The Analysis And Design Of Performance Ensurance Approach For The High-Reliability LAN

With the continual emergence of new network technology, and unceasing development of new network service, the demands on network performance become more and more high. Meanwhile, in the wake of further expanding the scale of the network, the maintenance of network high performance operation becomes more and more difficult. Therefore, the testing and evaluation of network performance are increasingly valued by people.The topic of performance safeguard plan is proposed according to this kind of condition, its core idea is to ensure maximum network performance at the design stage by thoroughly analysis of practical network application situation and each kind of factor which affects network performance at the beginning of the network design, and to monitor network actual operation situation through the specific network performance monitoring software after the network construction is completed.First, this thesis studied and summarized the network performance metric, analyzed the main factors affecting network performance and the corresponding network performance measurement technology and evaluation methods. Then, based on the needs of network performance in the project, this thesis analyzed and studied the LAN high-performance ensurance approach needed in such industries as the military, industrial control, negotiable securities, finance and so on. In view of high reliable system network performance metrics, this thesis proposed a high-performance LAN network design and implementation process. Then this thesis discussed a one-way network delay measurement approach which integrates the distributed test and the centralized control ways, including the discussion of the significance of the one-way delay measurement, measurement method, measurement architecture and implementation mechanisms.Finally, this thesis designed and realized a performance monitoring system which runs on the Windows platform developed with VC6.0 for the particular network architecture, and uses SNMP and active measurement ways for the management data collecting. This thesis discussed the realization process of the performance data acquisition module, performance data processing module, and network topology discovery module in detail.