A Leader Management Scheme for Fault-Tolerant Multiple Subnets

There have been a number of Ethernet-based fault-tolerant schemes which provide fast fault detection and recovery in a subnet environment. However, they are not scalable since Ethernet frame cannot be transmitted to the outside of a subnet. Our SAFE (scalable autonomous fault-tolerant Ethernet) scheme divides whole network into several subnets and manages leader nodes in each subnet. Leader nodes communicate each other for inter-subnet fault recovery. In this paper, we study a fault-tolerant leader management scheme for multiple subnet network. When one of leader nodes fails, the other node will be charged with task of previous leader quickly. Proposed scheme is performing in an autonomous way and network can operate continuously without interruption.