HyperBSA: A High-Performance Consortium Blockchain Storage Architecture for Massive Data

With the considerable exploration of blockchain in various industrial fields, the storage architectures of mainstream consortium blockchains exhibit significant performance limitations, which can't meet the requirements of efficient data access with massive data storage in enterprise-level business scenarios. In this paper, we creatively divided the underlying data of the consortium blockchain into two categories: continuous data and state data and proposed a new storage architecture to store and operate these two types of data efficiently. For continuous data, we designed a specialized index-based storage engine. For state data, we proposed a multi-level cache mechanism with a secure and integrated data persistence policy. In addition, a pluggable Client/Server mode is employed to achieve flexible distributed extension. A series of experiments are conducted to show the effectiveness of our architecture. Compared with mainstream consortium blockchain storage architecture based on LevelDB, the average time-consuming decreases 81.85%/82.47% for reading/writing continuous data and 22.21%/48.99% for reading/writing state data. Compared with the storage architecture based on distributed database TiKV, the time-consuming decreases more significantly. This storage architecture has been integrated into the enterprise-level consortium blockchain platform Hyperchain, which has supported the efficient running of dozens of large-scale commercial blockchain projects with massive data.