Brendan Gregg developed flame graphs, which provide a powerful way to visualize hierarchical data. Later on, Tanel Poder introduced the same concept for Oracle queries. This time, I’m bringing this idea for the first time as an open-source project to MySQL.

By using the JSON output of the MySQL optimizer, we can expose the query’s internal tree structure: nested-loop joins, temporary table creation, and sorting costs. Moving from traditional tabular EXPLAIN to this programmatically consumable format enables much deeper analysis of query behavior. We will explore the formats that myflames can generate for query profiling, such as flame graphs, treemaps, bar graphs, and the Visual Explain-style diagram familiar to MySQL Workbench users, and when each is most useful.

The session focuses on the logic needed to consume MySQL execution metadata and shows how these visualizations help diagnose complex query behavior that is often hidden in standard EXPLAIN output. Attendees will see how raw execution-plan data is transformed into actionable visual insights, making it easier to interpret and optimize mission-critical queries.