Chaos Engineering And Failure Injection

Use When

• Production incidents recur under dependency failure
• SLOs require proven recovery behavior

Workflow

1. Define failure hypotheses and blast radius
2. Run game days in staging with guardrails
3. Inject latency, error, and resource pressure
4. Measure recovery time and user impact
5. Track remediations with owners

Required Checks