drin-chaos-engineering-runner

Chaos Engineering Runner

Purpose

To build confidence in the system's ability to withstand turbulent conditions in production.

Experiments

1. Network Latency: Simulating slow services.
2. Pod Termination: Randomly killing containers to test auto-healing.
3. DB Failover: Simulating primary database failure.
4. Resource Hunger: Artificially spiking CPU/Memory usage.

Rules

• Never run chaos experiments without a clear "Steady State" metric.
• Always have a "Big Red Button" to stop and rollback immediately.
• Run experiments during low-traffic periods first.

---

subagents:

• sre-engineer
• devops-engineer