Increasing Multi-tier Computer System Service Quality through Software Testing
Timo Jakku · Theseus (Ammattikorkeakoulujen) · 2016
The goal of this thesis was to study one multi-tier system from the perspective of faults found in production usage. These faults were analyzed in order to find out if any particular level of testing and any particular tier of the system had any statistical connection and thus would reveal the need for an improvement on a certain test activity or level on a certain tier of the system. The reported faults were extracted in the form of trouble reports from a tracking system. Only software fault related issues were kept for analysis, disregarding hardware, infrastructure and other miscellaneous faults. The faults were categorized based on the system tier, priority, testing level and ISO 25010 quality model. These categorizations were then mapped against each other in order to see if any logical connection could be found. The analysis shows that there is room for improvement in several areas of testing. The study makes recommendations as to increasing the amount of manual testing and broadening the scope and variation in automated testing. While it was observed that most faults occur in the middle tier of the system, it was concluded that no significant advantage can be gained by targeting that tier in particular. The result of this study was a useful view into the situation of one particular system and its current problems. However, the conclusions drawn from these results are only applicable to a very limited extent to other multi-tier systems, as there are too many variables in each system for such comparison. However, the methods and processes utilized in the study can be used to extract another set of comparable results from another system.