🎓 Mission Debrief: Pod Logs Mystery

What Happened

The PostgreSQL container needed the POSTGRES_PASSWORD environment variable to initialize, but it wasn't provided. The container started, failed immediately, restarted, and repeated—entering CrashLoopBackOff. The only way to discover this was by checking the logs.

The Correct Mental Model

Logs are your debugging superpower. Not all failures are visible in kubectl describe. Some applications: - Start successfully (so the pod shows "Running") - Fail due to configuration errors - Exit immediately - Restart and repeat

Log locations in Kubernetes: - Container logs: Captured from stdout/stderr - Access via: kubectl logs - Stored temporarily on node - Rotated when they get too large

Commands You Mastered

# View current logs
kubectl logs <pod> -n <namespace>

# View previous container logs (after crash)
kubectl logs <pod> --previous -n <namespace>

# Follow logs in real-time
kubectl logs <pod> -f -n <namespace>

# Specific container in multi-container pod
kubectl logs <pod> -c <container> -n <namespace>

# Last N lines
kubectl logs <pod> --tail=50 -n <namespace>

# Logs since timestamp
kubectl logs <pod> --since=1h -n <namespace>

What's Next

Next: Init containers that block pod startup!