摘要： 现有的集群监控系统不能反映云计算集群具有海量节点且高耦合的特性。针对该问题，设计一种基于热力学相空间理论的云计算专用监控系统。采用将节点运行参数向相空间映射的方式，使节点参数的变化转换为相空间中投影点的运动，从而对相空间图像和参数进行分析。测试结果表明，该监控系统可以即时有效地反映云计算集群的整体负载和运行状态，实现对云计算集群的宏观监控。

关键词: 云计算, 监控系统, 热力学, 相空间, 宏观监控, 耦合关系

Abstract: In order to solve the problem which current monitoring system has in reflecting the mass nodes and high coupling of cloud computing cluster, this paper designs a cloud computing monitoring system based on thermodynamic space theory. By projecting node status parameters onto phase space, it converts node parameters change into the movements of projective points in phase space. Consequently, macro monitoring of cloud computing cluster is realized by utilization of phase space images and parameters. Tests results show that the monitoring system can real-time and effectively reflect the overall loading and working status of cloud computing cluster, accomplish monitoring the cluster from a macro level.

Key words: cloud computing, monitoring system, thermodynamics, phase space, macro monitoring, coupling relationship

中图分类号: 

• TP301