拓扑感知应用层组播故障检测机制研究

Research on Fault Detection Mechanism of Topology-aware ALM

应用层组播需要良好的故障检测机制,才能确保组播服务的鲁棒性。现有拓扑感知应用层组播中的故障检测机制极其简单,大大削弱了拓扑感知应用层组播的系统性能。为此提出一种反向抑制主动告警故障检测机制RAA,通过提前预警避免因缓冲区没有数据而导致的服务中止,同时通过反向抑制告警策略,减少重复和错误警报数量,为故障定位排除干扰。实验证明这种机制可以获得理想的告警准确率,并能降低组播成员因故障而导致的断连延时。

Application layer multicast needs an efficient fault detection mechanism to ensure the robustness of mulficast services. The fault detection mechanism of existing topology-aware ALM protocol is very simple which greatly weakens its performance. We propose a fault detection mechanism named RAA （Reverse-inhibition Active Alarm）. The mechanism can send alerts before the buffer data are constuned away to avoid the termination of service. It can reduce the number of repetitive and wrong alerts through reverse-inhibitlon strategy at the same time. Experiments show that the mechanism can get ideal alert accuracy rate and shorten the disconnection latency.