A Multi-Agent Framework for Testing 3-Tier Distributed Systems Architecture

Lack of time and capital for software testing is an oft encountered problem in most organizations. The plethora of possible test cases for each software component makes software testing a very expensive and time consuming activity and can, in extreme cases, account for up to 50% of the total cost of the software development process. The advent of distributed systems has augmented the complexity of the testing process and has made it an even more daunting task than before. This paper presents an innovative multi-agent framework to ease some of the burdens associated with the testing of generic distributed systems. The architecture of the proposed MAS (Multi-Agent System) is presented along with detailed descriptions of the various models arrived at following the analysis and design phases of the proposed system. The proposed framework is considered as a step towards the automation of the testing process for distributed systems in order to enhance their reliability within an acceptable range of cost and time.