A temporal agent based approach for testing open distributed systems

• We present some technical issues for testing distributed frameworks with timing constraints.
• We model an architecture taking into account the delay of messages exchanged between the components of the testing distributed applications.
• We propose a Multi-Agent based system to capture the complex monitoring tasks of distributed tester behaviors

The development of distributed testing frameworks is more complex, where the implementation process must consider the mechanisms and functions required to support interaction as long as the communication and the coordination between distributed testing components. The typical reactions of such systems are the generation of errors‘set: time outs, locks, observability, controllability and synchronization problems. In other side, the distributed testing process must not only check if the output events have been observed, but also the dates when these events have been occurred. In this paper, we show how to cope with these problems by using a distributed testing method including timing constraints. Afterwards, a multi-agent architecture is proposed in the design process to describe the behavior of testing a distributed chat group application on high level of abstraction.