| The increasing need of large-scale data centers has brought new challenges to the development of energy-efficiency techniques. Although many optimization techniques have been proposed, most of them neglect the characteristics of tasks and fail to consider the dependencies between tasks. Therefore, more comprehensive optimization is needed. The goal of this thesis is to combine capacity provisioning with scientific workflow scheduling to generate optimization in these two ways. Capacity provisioning is built upon a fuzzy logical mechanism to accomplish constructing computing resources proactively, and workflow scheduling is designed with the purpose of achieving energy efficiency. We aim at achieving optimization both in energy consumption and completion time. |
