When working out why an application has gone wrong we often find ourselves working with very limited information – we find ourselves looking at a stack trace caused by an uncaught exception, or looking at a ConcurrentModificationException where the interfering thread is unknown, or looking at a heap that has grown without bound.

We ask ourselves “What just happened?” and “How did I get here?”. Existing and new tooling including Java Flight Recorder (JFR) can provide some more info over a bare stack trace. But the ability to record and reversibly replay an application allows us to actually answer these questions.

In this talk we’ll demonstrate how to use LiveRecorder for Java to obtain a recording of a Java application, and how that recording can be used to diagnose a number of these difficult yet everyday problems.