摘要
针对流表的漏表项存储中产生的交换机之间拥塞的问题,提出一种基于遗传进化缓存策略的OpenFlow流表存储优化方案,首先在OpenFlow中心交换机中进行流高速缓存时态记录,构建OpenFlow流表的时态集合,对OpenFlow流表在缓存区域的特征聚集关系进行自适应控制演算,然后采用遗传进化方法进行流表项的动态平衡缓存设计,引入缓存替换策略实现OpenFlow流表存储优化,最后进行存储性能测试,结果表明,采用该方案能提高流信息的吞吐量,降低流表资源适配延时,改善了OpenFlow流表存储性能。
Aiming at the congestion between the flow table entries stored in the switch leakage problems, an optimization scheme of flow table to store the genetic evolution cache strategy is proposed based on OpenFlow, first recorded in the OpenFlow cache temporal flow the center of the switch, build the OpenFlow flow table table in the feature of temporal set, the cache area adaptive control algorithm of aggregation relationship OpenFlow flow, then the genetic evolutionary method of flow table design dynamic cache, cache OpenFlow flow table storage optimization strategy to replace, finally storage performance test results show that for OpenFlow flow table can improve the flow information throughput, reduce the flow table resource adaptation Delay, improve the OpenFlow flow table storage performance.
引文
[1]南洋,陈琳.基于客观权重确定的数据中心网络性能评估方法[J].计算机应用,2015,35(11):3055-3058.
[2]左青云,陈鸣,赵广松,等.基于OpenFlow的SDN技术研究[J].软件学报,2013,24(5):1078-109.
[3]刘中金,李勇,苏厉,金德鹏,曾烈光.TCAM存储高效的OpenFlow多级流表映射机制[J].清华大学学报(自然科学版),2014,54(4):437-442.