The Scientific Method for Resilience
summary
Uses a one-hour Failure Modes and Effects Analysis meeting to turn failure questions into explicit hypotheses (example: if the database dies, writes fail and reads use an in-memory cache). Runs chaos engineering by shutting down the DB in non-prod with realistic load, observes a retry storm that crashes the microservice and evicts the cache, documents dashboards and queries, then implements exponential backoff/circuit breakers, cache population fixes or autoscale tweaks and retests.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 0:52The Scientific Method
- 3:05Observation
- 4:24Meeting
- 5:39Discussion
- 7:09Chaos Engineering
- 8:33Analysis
- 10:15Conclusions
related talks
38mChaos Engineering
22mResilience Is Not Enough
03mDeveloping a Chaos Architecture Mindset
38mWhy software developers should build a resilience culture based on chaos engineering
41mOnce Upon a Vulnerability: Building Resilience in a Chaotic World
36mDesigning for failure - architecting fault-tolerant system
29mComplexity: The Crucial Ingredient in Your Kitchen
34mBuilding High Resiliency into Digital Solutions