A Storage Architecture of Blockchain for Time-Series Data

摘要

The blockchain technology is becoming more and more important in different areas. However, it is not suitable for storing time-series data because of its rapid growth. This paper starts from the traditional storage model of the blockchain and proposes a novel storage architecture intended for time-series data. Some implementation details are given as well. The system mainly includes three modules: block storage, index storage and time-series data storage. The time-series database is used as the basic storage tool so it can fully utilize the corresponding feature such as elasticity, automation, efficiency, etc. By proper parameter configuration, the system can retain data within a certain time window, automatically clear expired blocks, free up storage space for dynamic growth, and provide data aggregation statistics that can be directly applied to data analysis systems. Therefore, it's more suitable for time-series data in some areas.