Connecting RPC-Based Distributed Systems using Wide-Area Networks

Robbert van Renesse, Andrew S. Tanenbaum, Hans van Staveren, Joseph J. Hall · 1987

Remote Procedure Call (RPC) is a widely used communication mechanism in local network based distributed operating systems. It is simple, fast, and straightforward to implement. However, when two or more distant distributed systems are connected, problems arise concerning the protocols, locating services, and other issues. To solve these problems, gateways are introduced. In this paper we discuss various ways in which these gateways can be organized and show how their application in the Amoeba Distributed Operating System has solved the problems cited above. 1. INTRODUCTION As networks of high-performance personal workstations become more widespread, interest in distributed operating systems to make the whole system look like a single timesharing system is increasing. When the same distributed operating system is running on two widely separated local-area networks, it is natural to think about merging them into a single transparent distributed system. However, because local-area and wi...

Read the paper · More papers on PaperTik