Brian Rogers, Microsoft
Unit tests, regression suites, and end-to-end scenarios are invaluable in establishing a baseline measurement of software quality. They are also wholly insufficient for determining whether a product is truly ready to ship. To aid in this decision, we must regularly collect data on how well a system stands up to longer periods of use (and abuse). Long-haul testing is a methodology uniquely suited for this task.
This paper demystifies long-haul testing and shows how this valuable but often poorly understood technique can pay dividends if appropriately defined and implemented. Learn the principles and practices for building effective long-haul tests to simulate production usage patterns, leverage controlled chaos, and achieve high coverage while maintaining reasonable cost.
With ever-increasing user expectations around high availability and reliability, it is now more important than ever to adopt and embrace long-haul testing for your project. So are you satisfied with short-term quality indicators, or are you in it for the long haul?