Network Fault Management System Based On Expert System

In this paper, the whole infrastructure and the function of each part of fault management module in network management system are described and analyzed in detail. The following are the functions designed and implemented in this system: fault polling, fault accept, fault log and intelligent fault diagnosis.Fault management is a key part in the five function fields of network management. At present network fault management is mostly traditional fault diagnosis based on instance. Although this kind of fault management takes the effect, it is short of the diagnosis correctness as well as the value on reuse. Therefore, it can do nothing with the new style fault. Hence, a network fault management based on expert system was proposed. A network fault diagnosis system based on knowledge base and inference enginery was designed and implemented as well.The Inference enginery that is designed in the paper has four characteristics: non-recursive deep first search strategy; the strategy of collision detection based on time standard; control strategy combining with data and goal drive; recoding inference process based on storage mode by adopting data structure variable.