Atomic Broadcast in Heterogeneous Distributed Systems

Communication services have long been recognized as possessing a dominant effect on both performance and robustness of distributed systems. Distributed applications rely on a multitude of protocols for the support of these services. Of crucial importance are multicast protocols. Reliable multicast protocols enhance the efficiency and robustness of distributed systems. Numerous reliable multicast protocols have been proposed, each differing in the set of assumptions adopted, especially for the communication network. These assumptions make each protocol suitable for a specific environment. The presence of different distributed applications that run on different LANs and single distributed applications that span different LANs mandate interaction between protocols on these LANs. This interaction is driven by the necessity of cooperation between individual applications. The state of the art in reliable multicast protocols renders itself inadequate for multicasting in interconnected LANs. The progress in development methodology for efficient and robust LAN software has not been matched by similar advances for WANs. A high-latency, a lower bandwidth, a higher probability of partitions, and a frequent loss of messages are the main restrictive barriers. In our work, we propose a global standard protocol that orchestrates cooperation between the different reliable broadcast protocols that run on different LANs. Our objective is to support a reliable ordered delivery service for inter-LAN messages and achieve the utmost utilization of the underlying local communication services. Our protocol suite accommodates the existence of LANs managed by autonomous authorities. To uphold this autonomy (as a defacto condition), LANs under different authorities must be able to adopt different ordering criteria for group multicasting. The developed suite assumes an environment in which multicasting groups can have members that belong to different LANs; each group can adopt either total or causal order for message delivery to its members. We also recognize the need for interaction between different reliable multicasting protocols. This interaction is a necessity in an autonomous environment in which each local authority selects a protocol that is suitable to its individual needs. Our protocols are capable of interacting with any reliable protocol that achieves a causal order as well as with all timestamp-based total-order protocols. Our protocols can also be used as a medium for interaction between existing reliable multicasting protocols. This feature opens new avenues in interactability between reliable multicasting protocols. Finally, our protocol suite enjoys a communication structure that can be aligned with the actual routing topology, which largely minimizes the necessary protocol messages