How do you approach debugging a test script that intermittently fails in the test environment but passes locally?
