Distributed tracing is great — it helps you identify (micro)services within complex architectures having issues interfering with user experience, such as high latency or errors. But once a problematic service is identified, it can be difficult to find out which methods are to blame for the slowdown.

In this presentation, Felix Barnsteiner (one of the core developers of the APM Java Agent) will show you the different ways to get method-level insight into your application. Specifically, we’ll have a look at the newly added support for profiler-inferred spans that is based on a sampling profiler.