云平台下公共网络负载均衡调度方法研究

Research on Load Balance Scheduling Method of Public Network under Cloud Platform

针对当前云平台公共网络数据调度方法负载问题,提出了一种新型网络数据均衡调度方法。首先构建了一种新型云平台工作流调度模型,该调度模型根据时间负载和遗传算法,可以生成云平台数据流并执行分层调度,确定节点需求并建立映射关系,对应不同任务虚拟资源节点,提出ISFLA改进算法,加快完成平台数据流局部搜索和全局信息交互,依靠设计的调度模型在调度前进行数据流分层,提出优先层级,并基于分层结果通过虚拟机执行最优选项,完成数据均衡调度,达到负载均衡的目的。实验数据证明,与常用的PSO调度模型相比,设计的均衡调度方法可以有效缩短调度时延,并具有更高的数据流传输完整度,具有应用优势。

Aiming at the load problem of the current cloud platform public network data scheduling method,a new network data balancing scheduling method was proposed.First constructed a new type of cloud platform workflow scheduling model,the scheduling model based on time,load and genetic algorithm can generate cloud platform and implement layered dispatching data flow,determine and establish the mapping relationship between node demand,virtual resource nodes corresponding to different tasks,ISFLA improved algorithm is put forward,to speed up to complete the global search and local information platform data flow interaction,rely on the design before the dispatch of the dispatching model for the data flow stratification,priority level is put forward,and based on layered results through the virtual machine to perform the most preferences,complete the data balance scheduling,achieve the goal of load balance.Experimental data show that compared with the common PSO scheduling model,the designed balanced scheduling method can effectively shorten the scheduling delay,and has higher data flow transmission integrity,and has application advantages.