一种基于神经网络的时序库多线程快速导入导出方法
By optimizing the import and export process of InfluxDB using a neural network-based multithreaded method and a nonlinear fitting model, the problems of resource consumption and performance bottlenecks in time-series databases are solved, achieving efficient data processing and a flexible backup mechanism.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUODIAN NANJING AUTOMATION
- Filing Date
- 2024-08-27
- Publication Date
- 2026-07-17
AI Technical Summary
The existing InfluxDB time-series database consumes a lot of system resources during backup operations, resulting in performance degradation. Furthermore, the backup granularity is too large, which cannot meet the filtering backup requirements of specific resource tables. At the same time, the processing logic may become a performance bottleneck under high concurrency or large data volume conditions, lacking flexibility and scalability.
A multi-threaded fast import/export method based on neural networks is adopted. Time-series data is processed by slicing, and multi-threaded parallel processing of slices is used. A non-linear fitting model is built to optimize import/export parameters. Encryption technology and monitoring thread tracking mechanism are combined to ensure data integrity and efficiency.
It improves the efficiency of importing and exporting time series libraries, adapts to different operating environments and task requirements, reduces resource contention, and enhances system performance and flexibility.
Smart Images
Figure CN118964467B_ABST