Design And Implementaion Of Real-Time Network Analysis System Fault Analysis Functions

This paper discusses the network fault analysis function of real-time processing system. Network real-time processing system using distributed computing, stream computing, P2P and other technologies. As a distributed system, it can processing massive data in real time. The system composed of capture streaming software, fault analysis software, client adapter and user interface. Capture streaming software is responsible for data acquisition and pre-processing, which in the network environment. Fault analysis software is responsible for network fault analysis and processing. Client adapter is responsible for receiving and distributing the packets. User interface is responsible for interaction with the user.Failure analysis software can detected the transmission condition of network packet. The establishment of a typical system failure analysis model can analysis network packets timeout, network packets loss and packet path failure. Meanwhile, achieve fault analysis expansion module. The module allow customers to customize the new type of failure and new treatment failure process.Typical failure analysis model is based on Event-driven model. After pretreatment of network packets by capture streaming software, client adapter encapsulates data into events and troubleshooting module. We can analysis the network condition. Troubleshooting module is consisted of path dispatcher, IP reorganization marked dispatcher and timer.Troubleshooting Expansion Module is a combination of traditional software development method and extended scripting capabilities for handling variable processing characteristics achieved. Module uses Lua script as troubleshooting scripts. Lua script to achieve the sub-modules loaded. Users can customize the new type of troubleshooting and processes based on the needs of the network environment.The achievement of real-time failure analysis system is an innovation in network fault diagnosis technology. It is in line with the current network fault analysis of massive data-oriented trend.