Performance Evaluation of E-Voting Based on Hyperledger Fabric Blockchain Platform

Abstract

Permissioned blockchain platforms have become more prevalent in a wide range of applications. These, such as hyperledger fabric platforms, are sensitive to latency and throughput. In this work, the E-voting case study adopts a hyperledger fabric platform where performance evaluation has been studied in terms of scalability, latency, throughput, CPU usage, and memory allocation. Three scenarios were performed with varying transaction rates, block size, and organizations. Another two scenarios were performed, first with varying block timeout and second, measuring the impact of CPUs and memory allocation on the proposed fabric's entities (peers, orderer, couchDB, chaincode, etc.). The result shows that an increase in block size will significantly affect metrics such as latency and throughput. Good results were obtained with high transaction send rates on large block size. Similarly, low performance is obtained using a small block size with increased send rates. Also, it was noticed that increasing the number of organizations will increase latency and decrease the throughput. Therefore, in applications with a large number of concurrent transactions, to maintain high throughput, block timeouts and block size should be large. On The other hand, the number of CPUs and amount of memory allocation would impact hyperledger fabric performance.