Design And Implementation Of Scalable Campus Network Policy Management System

With the rapid development of science and technology, computer network is becoming more and more popular in people’s life. Many colleges also set up their own campus websites. As the scale and complexity of campus network are increasing, traditional network management gradually exposes its shortcomings. Traditional network require the manager to be familiar with network devices configuration commands and to know each network device configuration differences so that they can adjust the dynamic configuration according to different network, all of which makes it hard to maintain the efficiency and accuracy of network management. Policy-based network management is an effective measure to solve the above problems and is being paid more and more attention.Based on the analysis of comparing existing policy-based network management system structure, including strategies, information model and intelligent information extraction technology, this thesis focused on the universal description and mapping algorithm and link topology discovery algorithm of information extraction and implement a scalable network strategy management system.First, as the existing policy based network management lacks the unification of the operational strategy Description standard and high degree of dependence on foreign Information which lead to a weak scalability problem, the UIDR(Unified Information Description and Representation Method) Information Unified Description and mapping Method is proposed. The method is consisted of two parts: one is configuration policy based policy information unified description and mapping method. The other is used to describe the information needed for the equipment configuration of equipment information unified description and mapping method. The strategy information and equipment information unified description uses XML technology. Translation strategies and the technology of information extraction use Perl script. UIDR method establishes the unification of the policy information and equipment information description and opens the interface of description and mapping which effectively shield the differences of equipment command and third party equipment therefore significantly improve the scalability of the system. Information reverse extraction technology of UIDR method effectively reduces dependence on the system of third party information which increases the availability of the system.Secondly, in the process of Information extraction from a single device Information, the equipment can’t access to the network management domain of physical Topology relationship. This thesis puts forward ATDIE(the Algorithm of Topology Discovery-based on Information Extracting) link Topology Discovery Algorithm which is based on Information extraction. The algorithm is based on single device interface information, the ARP information. By recursively judging whether the IP address of different equipment is belong to the same network and whether the two ports directly connected through the ARP information, the full topology information of the network management domain is obtained. ATDIE algorithm effectively utilizes the single equipment information which solved the problems of physical topology discovery of a link and has the advantages of strong practicability.Finally, the thesis designed and implemented a scalable network strategy management prototype system. The system includes user interface, strategies, decomposition and distributing device, policy enforcement and validations, displaying module, etc. The information extraction module and strategies translation module of the strategy decomposition and distributing device does not directly operate database, which is beneficial to the scalability of the third party manufacturers and improve the stability of the system. Experimental results show that the system reaches the expected design requirements and can effectively configure the management of campus network.