基于流计算的实时监控预警架构研究及应用
Real-time Monitoring and Warning Architecture Based on Stream Computing
朱征 1于帅鹏 2赵志刚 3吴金龙1
作者信息
- 1. 国网上海市电力公司信息通信公司,上海200436
- 2. 国网电力科学研究院北京中电普华信息技术有限公司,北京100192
- 3. 聚云浩海科技有限公司,苏州215200
- 折叠
摘要
随着国家电网IT建设的迅速发展,大数据问题逐步呈现.根据国家电网运营监控系统中遇到的实际问题,提出一种基于流计算的实时监控架构,通过试验,该系统具有极强的扩展性,能够在百GB级别数据量下,获得秒级的实时监控响应,对比传统的批处理架构,基于流计算的实施监控架构在监控实时性、集群扩展性上具有更大的优势.
Abstract
With the rapid development of Information Technology construction in China,the State Grid is facing unprecedented challenges from big data deluge.In view of the actual problems in the national power grid operation monitoring system,this paper proposes a real-time monitoring architecture based on stream computing.Experiments have shown that the system has strong scalability,capable of acquiring second-grade real-time monitoring response with the hundred-GB data volume.Compared with the traditional batch processing architecture,this proposed framwork based on stream computing is superior in real-time monitoring and cluster extensibility.
关键词
流计算/电力系统/实时监控/大数据Key words
stream computing/power system/real-time monitoring/big data引用本文复制引用
出版年
2014
NETL
NSTL
维普
万方数据