【摘要】 随着高性能运算技术的发展,对于一个超过2800CPU以上多套PC集群的超大型计算中心,开发一体化的、能够全面监测系统运行状态的集群监控软件是十分必要的。通过大规模多架构PC集群一体化动态监控软件的研发,实现了在统一界面下,实时、动态地监控多套、不同架构PC集群各个节点的处理器、内存、网络和磁盘等关键子系统的性能数据和运行状态,同时,对各子系统的异常情况提供实时、准确的报警功能,对提高集群整体性能起到至关重要的作用。更多还原

【Abstract】 With the development of high-performance computing technology,for a super computing center of more than 2800CPU multiple PC clusters,it is very necessary. To develop integration,cluster monitoring software which can fully monitor the status of whole system running status. Through the development of large-scale architecture of PC cluster integration dynamic monitoring software,implemented it a unified interface, real-time, dynamic monitor all of the PC cluster node about processor,memory,network and disk and other key subsystem performance data and operation state,meanwhile,the abnormal situation of each subsystem provides real-time,accurate alarm function,a crucial role to improve the overall performance of cluster.更多还原