NetFlow流量采集与存储技术的研究实现  

Study on NetFlow-based network traffic data collection and storage

作　　者：陈宁[1] 徐同阁[1] 

机构地区：[1]北京航空航天大学软件开发环境国家重点实验室,北京100083

出　　处：《计算机应用研究》2008年第2期559-561,564,共4页Application Research of Computers

基　　金：国家"863"计划资助项目(2004AA1Z2240)

摘　　要：为解决高速大流量网络的流量监测与分析问题,提出了一套基于NetFlow的流量采集与存储方案,设计了多线程与双链表的NetFlow数据采集机制,有效提高了数据采集效率和可靠性。此外,在存储原始NetFlow数据的基础上,设计了一套NetFlow流量的三级聚合和存储方案。基于此方法可对纷繁复杂的原始流量信息进行有效整理,为前端静态、动态流量分析提供合理高效的数据支持。A network traffic data collection and storage solution based on NetFlow was presented, which was used to solve the problem of large traffic data analysis caused by high-speed network. By using multi-thread programming technique and doublelists cache structure, the efficiency and dependability of NetFlow data collection could be enhanced. Furthermore, a three-level aggregation and storage method based on the original NetFlow data was designed. By this method, the original flows could be sorted and simplified in order to offer a reasonable and efficient data support to the front static and dynamic analysis application.

关 键 词：NETFLOW 流量采集 双链表 流量聚合 时间粒度 

分 类 号：TP393.09[自动化与计算机技术—计算机应用技术]