Emergent Load Testing Rules for Organized Chaos
summary
Claims that emergent load testing finds what traditional scripts miss by giving tests agency and bidirectional feedback, demonstrated with working k6 examples that combine k6 core browser kubernetes disruptor modules to spin up clusters, inject delays, run browser flows, and trace requests into DBs. k6 is open source, scripted in JavaScript and runs in Go, emitting Prometheus-format metrics and logs to Loki and traces to Tempo. Tests are unpredictable, slow and nonrepeatable but reveal operational thresholds in complex systems.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:29What is emergence
- 6:29Emergence in software
- 14:40Challenges in emergent systems
- 22:29k6 as a multi-tool
- 24:43k6 core affordances
- 26:49k6 Kubernetes module
- 27:53Fault injection with k6
- 30:01Observability in load tests
- 32:14Tracing distributed requests
- 34:19k6 Redis storage
- 35:25Beyond API load testing
- 36:50Emergent load testing logic
- 38:37Emergent testing vs production
- 40:41When emergent testing works
- 41:03Summary
related talks
34mK8s load testing at scale with k6-operator
45mDistributed load testing with k6
46mDistributed load testing with k6
14mBreaking the build with K6 load tests
34mPerformance testing: from zero to hero with k6 & Azure
06mIt's Just a Jump to the Left...
38mMasterclass "Route to the Heights with k6"
35mChaos Engineering WG Deep Dive