为了解决虚拟计算环境中的资源合理调度问题,提出了一种基于信任的资源匹配模型——“资源滑动窗口”模型。首先对资源的静态属性进行分类,然后依据基于时间窗的贝叶斯信任模型对资源节点进行评价,同时考虑资源的负载,动态划分其实时性能。最后综合评估静态和动态属性,确定调度资源分配。该模型为不同任务和属性的资源调度策略奠定了基础,实现了“合适的资源服务于合适的任务”的目的。仿真实验表明所提的模型相对传统的调度算法,具有更高的系统任务执行成功率和资源利用率。
In order to solve the problem of reasonable resource scheduling in the virtual computing environment (iVCE), a resource matching model, based on trust "resource sliding window" model, was proposed. Firstly, the resources were classified in advance by their static properties. Then the resource nodes were evaluated by a trust model based on the time window and Bayesian algorithm, and the evaluated results as well as the resource load were fused to determine the levels of the dynamic real-time performance. Finally, the scheduling resources were allocated according to the comprehensive assessment of the static and dynamic properties of the resource nodes. The proposed model provides a basis of scheduling strategy for different resources with different tasks and attributes, thus achieving the goal of "proper resources' serving proper tasks". Simulation results show that, compared with the previous scheduling algorithm, the proposed model has better performance in the success rate of task execution and the rate of resource utilization.