A distributed algorithm for ordered, atomic and simultaneous group communication

Abstract

The group communication algorithms are employed in fault-tolerant distributed computing systems, where the members of a group coordinate by using multicast messages. The messages need to be totally ordered at the nodes in order to maintain global consistency among the nodes. In the large scale distributed systems having replicated servers, the asynchronous as well as simultaneous group communication mechanism is required. This paper proposes a distributed algorithm intended to unidirectional virtual ring for asynchronous and simultaneous group communication in a close group maintaining totally ordered as well as atomic communication. The algorithm is easy to implement and computationally inexpensive having reduced overall message complexity. The algorithm is implemented in distributed system setup in large scale and the experimental results illustrate that algorithm is scalable, where the message complexity per node is comparatively lower. This paper describes the design, analysis and experimental validation of the algorithm. (C) 2010 Elsevier B.V. All rights reserved.