As automated testing with OpenQA has expanded in Fedora, we have been met with a new issue - making sure that the results are analyzed without too much responsibility on a small group of people. OpenQA result analysis has an additional challenge when compared to most other automated test results - the output is heavily skewed towards visual data which cannot generally be analyzed with the same techniques as text results. This talk will go over the more unique challenges in processing OpenQA results, the experiments we have run so far, the results from experiments and the direction we plan to take going forward.
Video Stream
