针对耦合分布式系统中一个计算模块独自占用某台计算资源,导致其他计算模块无法调度到该计算资源的情况,提出了一种动态任务分配的调度算法.该算法能够根据计算任务的调度要求和计算资源的运行状态,动态进行任务分配,使计算能力强的计算资源能够运行更多的计算模块,从而实现多计算任务调度,使多个符合调度要求的计算任务同时处于运行状态,提高计算资源的利用率,保证了计算任务调度的有效性和高效性.仿真结果表明,该多任务动态调度算法能够在不影响计算速度的情况下,使一台计算资源同时为多个计算任务提供计算能力,大大提高了计算资源的利用率,并且使原本因计算资源的限制而无法运行的计算任务能够提前开始运行,提高了计算任务调度的高效性和灵活性.%In a coupling-distributed system,one calculation task is distributed in a single computing node so that other tasks cannot be dis-tributed in the same computing node,even with a strong computing power. A new dynamic task assignment algorithm is proposed. It can assign calculation task according to the scheduling requirements of computing task and running state of computing resources,which makes calculation resources with powerful computing capable of running more calculation modules for implementation of multi-tasks schedu-ling,and which enables computing tasks of meeting scheduling needs in the running state simultaneously, improving the utilization of computational resources and ensure the effectiveness and efficiency of scheduling. The simulation shows that the proposed algorithm can greatly improve the utilization of computing resources,not only ensuring the computing speed but also providing computing power for multiple computational tasks at the same time. The computational tasks that cannot be performed due to conditional constraints can run in time,thus improving the efficiency and flexibility of the scheduling system.
展开▼
