The Power of Distributed Profiling
summary
A demo shows distributed profiling that actually stitches traces across languages (PHP Python Go) by propagating an x-blackfire-query header and parent/child IDs so cascaded calls become one view — one example surfaces 22 requests in the cascade. The system requires a PHP extension probe (pip package for Python, import for Go), preserves headers, supports different agents and data centers, offers paid recommendations, and the cascade linking and programmatic subqueries work as demonstrated.
AI-written summary · about
chapters (13)
- 0:00Intro
- 0:50Demo
- 3:50What is Distributed Profiling
- 6:00Distributed Profiling Example
- 7:30Distributed Profiling Demo
- 9:10Linked Profiles
- 10:30Second Level
- 15:20Recommendations
- 16:30Functions
- 17:40Complex Architectures
- 18:40Manual Profiling
- 19:30programmatically
- 21:30Disable Distributed Profiling
related talks
36mContinuous Profiling in the Cloud-Native era
24mCommunity Driven Breakout Session: Continuous Profiling
36mProfiling in the cloud-native era - Matthias Loibl
26mProfiling Mastery: Decode Your Data and Elevate Observability
28mConprof - Profiling in the cloud-native era
56mContinuous Profiling the missing piece in your observability puzzle
33mInto the flamegraph: From the primitives through advanced concepts
36mContinuous Profiling with Parca- Beneath the Kernel to Beyond...