基于任务复制的多维QoS云计算任务调度

发布时间：2018-07-31 16:53

【摘要】：为了满足云计算环境下用户服务质量(QoS)需求和提高虚拟资源空闲时间段的利用率,提出了一种基于任务复制的多维QoS任务调度策略。首先,构建云资源模型和用户QoS模型,然后根据虚拟资源的利用情况和QoS的满意度对虚拟机进行性能测评,选择综合性能更高的虚拟资源进行任务的分配;在任务执行时为了缩短任务的完成时间,在调度过程中引入了在空闲时间段复制父任务的方式。通过仿真实验将该算法与HEFT、CPOP进行比较,实验结果显示:当用户偏好可靠性执行时,该算法平均可靠性比HEFT和CPOP高;当用户偏好完成时间和费用花费执行时,该算法平均完成时间比HEFT和CPOP少;当用户无偏好执行时,该算法平均完成时间和平均花费均比HEFT和CPOP少。结果表明该算法能有效提高资源利用率和用户的满意度。
[Abstract]:In order to meet the requirement of (QoS) in cloud computing environment and improve the utilization of idle time of virtual resources, a multi-dimensional QoS task scheduling strategy based on task replication is proposed. Firstly, the cloud resource model and the user QoS model are constructed, then the virtual machine is evaluated according to the utilization of virtual resources and the satisfaction degree of QoS, and the virtual resources with higher comprehensive performance are selected for task allocation. In order to shorten the completion time of the task, the method of replicating the parent task in the idle time is introduced in the scheduling process. The experimental results show that the average reliability of the algorithm is higher than that of HEFT and CPOP when the user preference reliability is executed, and when the user prefers the completion time and cost to execute, the simulation results show that the algorithm is more reliable than HEFT and CPOP. The average completion time of the algorithm is less than that of HEFT and CPOP, and the average time and cost of the algorithm are less than that of HEFT and CPOP when the user has no preference. The results show that the algorithm can effectively improve the resource utilization and user satisfaction.
【作者单位】：江南大学物联网工程学院;轻工过程先进控制教育部重点实验室(江南大学);
【基金】：国家自然科学基金资助项目(61170120) 江苏省自然科学基金资助项目(BK2011147)
【分类号】：TP393.09

【相似文献】