation phase, a data center with a given set of heterogeneous physical machines with different input workloads generated synthetically were modeled to evaluate the effectiveness of the EVO algorithm. The results obtained by EVO were compared with several heuristics. The evaluation results indicate that EVO algorithm is able to reduce power consumption from around 3 to around19 percent compared with FFD. Furhter, EVO outperformed all the evaluated heuristics in terms of power consumption. As a secondary parameter, resource wastage in the data center was also evaluated, which the obtained results show effectiveness of EVO when compared to other heuristics.
کلیدواژهها [English]
)