Common patterns of performance issues and how to identify them using visualisation.
In performance we often come up against similar categories of performance issue (such as global freezes, timeout/retry patterns). Like searching for constellations in the night sky, these patterns can be identified visually when we analyse our results and monitoring data. In this session we will look at some common and not so common patterns to look out for and how to interpret them. We will also cover the power of raw data, and the limitations and uses of aggregates (touching briefly on what this means for the future of automated performance analysis).
Stephen is a software performance specialist based in Auckland, New Zealand. For ten years he worked as an external consultant, providing performance testing and related services to a wide variety of customers. He is now working directly for an organisation with the hopes of building something lasting and adding value across the business. Stephen participated in the previous two PAC events, as well as WOPR 25 and 26. He has a keen interest in mentoring, communication, information retention, and making complicated things easier to understand. Stephen is also a professionally trained actor. The pinnacle of his acting career was starring as an extra in James Cameron’s ‘Avatar’. He switched careers to IT soon after because he wasn’t very good at acting.