A Termination Detection Technique Using Gossip in Cloud Computing Environments

Abstract

Termination detection is a fundamental problem in distributed systems. In previous research, some structures are used (e.g., spanning tree or computational tree) to detect termination. In this work, we present an unstructured termination detection algorithm, which uses a gossip based algorithm to cope with scalability and fault-tolerance issues. This approach allows the algorithm not to maintain structures during runtime due to node joining and leaving. These dynamic behaviors are prevalent in cloud computing environments and little attention has been paid by existing approaches. To measure the complexity of our proposed algorithm, a new metric, self-centered message complexity is used. Our evaluation over scalable settings shows that the unstructured approach can have a significant merit on performance over existing algorithms.