Testimise automatiseerimise juhtumiuuring
Files
Date
2017
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Viimase kahe aasta jooksul on LHV panga ühe arendustiimi testimisprotsess läbinud mitu parendustsüklit. Samas pole seda arengut metoodiliselt valideeritud. Selle töö eesmärk on analüüsida kolme võtmetähtsusega perioodi ning mõõta nende testimisprotsessi maksumust ja efektiivsust. Seejärel viiakse läbi mitmetasandiline analüüs, et tuvastada problemaatilised ja kasulikud mustrid ning nendega seotud tegurid. Analüüsi tulemusel seatakse eesmärgiks muuta testimisprotsess automatiseeritumaks. Sellest tulenevalt tegeleb ülejäänud lõputöö erinevate meetodite kombineerimisega, et muuta selline lähenemine läbi testide automatiseerimise protsessi teatud osade endi automatiseerimise saavutatavaks. Nendest tehnikatest moodustatakse pakutav lahendus, mis seejärel implementeeritakse ja selle lõputöö kontekstis valideeritakse.
Over the last two years, the testing process of one of the software development teams at LHV bank went through several development stages. However, there hasn't been any methodical approach towards validating that evolution. The aim of this thesis is to conduct an investigation of three key periods, and measure the cost and effectiveness of the testing process during each period. A multilevel analysis is then performed in order to identify problematic, as well as improvement patterns, and the factors associated with them. The analysis is concluded with setting the goal of shifting the testing process to a more automated model. Subsequently, the remainder of the thesis undertakes the task of combining a multiplicity of techniques that try to make such model achievable, by automating certain aspects of the test automation process itself. These techniques are articulated as a proposed solution, which is then implemented and validated in the context of this thesis.
Over the last two years, the testing process of one of the software development teams at LHV bank went through several development stages. However, there hasn't been any methodical approach towards validating that evolution. The aim of this thesis is to conduct an investigation of three key periods, and measure the cost and effectiveness of the testing process during each period. A multilevel analysis is then performed in order to identify problematic, as well as improvement patterns, and the factors associated with them. The analysis is concluded with setting the goal of shifting the testing process to a more automated model. Subsequently, the remainder of the thesis undertakes the task of combining a multiplicity of techniques that try to make such model achievable, by automating certain aspects of the test automation process itself. These techniques are articulated as a proposed solution, which is then implemented and validated in the context of this thesis.