Cluster Based Analysis of IPv6 Network Traffic

摘要

Classical IPv6 network traffic analysis system is built on one or more physical servers. It has low analysis efficiency, large process delay, and it could not reach the real-time demand. Cluster based IPv6 network traffic analysis system is composed of Front-analysis Node, Controller Node, Back-analysis Node and Data Warehouse Node. The functions of cluster-based analysis system are traffic collection, real-time analysis and off-line traffic analysis. Experimental results show that the Average Waiting Time (AWT) for jobs in case of Service Cluster is comparatively lesser with respect to Physical Server, and the Service Cluster shortens the service failover time by 89.8%.