The Design And Implementation Of QA Metrics

With the rapid development of software industry, software engineering is becoming more and more mature, the scale of software development has been expanded gradually, and software quality management becomes a very important role in this process. For the software quality management, defect tracking is a very essential part. Besides defect tracking, it can control the work flow to ensure the iterative development carry out smoothly. Defect tracking system solves the management problems of software projects through software technology.For the current defect tracking systems, they can trace the defect in the development process timely, accurately, but also can make the workflow of project to make the managment easily, and continuous integration and update. However, they can not provide the trend charts which are multi-angle and long time for the defects. They can not let the managers to moniter the risk clearly in the process of software development, so we need a tool to enhance this.This thesis is based on the existing defect tracking system, developing a customized software quality monitor system. It is responsible for collection of defects in each stage of project development process, according to different statistical algorithm to provide the intuitively, visually chart.This system mainly uses the flex of Web development technology and Ruby on rails server development technology, adopt three-tier architecture, using the MVC design pattern, where Flex is the view layer, Ruby on Rails model layer and control layer.The system is composed of the Dashboard page and background data statistics section.Among them, statistical data processing is the core part of the system, which contains each statical process for each chart, provide the data set for charts.This system has completed the development and testing of the first version, and has been on line in the team. It plays an important role of monitoring defects, but also provides more bases which evaluate the software development risk for manager.