Enhancing Fault Tolerance in Private Blockchain Networks through Modified Raft Leader Election Algorithm

Enhancing Fault Tolerance in Private Blockchain Networks through Modified Raft Leader Election Algorithm

In this paper, I study the Raft leader election process to enhance fault tolerance in a network composed of minimal nodes by considering various failure situations that may occur during consensus in a private blockchain network. In the process of processing network partition situations, node failure situations, and leader node failure situations, an Activity Score variable is set, so that the platform is configured with the minimum number of nodes in a network partition situation or node failure situation, and when successful leader election is required, it can be modified. Leader election is conducted according to the Raft algorithm, and leader node election and network failures are minimized based on trust to enhance fault tolerance even in a platform environment where the minimum number of nodes is operated. Excellent performance of over 12% on average was confirmed.