摘要
基于“天镜”系统对省级业务系统融入的要求,分析广西区域自动气象站中心站数据监控运维痛点,构建覆盖资源层、应用层、传输层的标准化监控体系。通过经验阈值法与标准参照法,确定涵盖CPU、内存、网络、日志等节点的19项监控指标的三级阈值,依托python开发环境,采用轻量级分层架构开发AWS-SCAN采集模块,通过超阈值数据处理及告警机制实现故障感知和分级告警。采集模块在广西3个省级中心站部署应用。结果表明,中心站监控数据实现5 min高频采集与“天镜”系统DI/EI规范对接,故障响应时间从45 min缩短至10 min内,为省级气象业务系统标准化运维提供可推广的技术方案。
In response to the integration requirements of the“Tianjing”system for provincial meteorological operational systems,this study analyzes the pain points in data monitoring and maintenance of the central terminal of regional automatic meteorological stations in Guangxi,and build a standardized monitoring framework covering resource,application,and transmission layers.Using the empirical thresholding and standard reference methods,three-tier thresholds are determined for 19 monitoring indicators across nodes including CPU,memory,network,and logs.Leveraging a Python development environment,the AWS-SCAN collection module is developed with a lightweight hierarchical architecture to enable fault detection and hierarchical alerting through thresholdexceedance data processing and alert mechanisms.The collection module has been deployed in three provincial central terminals in Guangxi.The results show that demonstrate high-frequency(5-minute interval)data collection aligned with the Tianjing system’s DI/EI specifications,reducing fault response time from 45 minutes to within 10 minutes.This provides a scalable technical solution for standardized maintenance of provincial meteorological operational systems.
作者
陈德诚
阮俊淞
李莹
CHEN Decheng;RUAN Junsong;LI Ying(Guangxi Meteorological Information Center,Nanning 530022,China)
出处
《气象研究与应用》
2025年第2期133-139,共7页
Journal of Meteorological Research and Application
基金
广西气象科研计划项目(桂气科2023M05)
广西气象大数据实验室项目。
关键词
区域自动气象站
状态采集
“天镜”融入
标准化
regional automatic weather station
state acquisition
“Tianjing”integration
standardization
