chaos-engineering.md

Chaos Engineering

Automated chaos experiments for GistPin to validate resilience and recovery.

Experiments

Experiment	Script	Description
pod-failure	chaos-tests/pod-failure.sh	Delete a random pod; validate K8s self-healing
network-latency	chaos-tests/network-latency.sh	Inject network latency via tc netem
resource-exhaustion	chaos-tests/resource-exhaustion.sh	CPU/memory stress on a running pod
failover	chaos-tests/failover.sh	Scale deployment to 0 and restore

Usage

# Run a specific experiment
NAMESPACE=gistpin DURATION=60 bash infrastructure/scripts/run-chaos-experiment.sh pod-failure

# Available experiments
bash infrastructure/scripts/run-chaos-experiment.sh <experiment>
# pod-failure | network-latency | resource-exhaustion | failover

Environment Variables

Variable	Default	Description
NAMESPACE	gistpin	Kubernetes namespace
DURATION	60	Experiment duration in seconds
LATENCY	200ms	Latency for network-latency experiment
DEPLOYMENT	backend	Deployment name for failover experiment

Recovery Validation

Each experiment logs pass/fail based on pod count and rollout status after the chaos window. Review stdout for PASS / WARN indicators.