摘要
经典的考虑滑动窗口的调度模型属于抢占式调度,调度作业间的切换浪费了系统资源,在系统过载时,过多作业调度失败,导致调度成功率较低。为解决上述问题,提出一种监控网络高可靠数据智能调度数据模型。模型通过对本地缓存高可靠数据的调度紧急性和稀缺性进行加权求和来设定调度优先级;考虑到监控网络系统采样周期的变化,结合调度算法的优先级分配实时调整采样周期以最大程度的避免抢占,提高调度成功率。实验结果表明,与考虑滑动窗口的调度模型相比,所建型在一定程度上避免了抢占的发生,作业间的切换次数大幅度减少,有效提高了重要任务的完成率。
In the classical scheduling model considering the sliding window,the switching operations waste system resource.When the system is overloaded,many job scheduling fail,resulting in low success rate.Therefore,an intelligent scheduling model of high-reliable data in monitoring network is presented.This model set the scheduling priority by weighting and summing the scheduling urgency and rareness of local high-reliability data.Considering the change of sampling period of monitoring network system and combining with the priority allocation of scheduling algorithm,this model adjusted the sampling period in real time,so as to avoid the preemption at utmost.Thus,the success rate of scheduling was improved.Simulation results show that,compared with the scheduling model considering the sliding window,the proposed model avoids the preemption to a certain extent.The switching frequency between jobs is greatly reduced,which effectively improves the completion rate of important tasks.
作者
瓦力斯·阿布力孜
Walisi Abulizi(Xinjiang University College of textiles and clothing,Urumqi Xinjiang 830046,China)
出处
《计算机仿真》
北大核心
2020年第11期431-435,共5页
Computer Simulation
关键词
监控网络
数据调度
本地缓存
高可靠数据
作业切换次数
Monitoring network
Data scheduling
Local cache
High-reliable data
Job switching times