Energy-sensitive Scheduling for Cloud Data Centers Prone to Failures*

摘要

With the rapid growth of cloud computing, its energy waste and excessive energy consumption have become a big issue. Cloud infrastructure is built on a large number of servers and devices. In the execution processes of computing tasks, faults of different components may occur in server hardware/software at any time. We propose a task scheduling method for high performance computing considering failures of servers and the transmission of task datasets in data centers. This approach optimizes two conflicting objectives: minimizing energy consumption during computation and transmission, and reducing application rejections or violations due to failures. The proposed method can also improve resource utilization. The experimental simulations via large scale parallel working datasets show that this method can obtain good energy saving benefit and high quality of service.