Research on Intelligent Operation and Maintenance Model of Large Database for Electric Power Industry
Chaoyi Yang, Ye Liu, Xiao Jianyi, Xiaobo Huang, Jiahua Chen · Journal of Physics Conference Series · 2021
Abstract Existing research on automated database operation and maintenance for the electrical industry mainly focuses on distributed and cloud platforms, and there is a lack of traditional large-scale database intelligent operation and maintenance research. This paper designs an overall operation and maintenance model framework of “intelligent perception-intelligent decision-intelligent execution”, and proposes feasible implementation plans, including: (1) Use the prophet time series forecasting model to perceive and predict important database performance indicators, and dynamically adjust the threshold of each performance indicator according to the predicted value; (2) Perform correlation analysis on abnormal indicators through the association rule model to construct “Indicators”->Operation” optimized combination operation strategy library for intelligent decision making; (3) According to the intelligent decision library, automatically restrict the associated operations under abnormal conditions to ensure the normal operation of the service and realize intelligent execution.