针对网元短时间大量接入网管系统产生大量告警并导致服务器宕机的现象进行了分析,深入讨论了综合网络管理系统中故障管理系统的体系结构,分析了告警风暴和告警振荡的产生原因,发现目前的网管系统无法对网络中出现的故障信息的真实性进行有效的判断,从而提出了有效的解决方案。 This paper analyzes the phenomenon that a large number of NEs access a large number of network management systems in a short time and causes server downtime. The architecture of fault management system in integrated network management system is discussed in depth. The causes of alarm and alarm oscillation are analyzed. It is found that the current network management system can not effectively judge the authenticity of the fault information appearing in the network, and thus puts forward an effective solution.