Flamegraphs 201
summary
Marcos Albe and Agustín Gallego explain advanced use of flame graphs for performance analysis, focusing on collecting and visualizing stack traces to find bottlenecks in applications and databases. They cover tools and workflows for generating flame graphs (e.g., perf, eBPF or similar profilers, and Brendan Gregg’s flame graph scripts) and practical recommendations for diagnosing real-world performance issues.
AI-written summary · about
chapters (14)
- 0:00Flamegraphs intro
- 1:26Technical definition
- 6:12Visual structure
- 10:15Capturing data
- 12:04Basic generation demo
- 17:47Advanced Flamegraphs
- 19:58Subsystem filtering
- 24:19Interpreting results
- 25:22Data cleanup
- 27:07Color map visualization
- 30:49Comparing captures demo
- 34:29Performance impact
- 38:02Additional resources
- 40:47Summary & Q&A
related talks
13mSto: A Better Way to Store and Query Profiler Data
36mFlare: Optimizing Apache Spark for Scale-Up Architectures and Medium-Size Data
33mCassandra Performance Tuning Like You've Been Doing It for Ten Years
33mJourney of developing a performance optimization feature in PGSQL
48mFlare: Scale Up Spark with Native Compilation and Set Your Data on Fire
18mGraph analytics: From zero to hero
29mAll about graphs in Database
19mA New Approach to Logging as a Stack: Fluent Bit + PostgreSQL (FPS)