On making your Rails App more transparent
summary
Adds OpenTelemetry to a Rails app with 3 gems and a tiny initializer, runs a local Jaeger UI via one Docker container, and produces a real trace of 12 spans showing controller, DB, Redis and view timings. Demonstrates head‑based vs tail‑based sampling (tail records everything then filters for errors at the collector), notes tail sampling increases runtime overhead, and says instrumentation works for monoliths and microservices alike while exposing inter-service dependencies and pinpointing latency bottlenecks.
AI-written summary · about
chapters (16)
- 0:00Intro
- 1:22Making things more transparent
- 1:44Joining a new team
- 2:13Onboarding
- 2:27Complexity
- 2:41Day job
- 2:57Rails
- 3:42Open Telemetry
- 6:16What does Open Telemetry do
- 8:37Supermini examples
- 9:52Traces and spans
- 11:46OpenTelemetry
- 12:48Traces
- 13:46Sampling
- 17:46Use cases
- 21:30Demo
related talks
01h 39mHow to instrument your Rails app with OpenTelemetry
21mBuilding a Performance Analytics Tool with ActiveSupport
26mRails Performance Monitoring 101: A Primer for Developers
36mHow to Build a Smart Profiler for Rails
35m"10 performance sins of Rails applications"
34mCode Spelunking: teach yourself how Rails works
57mFuture of Developer Acceleration with Rails
32mReading Rails: A Visual Walkthrough of the...