We define a "flaky" test as a test that both passes and fails with the same code. Test flakiness has many possible causes – it may occur no matter how smart our test framework is. Even at Google, every 7th test written by their stellar team occasionally fails without any code changes.
Flaky tests are not just annoying. They can be quite costly since test automation engineers often have to retrigger entire builds on CI – as a result, it significantly increases time to market. Not to mention the undermined confidence and low trust within teams.
How can we deal with and reduce the number of flaky tests? Join QATestLab and Miro as we discuss: