融合区分服务和速率调整的请求调度策略

摘要

Considering the difference of heterogeneous user demands for quality of service ( QoS ) and the energy con-sumption of cloud data centers , a request scheduling strategy integrating differentiation service and speed adjust -ment is proposed .The strategy divides an IaaS ( infrastructure as a service ) cloud data center into the rapid cloud with speed adjustment and the reliable cloud with fault restoration , then a novel hybrid cloud computing paradigm is investigated .Accordingly , a system model based hybrid queueing network is established .From the view of the av-erage response time of delay sensitive requests , the energy saving level of virtual machines in rapid clouds and the utilization of virtual machines in reliable clouds , the system performance of the request scheduling strategy is esti-mated.The effectiveness of the proposed strategy and the rationality of the established system model are verified by experiments .By constructing a system cost function and improving the Jaya intelligent searching algorithm , the pro-posed request scheduling strategy is optimized .%针对异构用户服务质量(QoS)需求不同及云数据中心能源浪费等问题,提出了一种融合区分服务和速率调整的请求调度策略.该策略将IaaS云数据中心分为提供速率调整机制的快速云和提供差错恢复机制的可靠云,然后研究新型混合云计算模式.构建了混合排队网络系统模型,从延迟敏感请求的平均响应时间、快速云中虚拟机的节能水平及可靠云中虚拟机的利用率等方面评估了请求调度策略的系统性能.进行了系统实验,验证了策略的有效性及模型的合理性.通过构造系统成本函数,改进Jaya智能寻优算法,给出了请求调度策略的优化方案.