Multi-Agent based Functional Testing in the Distributed Environment

Abstract

Verification and testing are two formal techniques of defect reduction applied on designing and development phases of SDLC to rationalize quality assurance activities. The process of testing applications in the distributed environment becomes too complex. This study discusses a distributed testing framework that consists of many parallel tester components. The idea is based on utilizing client server environment to conduct software testing efficiently and in a short span of time. It is pertinent to mention that this study is restricted to testing of functional aspects of the software while testing of performance and other quality-of-service aspects are outside the scope of the study. An important factor influencing the use of agent technology in software testing is the dynamic nature of events. Since agents are characterized by intelligence and autonomy, their ability to interact with the environment offers added functionality to make decisions based on the needs of the scenarios that are dynamic in nature. This study shows that the use of agents to build a dynamic model for software testing in the distributed environment results in a more robust and efficient design. The proposed framework is based on distribution of test cases among multiple agents deployed across a distributed system which collaborate with each other to perform testing in an efficient manner. The proposed framework also provides an in-depth visibility into the software quality by providing the defect statistics on-the-fly. The experiments have been conducted using Selenium test automation tool. The test cases along with their test scripts and the test run results are described herein.