Evaluation of methods for automated testing in large-scale financial systems
Maryna Shtakova · KTH Publication Database DiVA (KTH Royal Institute of Technology) · 2012
Nowadays automated testing technologies are widely used by companies to make testing of software applications effective and save time and effort applied during a manual testing. Automated testing uses different approaches: from the oldest “record and playback” test method to the modern non-scripting method. All existing testing tools are based on one of these testing methods or their combinations and can be used with varying effectiveness in testing of real software products. The purpose of this thesis work was to compare testing methods in “real life” – by applying them to the large-scale financial system “Scila Surveillance”. To achieve this goal, it was decided to select several testing tools with different testing methods in a basis, evaluate them and select the best one for the automation of functional testing of the graphical user interface in the Scila company. Evaluation was done based on the list of criteria, defined at the beginning of the thesis work and with attention to requirements for a testing tool stated by Scila. Decision matrix method was used to make a final decision about the best testing tool. It was found that it is complicated to select a testing method which will fulfill all the requirements and expectations completely. The solution taken was to combine a non-scripting testing tool, which includes basic necessary functionality and allows producing test cases quickly with a scripting testing tool, which supports extraordinary test scenarios in case of need.