Multi-faceted quality and defect measurement for web software and source contents
Journal of Systems and Software
| Hi-index | 0.00 |
With the prevalence of the World Wide Web and people's reliance on it in today's society, ensuring satisfactory performance and reliability of Web servers and websites is becoming increasingly important. Various techniques exist today to test and measure the reliability of software systems. However, the Web environment presents many new challenges, and requires adapted or newly developed techniques. In this dissertation, based on the characterization of Web applications and the overall Web environment, we classify Web problems, focus on the subset of source content problems, and further develop an integrated framework for Web testing and reliability analysis. Using information about Web accesses and related failures extracted from existing Web server logs, we validate the memoryless property for Web testing based on our Unified Markov Models (UMMs), characterize website workload at different levels and from different perspectives, and evaluate Web software reliability. Finally, we identify high-risk components of websites, and explore other alternatives to measure Web quality by using other data sources and from different perspectives. We applied our approach to different kinds of websites: one educational website, one open source software development website, and one commercial website. The results demonstrated the combined viability and effectiveness of our approach.