We’ve all been there: tests fail for no obvious reason. The team says that it is random or that it cannot be reproduced and the related bugs get closed with no fix. Over time, the failures continue, sometimes mutating and sometimes multiplying. While not prioritized by developers, these intermittent failures are often felt acutely while testing and will ultimately be painful for customers. These bugs are fixable if you take the correct approach.
These are the techniques I will demonstrate to overcome this problem:
– How to change the language used to report and discuss bugs.
– How to refocus the testing effort to tease out details of the failures
– How to apply visualization techniques to clearly demonstrate the impact of the issue and possible root causes.