Authors : N. Gopikarani, G. Sudha Sadasivam and G. Kavitha
Abstract: In distributed environment, the leader election process is to main system consistency and to make the system a reliable one. Leader election process is to select a single node as leader among other members. High availability is a system design approach. It ensures that a pre arranged level of operational performance will be met during a contractual measurement period. High availability is essential for mission critical applications. The term highly available refers that the systems can continue providing services even when master node (i.e., coordinator) fails. Goal of the proposed work is to use a suitable distributed election algorithm to improve high availability features. Stable leader election is implemented to increase robustness and efficiency. Election commission concept is used to minimize message passing. This algorithm is used to identify a temporary master when there is any failure in existing master node and reduces message passing. At the back-end the implementation will be invoked inside a cluster to achieve scalability in addition to high availability features.
N. Gopikarani, G. Sudha Sadasivam and G. Kavitha, 2016. A New Improved Election Algorithm for Achieving High Availability in Distributed Systems. Asian Journal of Information Technology, 15: 4482-4488.