Hybrid Fault Categorization Approach To Reduce Thetesting Cost Of Combinatorial Test Cases
i-manager’s Journal on Software Engineering · 2015
In our modern software systems, customizable systems can have more number of configurations with the changes in run time or compile time preferences. These customizable systems includes application servers like Tomcat, web servers like Apache, databases like MySQL and also some of the office applications like MS Word. In order to validate the correctness of the system, it is impossible to test all the configurations. Systematic testing is not so easy as the systems have many configurations and in turn have many optional features and whole system testing takes more time to test. This problem can be resolved by Combinatorial Interaction Testing (CIT) and thus the effectiveness can be improved. Combinatorial interaction testing is a black box system testing technique that samples inputs, configurations and parameters and combines them in a systematic fashion. To ensure dependability across the configuration, researchers and practitioners are using CIT so as to get a quantifiable set of configurations that are to be tested. When used properly, it can be a powerful tool in the hands of a software tester. CIT was first introduced in the mathematical study of interactions. One of the major aspects of CIT is finding what to test. For instance, a web browser has many configurations that may change the way one views or accesses data over the web. It is very important that all the configurations work with each other and do not negatively affect the ability to use the internet. Random testing should not pair with some of the options together and if so, the problem would never be detected. CIT is used to design tests that should focus on catching these kind of errors. Initially the authors have to find out all the different configuration combinations. Combinatorial Interaction Testing is a powerful tool for testing with practical applications in Software Engineering. By finding out the total number of possible configurations and then designing an array that tests pairwise configurations so that one can reduce the size of the test suite. In this section, the authors provide the background information on covering arrays, combinatorial methods in testing. Creation of a covering array will guarantee that all pair