gabrielmoreira Kubernetes Patterns
Deploy and manage production Kubernetes clusters.
Core Objects
Workloads
| Object | Use Case | Scaling |
|---|---|---|
| Deployment | Stateless apps | Replicas, HPA |
| StatefulSet | Stateful apps (DBs) | Stable network IDs |
| DaemonSet | Per-node agents (logging, monitoring) | Node count |
| Job/CronJob | Batch tasks, scheduled jobs | Completion |
Networking
- Service: Stable endpoint for pods (ClusterIP, NodePort, LoadBalancer)
- Ingress: HTTP routing, TLS termination, path-based routing
- Network Policies: Pod-level firewall rules
Production Best Practices
Resource Management
resources:
requests:
cpu: 100m
memory: 128Mi
limits:
cpu: 500m
memory: 256Mi
- Always set requests AND limits
- Use LimitRange for namespace defaults
- Use ResourceQuota for namespace caps
Pod Anti-Affinity
Spread pods across nodes for HA.
Readiness & Liveness Probes
- Readiness: traffic starts flowing
- Liveness: pod gets restarted
- Startup: for slow-starting containers
RBAC
- Least-privilege service accounts per app
- Namespace-scoped roles (not cluster-wide)
- Regularly audit permissions
- Use groups, not individual users
Autoscaling
- HPA: scale by CPU/memory or custom metrics
- VPA: adjust resource requests automatically
- Cluster Autoscaler: add/remove nodes
- KEDA: event-driven scaling (SQS queue depth, etc.)