EAST Experiment Data Storage Monitoring System Based On SpringBoot

With the continuous progress of EAST experiments,a large amount of experimental data,diagnostic data and engineering data will be generated and stored on the MDSplus(Model Drive System Plus)server in real time.At present,the storage scale has reached PB level,however,the number of data files stored in the file system of the MDSplus server is huge,the storage locations are disorderly,the distribution is uneven,mixed with many invalid files,and there is no reliable monitoring.It is easy to cause data loss or duplication and no leak detection.In addition,the storage volume and semaphore size of the statistical sub-experiment still need to be handled through the traditional command line under Linux,which is not easy to operate.At this time,the monitoring of data storage status and user behavior becomes particularly important.Therefore,it is necessary to build an EAST experimental data storage monitoring system on the existing MDSplus storage server,which can centrally manage and monitor these data files,quickly statistically analyze the effective information of each experimental data,and help predict the experimental development trend,and it can promptly warn the abnormal state of the experiment and effectively prevent data loss or overflow.This paper introduces the design and implementation of the EAST experimental data storage monitoring system,and make plans to improve system functions.The system architecture includes EPICS+Linux driver layer,data monitoring layer,data processing layer,statistical analysis layer,display layer,and function layer.Each layer provides corresponding interfaces,which reduces the degree of coupling of the system.The system recursively reads the massive data file system on the MDSplus server disk,performs a series of operations such as splitting,stitching,and reorganizing the file information to obtain metadata describing the status of the data file and stores it into the relational database MySQL to support metadata Storage,multiple copies,and secure backups.Then,a series of SQL operations are used to perform statistical analysis on the metadata to obtain valuable experimental information to realize the information mining of massive metadata.The web system of the entire system uses front-end and back-end separation.The back-end uses the SpringBoot+Nginx+Mybatis technology stack to visualize the statistical analysis layer data.The web front-end is developed using pure static web page html+js+css,and cooperates with the Bootstrap framework and Echarts plug-in to display the data in the form of charts,making the entire web UI design more convenient,fast and beautiful,and intuitively reflecting the entire EAST data storage status to help laboratory personnel quickly analyze the experimental conditions,grasp effective information,and warn of abnormal conditions.The EAST data storage monitoring system provides a standardized monitoring solution for the entire EAST experimental data.The test shows that the EAST experimental data storage monitoring system runs stably,the system works to meet the design requirements,the front-end interface is simple,the frame is clear,and the appearance is elegant,which has great practical significance for the monitoring management of the fusion experiment data.