Testing in Production at Scale
summary
Argues that running testing in production at Uber scale (600 cities, 75M riders, 15M trips/day) is viable by attaching a tenancy context to every request and persisting artifact, propagating it with Jaeger, and routing via a sidecar so instances handle mixed or isolated test traffic. Describes gateway vs language-specific client libraries, staging separate-production instances as a stepping stone, and extensions like tenancy-based rate limits, shadowing, canaries, and recording/replay for offline validation.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 3:08Agenda
- 3:49Uber Scale
- 4:16Microservice Architecture
- 5:48Capacity Planning
- 7:10Testing in Production
- 8:30Design Considerations
- 10:28Building Blocks of Tenancy
- 11:12Context Propagation
- 12:34Infrastructure Components
- 13:33Services in Production
- 15:04Tenancy
- 18:50Extensions
- 20:26Final Thoughts
