This example shows that NS Clients and NSE on the one node can find each other.
NS Clients and NSE are using the kernel
mechanism to connect to its local ovs forwarder.
The NS Client connections are multiplexed over single veth pair interface on the NSE side.
Make sure that you have completed steps from ovs setup. There is more consumption of heap memory by NSE pod due to vpp process when host is configured with hugepage, so in this case NSE pod should be created with memory limit > 2.2 GB.
Deploy NSC and NSE:
kubectl apply -k ../../../examples/use-cases/Kernel2KernelVLAN
Wait for applications ready:
kubectl wait --for=condition=ready --timeout=1m pod -l app=nsc-kernel -n ns-kernel2kernel-vlan
kubectl wait --for=condition=ready --timeout=1m pod -l app=nse-kernel -n ns-kernel2kernel-vlan
Choose one ns client pod and nse pod by labels:
NSC=$((kubectl get pods -l app=nsc-kernel -n ns-kernel2kernel-vlan --template '{{range .items}}{{.metadata.name}}{{" "}}{{end}}') | cut -d' ' -f1)
TARGET_IP=$(kubectl exec -ti ${NSC} -n ns-kernel2kernel-vlan -- ip route show | grep 172.16 | cut -d' ' -f1)
NSE=$(kubectl get pods -l app=nse-kernel -n ns-kernel2kernel-vlan --template '{{range .items}}{{.metadata.name}}{{"\n"}}{{end}}')
Ping from NSC to NSE:
kubectl exec ${NSC} -n ns-kernel2kernel-vlan -- ping -c 4 ${TARGET_IP}
Delete ns:
kubectl delete ns ns-kernel2kernel-vlan