How much pressure can your app handle - Performance testing with Ruby
summary
Airtasker builds a Ruby-based performance framework that generates JMeter JMX files via ruby-jmeter, spins production-like environments on demand with Terraform and Kubernetes, replays anonymized production data and simulates spikes (they modelled a 300% traffic campaign and recall a 1,700-task surge), produces live dashboards and HTML reports, and surfaces real bottlenecks — a single slow endpoint caused by an N+1 query — so the team fixes queries, adjusts instance sizing and autoscaling, and reruns tests to verify improvement.
AI-written summary · about
chapters (18)
- 0:00Start
- 2:26Why performance testing?
- 3:22Reactive Problem Solving
- 4:25Performance testing process
- 6:08Types of performance testing
- 7:52The Wrong Questions
- 8:43The Right Questions
- 11:06e Production equivalent environment
- 13:216 Tool selection criteria
- 14:33Tools Comparison
- 15:47Winner winner chicken dinner
- 16:39Best Practices
- 17:49Traffic Profile by Endpoint/HTTP method
- 18:11Test environment
- 19:59Test cases
- 28:33Example report
- 29:14Next steps
- 32:07Nancy's stress testing result
related talks
01h 03mRuby at Scale
16mPerformance testing – under pressure!
33mEnsuring High Performance for Your Ruby App
34mMetric Driven Development with Ruby on Rails
29mHow to Performance
43mReal World Ruby Performance at Scale
26mHow to Performance
41mMartijn Goossens: Become a Performance Test Master - the Non-Technical Side of Things