Job classification in cloud computing: the classification effects on energy efficiency

摘要: One of the recent and major challenges in cloud computing is to enhance energy efficiency data centers. Such enhancements can be done by improving resource allocation management algorithms. In this paper, a model that identifies common patterns for jobs submitted proposed. This able predict type job submitted, accordingly, set users' classified into four subsets. Each subset contains have similar requirements. addition jobs' pattern requirements, history considered prediction model. The goal classification find way propose useful strategy helps improve efficiency. Following process classification, best fit virtual machine allocated each job. Then, machines are placed physical according novel called Mixed Type Placement strategy. core idea proposed place different types same whenever possible, based on Knapsack Problem. because do not intensively use compute or storage resources machine. reduces number active which leads reduction total consumption center. A simulation results shows presented outperforms both Genetic Algorithm Round Robin from an perspective.