Heat Prediction of High Energy Physical Data Based on LSTM Recurrent Neural Network

¹ IHEP computing center, 19B Yuquan Road, Beijing 100049, China
² University of Chinese Academy of Sciences, 19A Yuquan Road, Beijing 100049, China
³ Tianfu Cosmic Ray Research Center, IHEP, 9 Renmin South Road, Chengdu 620500, Sichuan, China

^* chengzj@ihep.ac.cn, wanglu@ihep.ac.cn, chyd@ihep.ac.cn, gang.chen@ihep.ac.cn, chengzj@ihep.ac.cn

Published online: 16 November 2020

Abstract

High-energy physics computing is a typical data-intensive calculation. Each year, petabytes of data needs to be analyzed, and data access performance is increasingly demanding. The tiered storage system scheme for building a unified namespace has been widely adopted. Generally, data is stored on storage devices with different performances and different prices according to different access frequency. When the heat of the data changes, the data is then migrated to the appropriate storage tier. At present, heuristic algorithms based on artificial experience are widely used in data heat prediction. Due to the differences in computing models of different users, the accuracy of prediction is low. A method for predicting future access popularity based on file access characteristics with the help of LSTM deep learning algorithm is proposed as the basis for data migration in hierarchical storage. This paper uses the real data of high-energy physics experiment LHAASO as an example for comparative testing. The results show that under the same test conditions, the model has higher prediction accuracy and stronger applicability than existing prediction models.