Use NGINX Ingress

Procedure

1. Deploy the NGINX Ingress controller in the cloud and edge node pools. See Deploy the NGINX Ingress controller. After deployment, check the controller status and Service exposure.

   1. Check the NGINX Ingress controller pod status in the cloud node pool:

      kubectl get po -n ingress-cloud -o wide

      Expected output:

      NAME                                                     READY   STATUS    RESTARTS   AGE     IP           NODE           NOMINATED NODE   READINESS GATES
      ack-ingress-nginx-v1-cloud-controller-7d7b555f8b-5kgqb   1/1     Running   0          20s     10.10.0.1    cloud-node-1   <none>           <none>
      ack-ingress-nginx-v1-cloud-controller-7d7b555f8b-725cl   1/1     Running   0          20s     10.10.0.2    cloud-node-2   <none>           <none>

   2. Check the NGINX Ingress controller Service status in the cloud node pool:

      kubectl get svc -n ingress-cloud

      Expected output:

      NAME                                    TYPE           CLUSTER-IP      EXTERNAL-IP       PORT(S)                      AGE
      ack-ingress-nginx-v1-cloud-controller   LoadBalancer   172.12.0.1     xxx.xxx.xxx.xxx   80:30080/TCP,443:30443/TCP    30S

   3. Check the NGINX Ingress controller pod status in the edge node pool:

      kubectl get po -n ingress-edge -o wide

      Expected output:

      NAME                                                    READY   STATUS    RESTARTS   AGE   IP           NODE          NOMINATED NODE   READINESS GATES
      ack-ingress-nginx-v1-edge-controller-7d57b84c88-4wsm4   1/1     Running   0          17s   10.10.0.1    edge-node-1   <none>           <none>
      ack-ingress-nginx-v1-edge-controller-7d57b84c88-fqvj8   1/1     Running   0          17s   10.10.1.1    edge-node-2   <none>           <none>

   4. Check the NGINX Ingress controller Service status in the edge node pool:

      kubectl get svc -n ingress-edge

      Expected output:

      NAME                                   TYPE       CLUSTER-IP      EXTERNAL-IP   PORT(S)                      AGE
      ack-ingress-nginx-v1-edge-controller   NodePort   172.12.0.2      <none>        80:32080/TCP,443:32443/TCP   3m51s

2. Create a file named cube.yaml with the following content.

   apiVersion: v1
   kind: Service
   metadata:
     name: cube-svc
     annotations:
       openyurt.io/topologyKeys: openyurt.io/nodepool
   spec:
     type: ClusterIP
     selector:
       app: cube
     ports:
       - port: 80
         targetPort: 80
   ---
   apiVersion: apps/v1
   kind: Deployment
   metadata:
     name: cube
     labels:
       app: cube
   spec:
     replicas: 4
     selector:
       matchLabels:
         app: cube
     template:
       metadata:
         labels:
           app: cube
       spec:
         containers:
         - name: cube-web
           image: registry.cn-hangzhou.aliyuncs.com/acr-toolkit/ack-cube:1.0
           ports:
           - containerPort: 80

3. Create the cube test application and its Service:

   kubectl apply -f cube.yaml

   Verify that pods are deployed to both the cloud and edge node pools:

   kubectl get pod  -o wide

   Expected output:

   NAME                    READY   STATUS    RESTARTS   AGE   IP            NODE           
   cube-757558c974-9zfkn   1/1     Running   0          11m   10.10.0.1     cloud-node-1   
   cube-757558c974-cw72m   1/1     Running   0          11m   10.10.0.2     cloud-node-2   
   cube-757558c974-fbvlf   1/1     Running   0          11m   10.10.1.1     edge-node-1    
   cube-757558c974-ngwxt   1/1     Running   0          11m   10.10.1.2     edge-node-2    

4. Create a file named ingress.yaml with the following content to define Ingresses for the cloud and edge controllers.

   apiVersion: networking.k8s.io/v1
   kind: Ingress
   metadata:
     name: cube-ingress-cloud
     namespace: default
   spec:
     ingressClassName: ack-nginx-cloud               # Create an Ingress for the cloud Ingress controller.
     rules:
     - host: example.cube.com
       http:
         paths:
         - path: /
           backend:
             service:
               name: cube-svc
               port:
                 number: 80
           pathType: ImplementationSpecific
   
   ---
   apiVersion: networking.k8s.io/v1
   kind: Ingress
   metadata:
     name: cube-ingress-edge
     namespace: default
   spec:
     ingressClassName: ack-nginx-edge                 # Create an Ingress for the edge Ingress controller.
     rules:
     - host: example.cube.com
       http:
         paths:
         - path: /
           backend:
             service:
               name: cube-svc
               port:
                 number: 80
           pathType: ImplementationSpecific

   Create the Ingress resources:

   kubectl apply -f ingress.yaml

5. View the Ingress resources:

   kubectl get ingress

   Expected output:

   NAME                 CLASS             HOSTS              ADDRESS           PORTS   AGE
   cube-ingress-cloud   ack-nginx-cloud   example.cube.com   139.224.xxx.xxx   80      24m
   cube-ingress-edge    ack-nginx-edge    example.cube.com   172.20.xxx.xxx    80      24m

   • A LoadBalancer Service exposes the NGINX Ingress controller in the cloud node pool. The Ingress ADDRESS field is populated with the load balancer address. Use this address to access the Service.

   • A NodePort Service exposes the NGINX Ingress controller in the edge node pool. The Ingress ADDRESS field shows the Service ClusterIP. Access the Service using a node IP address and the NodePort.

References

ACK Edge clusters extend ACK managed Pro clusters with edge node pools for edge nodes and on-premises IDC machines. extends ACK managed Pro cluster with edge node pools. NGINX Ingress in an ACK Edge cluster also supports these capabilities: