本文關鍵詞： 基于策略的網絡管理 可擴展 策略描述 信息提取 鏈路發(fā)現(xiàn)　出處：《國防科學技術大學》2014年碩士論文　論文類型：學位論文

【摘要】：隨著科學技術迅猛發(fā)展,計算機網絡在人們生活中得到越來越廣泛應用,各高等院校也紛紛建立專門的校園網站。校園網規(guī)模不斷擴大、復雜性不斷增加,傳統(tǒng)網絡管理手段正逐漸暴露出其弊端和不足。傳統(tǒng)網絡管理要求管理人員熟悉網絡設備配置命令,了解各型網絡設備配置差異,能夠根據不同網絡需求動態(tài)調整配置,這一切使得網絡管理的效率和正確性難以得到保障�；诓呗缘木W絡管理是解決上述問題的有效手段,也是當前的研究熱點。論文在分析比較現(xiàn)有基于策略的網絡管理體系結構、策略信息模型、智能信息提取等技術基礎上,重點研究信息統(tǒng)一描述與映射算法以及信息提取的鏈路拓撲發(fā)現(xiàn)算法,并設計實現(xiàn)了一個可擴展的校園網策略管理系統(tǒng)。首先,針對現(xiàn)有基于策略的網絡管理缺乏可操作的統(tǒng)一策略描述標準,對外來預知信息依賴程度高,系統(tǒng)可擴展性差等問題,論文提出了UIDR(Unified Information Description and Representation Method)信息統(tǒng)一描述與映射方法。該方法包含兩部分內容:一是用于描述配置策略信息的策略信息統(tǒng)一描述和映射方法;二是用于描述設備配置時需要的相關信息的設備信息統(tǒng)一描述與映射方法。其中策略信息和設備信息的統(tǒng)一描述采用XML技術,策略翻譯和信息提取采用Perl腳本技術。UIDR方法建立策略信息和設備信息的統(tǒng)一描述方式,開放描述與映射接口,有效屏蔽了設備的命令差異,第三方設備可遵循接口規(guī)范編寫腳本,顯著提高了系統(tǒng)的可擴展性。UIDR方法的設備信息反向提取技術,有效降低了系統(tǒng)對第三方信息的依賴程度,提高了系統(tǒng)的可用性。其次,針對設備信息提取過程中從單臺設備信息無法獲得網絡管理域的物理拓撲連接關系的問題,論文提出一種ATDIE(the Algorithm of Topology Discovery based on Information Extracting)基于信息提取的鏈路拓撲發(fā)現(xiàn)算法。該算法基于單臺設備的接口信息、ARP信息,通過遞歸查找不同設備的兩個端口的IP地址是否屬于同一網段,并且通過ARP信息判斷兩個端口是否直連,從而獲得網絡管理域的完整拓撲信息。ATDIE算法有效利用了單臺設備的提取信息,為物理鏈路層的拓撲發(fā)現(xiàn)提供了新的解決途徑,也為維護管理網絡提供了便利。最后,論文設計實現(xiàn)了一個可擴展的校園網策略管理原型系統(tǒng),系統(tǒng)包括用戶界面、策略分解與派發(fā)器、策略執(zhí)行與驗證器、態(tài)勢展示模塊等。策略分解與派發(fā)器中的信息提取模塊和策略翻譯模塊不直接操作數(shù)據庫,便于第三方廠家的擴展,提高了系統(tǒng)的可擴展能力,以及系統(tǒng)穩(wěn)定性。實驗結果表明:該系統(tǒng)達到了預期的設計要求,可對校園網進行有效配置管理。
[Abstract]:With the rapid development of science and technology, the computer network has been more and more widely used in people's life. Colleges and universities have also set up special campus websites. Traditional network management means are gradually exposing its disadvantages and shortcomings. Traditional network management requires managers to be familiar with network equipment configuration commands, to understand the differences of network equipment configuration, and to be able to dynamically adjust configuration according to different network requirements. All these make it difficult to guarantee the efficiency and correctness of network management. Policy-based network management is an effective means to solve the above problems. This paper analyzes and compares the existing policy-based network management architecture, policy information model, intelligent information extraction and other technologies. This paper focuses on the unified description and mapping algorithm of information and the link topology discovery algorithm for information extraction, and designs and implements an extensible campus network policy management system. In view of the lack of an operational unified policy description standard for policy-based network management, the high degree of dependence on external predictive information and the poor scalability of the system, etc. This paper proposes a unified description and mapping method of UIDR(Unified Information Description and Representation method, which consists of two parts: one is the unified description and mapping method of policy information used to describe the configuration policy information; The second is the unified description and mapping method of device information, which is used to describe the relevant information when the device is configured, in which the unified description of policy information and equipment information is based on XML technology. Policy translation and information extraction using Perl script technology .UIDR method to establish a unified description of policy information and device information, open description and mapping interface, effectively shield the device command differences. Third-party devices can follow the interface specification to write scripts, which can significantly improve the extensibility of the system. UIDR method of equipment information reverse extraction technology, effectively reduce the system's dependence on third-party information, improve the system's usability. In order to solve the problem that the physical topological connection of network management domain can not be obtained from single device information in the process of equipment information extraction, This paper presents a link topology discovery algorithm based on information extraction for ATDIE(the Algorithm of Topology Discovery based on Information extraction. The algorithm is based on the interface information of a single device and finds out recursively whether the IP addresses of two ports of different devices belong to the same network segment. And through the ARP information to judge whether the two ports are directly connected or not, the complete topology information of the network management domain. ATDIE algorithm effectively utilizes the information extracted from a single device, and provides a new solution for the topology discovery of the physical link layer. Finally, the thesis designs and implements an extensible campus network policy management prototype system, which includes user interface, policy decomposition and distribution, policy execution and verification. The information extraction module and the policy translation module in the policy decomposition and dispatcher do not operate the database directly, which is convenient for the expansion of the third party manufacturers and improves the extensibility of the system. The experimental results show that the system meets the expected design requirements and can effectively configure and manage the campus network.
【學位授予單位】：國防科學技術大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TP393.07
，

本文編號：1526573