Multus CNI 12 Factor App — Multi-Network

Multus CNI 12-Factor

Multus CNI 12 Factor App Kubernetes Multi-Network Pod Interface macvlan ipvlan SR-IOV Storage Management Telco NFV Production

Network	CNI Plugin	Use Case	Performance
Default (Pod)	Calico/Cilium/Flannel	Service-to-Service Traffic	Standard
Storage	macvlan/SR-IOV	Ceph NFS iSCSI Replication	High (Dedicated)
Management	macvlan/bridge	Monitoring SSH Admin	Standard
Data Plane	SR-IOV/DPDK	NFV 5G Packet Processing	Very High (Near-native)

Production Monitoring

# === Production Monitoring & Troubleshooting === # Troubleshooting Commands # kubectl exec my-pod -- ip addr # ดู Interface ทั้งหมด # kubectl exec my-pod -- ip route # ดู Routing Table # kubectl exec my-pod -- ping 10.10.0.100 # ทดสอบ Connectivity # kubectl describe net-attach-def storage-net # ดู Network Config # kubectl get net-attach-def -A # ดู Network ทั้งหมด # journalctl -u kubelet | grep multus # ดู Multus Log @dataclass class MonitorItem: metric: str check_command: str target: str alert: str monitoring = [ MonitorItem("Pod Network Interfaces", "kubectl exec pod -- ip addr | grep net", "ทุก Pod มี Interface ครบตาม Annotation", "Missing Interface → P1 Check Multus CNI Log"), MonitorItem("IP Address Pool Usage", "kubectl get ippools (whereabouts)", "< 80% ของ Subnet ที่จัด", "> 90% → P2 Expand Subnet หรือ Cleanup"), MonitorItem("Network Connectivity", "kubectl exec pod -- ping ", "RTT < 1ms (Same Node) < 5ms (Cross-node)", "Timeout → P1 Check Host Interface CNI Config"), MonitorItem("Bandwidth per Network", "iperf3 between Pods on each network", "Storage > 10Gbps | Mgmt > 1Gbps", "< 50% expected → P2 Check MTU Congestion"), MonitorItem("Multus DaemonSet Health", "kubectl get ds -n kube-system multus", "Running on all nodes READY = DESIRED", "Not Ready → P1 Node ไม่มี Multi-network"), ] print("=== Production Monitoring ===") for m in monitoring: print(f" [{m.metric}]") print(f" Check: {m.check_command}") print(f" Target: {m.target}") print(f" Alert: {m.alert}")