Providing fault-tolerant services in an open client/service paradigm

The design and understanding of fault-tolerant open distributed systems is a very difficult task. We have to deal with not only all the complex problems of open distributed systems when all the components are well, but also the more complex problems when some of the components fail. This paper proposes a fault-tolerant computing model on open distributed systems. The performance of the model is simulated and analysed. A prototype implementation of the model using remote procedure calls (RPCs) is also described.