floating_Memif2IP2Memif

Floating interdomain memif to IP to memif example

This example shows that NSC can reach NSE registered in floating registry.

NSC and NSE are using the memif mechanism to connect to its local forwarder. Forwarders are using the IP payload to connect with each other.

Important points:

• nsc deploys on cluster1 and requests network service from cluster3.
• nse deploys on cluster2 and registers itself in cluster3 with IP payload.

Requires

Make sure that you have completed steps from interdomain

Run

1. Deploy network service on cluster3

Deploy NS:

kubectl --kubeconfig=$KUBECONFIG3 apply -k https://github.com/networkservicemesh/deployments-k8s/examples/interdomain/usecases/floating_Memif2IP2Memif/cluster3?ref=7a2735f6f8c8ed02d058c1a6a2f04846a3d88cad

2. Deploy endpoint on cluster2

Deploy NSE:

kubectl --kubeconfig=$KUBECONFIG2 apply -k https://github.com/networkservicemesh/deployments-k8s/examples/interdomain/usecases/floating_Memif2IP2Memif/cluster2?ref=7a2735f6f8c8ed02d058c1a6a2f04846a3d88cad

Wait for applications ready:

kubectl --kubeconfig=$KUBECONFIG2 wait --for=condition=ready --timeout=2m pod -l app=nse-memif -n ns-floating-memif2ip2memif

2. Deploy client on cluster1

Deploy NSC:

kubectl --kubeconfig=$KUBECONFIG1 apply -k https://github.com/networkservicemesh/deployments-k8s/examples/interdomain/usecases/floating_Memif2IP2Memif/cluster1?ref=7a2735f6f8c8ed02d058c1a6a2f04846a3d88cad

Wait for applications ready:

kubectl --kubeconfig=$KUBECONFIG1 wait --for=condition=ready --timeout=2m pod -l app=nsc-memif -n ns-floating-memif2ip2memif

3. Check connectivity

Ping from NSC to NSE:

result=$(kubectl --kubeconfig=$KUBECONFIG1 exec deployments/nsc-memif -n "ns-floating-memif2ip2memif" -- vppctl ping 172.16.1.2 repeat 4)
echo ${result}
! echo ${result} | grep -E -q "(100% packet loss)|(0 sent)|(no egress interface)"

Ping from NSE to NSC:

result=$(kubectl --kubeconfig=$KUBECONFIG2 exec deployments/nse-memif -n "ns-floating-memif2ip2memif" -- vppctl ping 172.16.1.3 repeat 4)
echo ${result}
! echo ${result} | grep -E -q "(100% packet loss)|(0 sent)|(no egress interface)"

Cleanup

Cleanup resources for cluster1:

kubectl --kubeconfig=$KUBECONFIG1 delete ns ns-floating-memif2ip2memif

Cleanup resources for cluster2:

kubectl --kubeconfig=$KUBECONFIG2 delete ns ns-floating-memif2ip2memif

Cleanup resources for cluster3:

kubectl --kubeconfig=$KUBECONFIG3 delete ns ns-floating-memif2ip2memif