基于多采样频率云环境监控调度策略  

Cloud Computing Monitoring Schedule Strategy of Multi-sample Frequency

作　　者：马文涛[1] 韦海亮[1] 王宇[1] 周浩杰[1] 

机构地区：[1]江南计算技术研究所,无锡214000

出　　处：《高性能计算技术》2012年第5期46-51,共6页

摘　　要：云计算是一种新型的IT基础设施构建及服务模式，云环境具有资源种类多、规模大、高弹性等特征，对云环境的高效监管是其可靠运行的基础。云环境下监控调度的状态信息采样频率直接关系到故障发现的效率和被监控资源及网络的负载，本文比较了当前广泛使用的四种主要监控调度模式的优点与缺点，在主动监控调度的基础上提出了一种多区间多采样频率监控调度策略，并对该策略进行了性能分析评估及相关优化。Cloud computing is a new type of IT infrastructure and service delivery. The cloud environment is characterized by the variety of resources, large-scale and high elastic. The efficient supervision of the cloud environment is the basis of reliability of operation. The sampling frequency of monitoring and dispatching status information in the cloud environment is directly related to the efficiency of fault detection and the workload of resources and network. This paper compares the advantages and disadvantages of four current widespread use of monitoring and dispatching mode, then base on basis of active monitoring and dispatching proposed a multi-range multi-sampling frequency of monitoring and dispatching strategy, and analysis the strategy performance. At last we gave some related optimizations.

关 键 词：云计算 资源监控 调度策略 数据采集 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]