Production-grade infrastructure for Google Kubernetes Engine — private clusters, Workload Identity, zero-trust networking, and fully automated CI/CD.
Regional cluster, no public node IPs, Workload Identity Federation, Shielded Nodes, Binary Authorization, master authorized networks.
VPC-native, Cloud NAT for private egress, Calico network policies, default-deny per namespace via Helm.
Namespaces, ResourceQuotas, NetworkPolicies, LimitRanges across production, staging, and monitoring.
Zero-config GMP with 8 production alert rules — nodes, pods, deployments, HPA saturation.
Every push triggers tfsec against 200+ GCP rules. ClusterRole and RoleBindings properly wired.
System (tainted) and app pools. SURGE upgrades, zero max-unavailable, BALANCED zone distribution.
| Feature | This Repo | Typical Portfolio |
|---|---|---|
| Private nodes + Cloud NAT | ✅ | ❌ usually public |
| Workload Identity — no key files | ✅ | ❌ SA keys |
| Binary Authorization | ✅ | ❌ rarely implemented |
| Separate system + app node pools | ✅ | ❌ single pool |
| tfsec in CI pipeline | ✅ | ❌ |
| Google Managed Prometheus | ✅ | ❌ self-managed |
| RBAC with RoleBindings | ✅ | ❌ roles only |
| Prometheus ClusterRole | ✅ | ❌ scraping broken |
| Dependabot auto-updates | ✅ | ❌ |
| make validate passing clean | ✅ | ❌ |
| Live demo URL | ✅ | ❌ |