The EAST MDSplus Access Data Monitoring and Early Warning System Based on Elastic Stack
Y. Zhang, Feng Wang · IEEE Transactions on Plasma Science · 2025
The exponential growth of experimental data generated by the Experimental Advanced Superconducting Tokamak (EAST) tokamak device necessitates enhanced monitoring of MDSplus data access patterns. While existing MDSplus infrastructure records basic connection metadata, critical gaps persist in auditing granular operational details (e.g., specific data operations, user attribution, and access patterns), creating vulnerabilities for large-scale fusion data security. These gaps hinder the ability to detect unauthorized activities, ensure compliance with data governance standards, and protect sensitive experimental data from misuse or breaches. To address MDSplus’s auditing limitations in fusion data security, we develop an access data monitoring and early warning system with three core contributions: A multidimensional logging architecture designed to track user-specific operations and access patterns, thereby obtaining comprehensive audit trails; an elastic stack-based infrastructure enabling real-time analysis and storage of Terabyte-scale audit trail; and dual monitoring interfaces integrating anomaly detection algorithms for both real-time alerts and offline investigations. Our framework establishes a paradigm for high-resolution diagnostics in fusion experiment data governance, which can be readily applied to other fusion experimental devices that use MDSplus for data storage, and provides significant value for monitoring fusion experiment data and ensuring data security.