Demystifying the Performance of Consortium Blockchain Under the Industrial Internet Application

The Industrial Internet significantly mines the value of industrial data and improves the efficiency of industrial production by collecting, sharing, and intelligently processing data generated by multiple entities involved in the industrial lifecycle. However, data sharing and interaction between multiple data holders not only require the security of the data itself but also require the construction of a trusted and secure infrastructure. Consortium blockchain has the characteristics of data nonrepudiation, tamper resistance, and traceability, and is expected to become a promising distributed ledger technology for trusted data sharing in Industrial Internet environments. This paper attempts to reveal the performance of a representative consortium blockchain system, Hyperledger Fabric, in typical Industrial Internet environments through experimental analysis methods. We consider factors such as network topology, transaction rate, and scale of participating nodes, and comprehensively analyze the key system metrics including system throughput, transaction latency, and scalability through actual deployment and measurement. The experimental results indicate that the generation rate of blocks, the scale of devices, and endorsement strategy in the industrial internet have a major impact on the performance of Hyperledger Fabric, and there will be many invalid transactions under the condition of high complexity of Chaincode. In addition, there is also potential for optimization in the underlying architecture.