基于Zabbix的地震数据计算集群监控策略  

Developing A Calculation Cluster Monitoring Strategy for Seismic Data Based on Zabbix

作　　者：余卫江 林威扬 宋志君 陈雄斌 YU Weijiang;LIN Weiyang;SONG Zhijun;CHEN Xiongbin(China Oilf ield Services Limited Geophysical BU Geophysical Research Institute Tepu Data Center(Zhanjiang),Zhanjiang 524057,China)

机构地区：[1]中海油田服务股份有限公司物探事业部物探研究院特普数据中心(湛江),广东湛江524057

出　　处：《化工管理》2024年第19期74-77,共4页Chemical Management

摘　　要：计算机集群在地震储存和处理领域运用广泛,被誉为地震勘探领域的硬件基石。当服务器中任意环节发生故障且不能准确定位迅速恢复时,可能引发整个生产业务链瘫痪。文章提出了一种基于Zabbix的地震数据多源计算集群监控策略,其中基于Zabbix平台的监控引擎实现了多系统源服务器资源监测,通过二次开发实现了Web页面大屏显示,基于K-means算法完成了故障数据预处理以及自动恢复功能。经实践,该系统监控功能完备,实现了资源的多维度显示和故障智能处理等功能,整体监控效果良好。Computer clusters are widely used in the f ield of earthquake data storage and processing,and are considered the basic hardware in seismic exploration.When any part of the server fails and cannot be accurately located for quick recovery,it may cause the entire production business chain to be paralyzed.This article proposes a seismic data multi-source computing cluster monitoring strategy based on Zabbix.The monitoring engine,which is based on the Zabbix platform,realizes resource monitoring of multiple system source servers.Achieved a large-screen display for web pages via secondary development,and completes fault data preprocessing and automatic recovery functions using the K-means algorithm.In practice,the system’s monitoring capabilities are comprehensive,achieving multi-dimensional resource display and intelligent fault handling,with the overall monitoring performance being satisfactory.

关 键 词：Zabbix监控引擎 K-MEANS算法 集群监控 

分 类 号：P631.4[天文地球—地质矿产勘探]