基于Hadoop技术的批量录波数据高性能处理方法  

Research on High-performance Treatment of Batch Wave Recording Data Based on Hadoop Technology

作　　者：张琳波 李本瑜 翟海燕 廖晓春 陈益[3] 

机构地区：[1]云南电力调度控制中心,昆明650011 [2]武汉顺承科技有限公司,武汉430071 [3]湖北工业大学理学院,武汉430068

出　　处：《云南电力技术》2015年第2期135-138,共4页Yunnan Electric Power

摘　　要：介绍了批量录波数据的特点及产生原因,分析了现有应对措施及存在不足,并提出基于Hadoop虚拟化和分布式技术改善和优化主站性能的方法,采用通信服务矩阵替代原双重化模式下单一服务器来承载批量数据处理任务,保证了主站系统整体负载平衡,有效降低了CPU负荷率。This paper describes the characteristics and causes of batch wave recording data, analyses the existing treatments and shortcomings, proposes Hadoop-based virtualized and distributed technologies to improve and optimize the performance of Main Sta-tion System, which adopts communications service matrix to replace the originally single server mode to Carry batch data processing tasks, in order to ensure a integral load balance of the main station system, effectively reduce the CPU load rate.

关 键 词：主站系统 通信服务矩阵 CPU负荷率 

分 类 号：TM76[电气工程—电力系统及自动化]