摘要
为了对大型电信数据网进行综合统一管理,设计并开发了基于SNMP(S imp le Network M anagem ent Pro-tocol)的综合数据网络管理系统(Sp lenNMS:Sp len Network M anagem ent System)。从网络资源、性能、故障3个方面,着重论述了该系统实现的关键技术,叙述了网络故障信息采集方法和Sp lenNMS的体系结构。针对资源信息种类繁多、差异性大的采集问题,采用基于任务规则的设备资源配置信息采集方法;提出一种改进的性能数据采集及存储方案,解决了海量性能数据采集管理的困难。与流行的MRTG(Mu lti Router Traffic G rap-her)软件相比,该方案减小了最小执行周期;由于采用了基于Trap扩展定义的性能告警处理方法,使性能告警的处理在形式上与设备告警一致。通过在某省会城市大型电信数据网络上的试运行表明,该系统能适应2000台网络设备左右的规模。
In order to manage the large-scale data network in telecom, the implementation method of an integrated data network management system is presented. The pivotal technologies of SplenNMS (Splen Network Management System) are described in detail. The algorithms of discovering network devices and topology are shown, and a model based on task rules is also proposed in collecting network resource information, which overcomes the difficult problem in resource collection. A scheme of collecting and storing performance data is presented, which resolves the problems in management of large-scale data, and it can reduce the implementation cycle in the comparison to MRTG (Multi Router Traffic Grapher). The definition of extended Trap types is applied to handle performance alerts. Finally, a method of collecting fault management information is given, and the architecture of SplenNMS is described. As shown in our practical application, SplenNMS can meet the requirement of network with about 2000 devices.
出处
《吉林大学学报(信息科学版)》
CAS
2006年第2期113-119,共7页
Journal of Jilin University(Information Science Edition)
基金
国家科技攻关基金资助项目(2004BA907A20)
吉林省科技发展计划基金资助项目(20040304)
关键词
计算机网络
网络管理
拓扑发现
性能监测
故障管理
computer network
network management
topology discovery
performance monitoring
fault management