Abstract: Increasing complexity in broadband access network development is a great challenge for network management system (NMS). Fault management is the core of network management. This paper presents a new algorithm-event stream algorithm. It takes use of fault polling mechanism and sets up a listen daemon to receive trap message from management agent (MA). This algorithm has high efficiency and good outcome.

摘要： 随着宽带接入网络技术的发展，如何有效地进行网络管理的问题日益突出。故障管理是网络管理的核心内容。文章提出了将故障轮询机制与监听来自管理代理的Trap消息来检测故障的方法相结合的事件流分析方法。该算法很好地实现了故障检测的功能。

关键词: 事件流, 故障诊断, 轮询, Trap, SNMP

CLC Number: 

• TN915.61