Design and implementation of a distributed applications testbed

In order for designers and implementers to evaluate interoperability, trade-offs, performance of algorithms, and protocols within distributed applications, experiments are often necessary. Results obtained from experiments can be ultimately used to improve on the performance, efficiency and reliability of the applications during their operations. We propose a general architecture for a distributed applications testbed. This general architecture has emerged from a specific distributed applications testbed for the experimental study of problems related to X.500 distributed directories. Our design and experience suggest that a general architecture and tools can facilitate the development of similar testbeds for other distributed applications.<
>