Snmp-based Enterprise Backbone Network Monitoring And Management System

With the high-speed development of computer network technology, network scale becomes more and more wide and network structure grows more and more complex. So network management has been the hottest spot in network researches and it has influenced a lot in the development of network. ISO has defined the five functions of network management: performance management,accounting mangment,security mangment,configuration mangment and fault mangment and they have been widely accepted. The wide use of Control Protocol/Internet Protocol(TCP/IP) networking technology has led to the origin and development of network management. Simple Network Management Protocol(SNMP) has been widely used and has actually become a industrial standard for TCP/IP-based network.At present, in network performance management, the accuracy of analysis and evaluation from the data connected with network performance needs improvement. As for network fault mangment,any documents can not exactly define the reson for deficiency that the MIB(Management Information Base)data probably reflects because of its complexities. So it is necessary for us to research these problems.Based on the intensive studies of SNMP, the author designs a SNMP-based enterprise backbone network monitoring and management system aiming at the enterprise users of the Internet and discusses the plan of each part of the system. Seeing that both SNMPvl and SNMPv2 exist at present, the MIB data acquisition containing SNMPvl and SNMPv2, MIB Browser, is designed and realized. All the MIB data of managed equipments which own SNMP agents can be obtained through using MIB Browser. The network management information database is established. The operations of the database are achieved by DAO programing. In performance management, the paper presents the mathematic method of performance analysis and evaluation in order to increase the precision of valuation . The mathematic method is used in the system realization and its effect is good. In fault managemeng, the thesis makes detailed discussion on fault diagnosisjeason for deficiency and failure prediction. As far as reason for deficiency, the fault expert database is built for helping administrators determine the reason for deficiency;. As for failure prediction,the mathematic models,the logarithmic normal distribution model of network software faults and the exponential distribution model of network hardware faults,are established. In accounting(flow) management,the mathematic formulas on counting the flow are raised. In configuration management, a lot of MIB data that can be used for configuration management are listed. Due to the limited time, performancemanagement,accounting(flow) management and a part of fault management are realized on Windows2000 platform. There may be some mistakes in this article because what the author knows is limited. Welcome to give some advice if you'd like to.