Controlled Chaos: Managing your Microservice Ecosystem for Business Agility
summary
Argues that microservice “service landscapes” need active, real‑time governance rather than deeper tracing or static configs, claiming runtime control and a Mission‑Control mindset reduce ripple failures. Shows concrete remediation patterns — bulkheads and backpressure plus quarantines and circuit breakers — and cites an outage where one router causes an eighty‑million dollar loss and five‑day outage. Demos include Kafka/container noise, segmented video conferencing, connected‑car fleets, and removing staging by containing deployments in production.
AI-written summary · about
chapters (15)
- 0:00Start
- 2:20Southwest meltdown
- 5:06Overview
- 9:09The agile operating model
- 11:56Service landscape
- 14:18Stability complex emergent behaviors
- 19:17Do nothing
- 22:19Monitoring
- 27:23Service mesh
- 32:46Golden signals
- 33:29Operational patterns
- 46:34Runtime control examples
- 47:38Summary
- 49:11Stages of acceptance
- 56:53Takeaways
related talks
01h 01mEmbrace chaos to achieve stability
46mEmbracing Collaborative Chaos: Chaos Days on Large Platforms
39mResilience & Chaos Engineering in the Cloud
01h 18mIn Defense of Microservices
45mDeveloping a Chaos Architecture Mindset
20mThe Business Benefits of Microservices
43mPanel: Microservices, Where Did It All Go Wrong
37mKnee Deep in Microservices