Improving Software Quality and Reliability Through Analysing Sets of System Test Defects

摘要

Telecommunications networks support many critical services, leading users to demand very high levels of quality and reliability from these networks. The quality and reliability of these services is mainly dependent on the network's software. At the same time, competition is driving the demand for new software features in short delivery cycles. There are many challenges to delivering high quality, highly reliable software in these short cycles. Overcoming these challenges requires fast feedback to the development processes to minimize the number of escaped defects. This fast feedback can be achieved through the systematic analysis of system test defects. This method contrasts with the typical practice of analysing customer found defects. This improved method analyses each system test defect as it is fixed and stores this data. A set of defect data is then analysed to identify the most common defect type and where they are injected. This enables teams to focus improvement efforts on their largest source of defects. By automating this method, teams can continually fine tune their development processes to minimise the number of escaped defects. This results in a steady improvement over time in the quality and reliability of the software.