实时高可用测控集群关键技术研究

针对实时测控信息类型复杂、信息处理和控制方法多样、对实时性和可靠性要求高等特点,构建了一种基于集群计算平台的实时、高可用测控系统,在此基础上提出了一主一备双工热备节点的分布式集群管理控制模式,即在系统运行过程中任务除在一个节点上进行处理外,另一节点同时运行该任务处理(但不输出处理结果),在此基础上设计了任务故障监测的心跳机制、任务故障迁移和恢复的分配方法。通过实验证明,该方法能在一个处理周期内发现任务故障,并实现故障任务的迁移和恢复,不仅减少了系统建设成本,而且缩短了故障任务迁移和恢复时间,提高了测控系统的实时性、可靠性。

Key Technologies for Realtime High Availability TT&C Cluster Systems

A realtime high availability TT&C (Tracking, Telemetry and Command) system based on cluster computing platform is constructed in the backdrop of complicated types of realtime TT&C information, diversified information processing and control methods, realtime performance and high availability requirements.A distributed cluster management control mode featuring duplex hot standby nodes is proposed.During system operation, a task is processed simultaneously at two nodes, but only one node outputs the processing results.Based on this, a heartbeat mechanism of task fault monitoring and a distribution method for task migration and restoration are designed.The experimental results show that the method finds faults in one task processing cycle and fulfills migration and recovery of fault tasks.This not only reduces system construction costs, but also shortens the time of fault task migration and recovery.Moreover, it enhances the realtime performance and reliability of TT&C systems.