摘要
天津市地震局通过"九五"、"十五"数字地震观测网络的建设,实现了测震台站数据网络化传输,将波形数据实时存储于AWS数据存储服务器中,用户通过人机交互系统从存储数据中调取波形数据进行分析并提交分析结果。测震系统在日常运行时会对存储服务器端软硬件造成一定程度的负荷,影响系统的整体应用效能。针对这一问题,利用Python时间序列算法实现对AWS数据存储磁盘容量进行预测,判断存储服务器的负载率情况,避免应用系统由于存储容量耗尽造成的系统瘫痪。根据该算法同时可监控台站实时数据量,生成台站数据流量变化曲线,从而成为一种评估台基噪声情况的新方法。
The Earthquake Agency of Tianjin Municipality realized a seismic data transmission network by constructing a digital seismological observation network.The seismic waveform data is real-time stored in the AWS data server,and users can call the data from the AWS server by a man-machine interface system.Therefore,the seismic system operation generates load in the server hardware and software,thus affecting the system overall efficiency.In this study,we use a Python-based algorithm of time series to predict the disk capacity of the AWS server and estimate the load rate of the storage server,thus avoiding the occurrence of system paralysis because of storage capacity depletion.The real-time data of stations can also be monitored by the algorithm,and variation curves of station data can be accordingly generated.
作者
孙路强
刘磊
卞真付
许贺
SUN Luqiang;LIU Lei;BIAN Zhenfu;XU He(Earthquake Agency of Tianjin Municipality,Tianjin 300201,China)
出处
《地震工程学报》
CSCD
北大核心
2018年第A01期170-173,196,共5页
China Earthquake Engineering Journal
基金
中国地震局监测
预测
科研"三结合"课题(160201)
