LoCom: an Efficient Load-Balanced Concurrency Optimization Method for Bulk-Commodity Regulatory Systems

Hengqi Guo,Lianggui Chen,Wenbin Huang,Junxiong Lin,Zhihui Lu,Shijing Hu,Gao Zou,Dayao Sha

DOI: https://doi.org/10.1007/978-981-99-4402-6_35

2023-01-01

Abstract:To address the lack of effective regulatory measures for the bulk-commodities electronic trading market, a trusted transaction depository system was built using Hyperledger Fabric as a framework to meet the needs of the bulk-commodities regulatory scenario, but the performance of the current Fabric can hardly meet such a large throughput. In this paper, an efficient load-balanced concurrency optimization method (LoCom) for bulk-commodity regulatory systems is proposed. Parallelisation of the endorsement phase is accomplished by transforming the underlying logic of the endorsement phase in this method. For the modified multi-container endorsement scenario, external chain code technology is introduced to manage the endorsement containers, and the transactions are distributed to different containers through dynamic load balancing algorithms. Meanwhile, a multi-threaded approach was adopted to verify the read and write sets of transactions, while parallelising the writing to the historical database and updating the ledger. Through stress testing, the system throughput is increased by nearly two times, close to the limit value in the ideal state, and the number of transactions distributed by the algorithm is basically even.