论文部分内容阅读
针对现有的云计算任务调度策略仅考虑单数据中心内部负载均衡及平等看待各项任务的问题,研究了基于服务等级协议(service level agreements,SLA)的多云数据中心任务调度机制,设计了相应的任务调度效益模型和任务准入控制策略,提出了基于SLA的最大化收益任务调度算法(SLA-MPS算法),实现了在多个云数据中心间调度资源、优先处理紧急任务的同时保证云服务商利益最大化。在CloudSim上的实验证明,SLA-MPS算法能加快任务响应速度,降低云服务商违约率并提高其收益。
In view of the existing cloud computing task scheduling strategy, we only consider the issues of internal load balancing and equal treatment of tasks in a single data center. The task scheduling mechanism of multi-cloud data center based on service level agreements (SLA) is studied. Correspondingly, Task scheduling benefit model and task admission control strategy, a SLA-based maximized revenue task scheduling algorithm (SLA-MPS algorithm) is proposed, which enables resource scheduling among multiple cloud data centers, prioritizing emergency tasks and guaranteeing cloud Service providers to maximize the benefits. Experiments on CloudSim prove that the SLA-MPS algorithm can speed up the task response, reduce the default rate and increase the profit of cloud service providers.