Markov chain usage models provide significant automation capabilities, including test case generation, test execution, and result analysis. Kirk Sayre presents a real-world case study describing how Markov chain testing techniques were applied to the testing of dynamic Web applications. Learn how Markov chain models are used to generate random and non-random test and provide you with reliability and risk estimation information. Use this technique for load and performance testing, system testing, and acceptance testing.
STARWEST 2003 - Software Testing Conference
Programming methods and tools are changing at a pace unrivaled since the Seventies. The economy has changed dramatically and continues to change now. New technology appears continuously and is becoming cheaper to deploy. Many testers who remain static in this rapidly changing climate will find themselves irrelevant and unemployed. Testers who change with the times not only find their jobs more stable, but more satisfying than ever before. The choice is yours.
Because Web services are almost completely dynamic, there is an increased chance of errors in applications using these services. In addition, applications often are more closely tied to business transactions, increasing the business risks whenever those errors do occur. By design, Web services allow access from anywhere in the world and provide its services in real-time.
Beta programs and early release programs are commonly used in software release cycles. The next level of partnership is bringing actual customers into the test lab. When customers test, even for short periods of time, their contribution can change your test practices and test environments forever. The best part is, they’ll do it for free! Learn how to recruit customers for testing and then use them to improve test practices within your test team.
Reliable and consistent performance must be an integral part of your software's release criteria and specifically tested during quality assurance. Learn the key elements of building a performance benchmark for your application. Steve Rabin describes a roles-based approach for performance testing benchmark and shares the methodology he has used numerous times. With this process, you define metrics, workload characteristics, transactional definitions, and utilization assumptions.
Organizing test artifacts and having an automated process to manage them is critical to the success of test groups in complex application development lifecycles. Managing test cases in MS-Word or a spreadsheet no longer work. From the initial business case for upper management and achieving buy-in within the test team to tool acquisition and implementation, learn how to deploy a structured testing management system in your organization.
Many organizations are finding out that their customers and users won't accept poorly designed Web sites and other critical applications. Through usability testing, you'll increase sales, increase traffic, and improve user productivity. In this case study from a major retailer, Antonet Bekker describes how to evaluate a system's usability in step-by-step approach. Often, recommended improvements to the software's usability design before (or after) the software goes live will result in dramatic increases in efficiency and effectiveness.
A large system test can consist of hundreds or even thousands of test cases, making it difficult to report results to management in a meaningful way. We typically use summary metrics, but they don't always present a clear picture. In this session, Jan Scott shows you how to develop business-driven test objectives, measure your testing progress against these objectives, and present your results to management. Improve your test process while giving management a better tool for deciding when the software is ready to go into production.
With Web services touted as revolutionary technology for eBusiness, many companies are beginning to add this dynamic element to their internal and external Web applications. If your company is deploying Web services, you need a new test plan for your application. Rather than the usual independent tester approach, testing Web services requires a cross-functional team of developers, testers, and performance engineers in an interactive, iterative development and testing cycle.
Distributed testing is often used to leverage outsourced or offshore test and development labs. Managing distributed projects does not automatically begin when the contract is signed, or during the kick-off meeting. It begins when you develop an effective framework for executing a distributed project. Jacob Hsu identifies the tools you need to successfully manage distributed teams, including defining methodologies, creating an enabling infrastructure, and improving corporate culture.