The Future of Ruby Performance Tooling
summary
Ruby 2.1 exposes hooks that make concrete work possible: rblineprof gives line-by-line relative timings but inflates runtime so it stays off production; ppprofiler automates 100-run bench runs to compare cache-on vs cache-off; StackProf is a low-overhead sampling profiler that records 30s production dumps and produces flame graphs revealing ActiveRecord hotspots; object_space.dump_all exports the entire heap to JSON (87 MB example, ~13 minutes to import) but pauses the app so it’s heavy yet precise for memory forensics.
AI-written summary · about
chapters (14)
- 0:00Intro
- 0:26Welcome
- 1:01Paperless Post
- 1:29Performance Matters
- 2:45Ruby Performance
- 3:06Application Performance
- 5:48The Past
- 11:53Active Support Notification
- 14:03Line Profiler
- 16:16Line Profit
- 19:42Stack prof
- 23:27Stack prof overview
- 27:28RBTrace overview
- 29:36RBTrace impact
related talks
43mReal World Ruby Performance at Scale
26mHow a Ruby profiler works: Stackprof under a microscope
42mDebugging Ruby
33mFixing Performance & Memory problems
34mMetric Driven Development with Ruby on Rails
29mHow to Performance
24mRuby Performance Secrets and How to Uncover Them
34mHow to optimize Ruby internal.
