Have you ever wondered how profiling, like JFR, works in OpenJDK and how we can improve it? In this talk, I'll take you on my three-year journey to improve profiling, especially method sampling, with OpenJDK: from the initial ideas and problems of existing approaches to my different draft implementations and JEP versions, with all the setbacks and friends I made along the way. It's a story of blood, sweat, and C++.