Skip to content

Supervisor with "NSX + CTGW/Edge/T0"

This section describes the procedures for deploying an application (VMs/K8s) into the VKS Namespace utilizing an "NSX + CTGW/Edge/T0" architecture inside a vSphere environment.

VDS Architecture

Deploy App (K8s)

Topology

Client Operating System

While the command outputs below are captured from a Windows client, the vcf and kubectl CLI tools operate identically across Linux and macOS environments.

Deploy a Full Application (Load Balancer + Pods)

Connect to the K8s Cluster

See Connect to the K8s Cluster

(Optional) Create a Namespace in the K8s Cluster

Create a Namespace if you want to deploy the application (pod) in a specific K8s Cluster Namespace (not default).

kubectl create namespace ns1

Output example 

PS C:\Users\Administrator\Documents> kubectl create namespace ns1
namespace/ns1 created

Allow the deployment of applications in the Namespace.

kubectl label ns ns1 pod-security.kubernetes.io/enforce=baseline
Output example 

PS C:\Users\Administrator\Documents> kubectl label ns ns1 pod-security.kubernetes.io/enforce=baseline
namespace/ns1 labeled

Create application (LB + 2 Pods apache) yaml file

Create file "apache-k8s.yaml"

apache-k8s.yaml file 
apiVersion: v1
kind: Service
metadata:
  name: apache-vip-service
  namespace: ns1
spec:
  type: LoadBalancer
  ports:
    - port: 80
      targetPort: 8080
      protocol: TCP
  selector:
    app: apache-web
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: apache-deployment
  namespace: ns1
spec:
  replicas: 2
  selector:
    matchLabels:
      app: apache-web
  template:
    metadata:
      labels:
        app: apache-web
    spec:
      containers:
        - name: apache
          image: public.ecr.aws/docker/library/httpd:2.4
          ports:
            - containerPort: 8080
              protocol: TCP
          # Intercept startup to write the file, THEN start Apache
          command: ["/bin/sh", "-c"]
          args: 
            - 'sed -i "s/Listen 80/Listen 8080/g" /usr/local/apache2/conf/httpd.conf && echo "<h1>It works - Pod: $HOSTNAME</h1>" > /usr/local/apache2/htdocs/index.html && httpd-foreground'

Deploy the application (LB + 2 Pods apache)

kubectl apply -f apache-k8s.yaml
Output example 

PS C:\Users\Administrator\Documents> kubectl apply -f apache-k8s.yaml
service/apache-vip-service created
deployment.apps/apache-deployment created

Validate deployment of the application (LB + 2 Pods apache)

  • Check application VIP

    kubectl get service apache-vip-service -n ns1

    Output example 

    PS C:\Users\Administrator\Documents> kubectl get service apache-vip-service -n ns1
NAME                 TYPE           CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
apache-vip-service   LoadBalancer   10.101.142.113  10.150.0.9    80:30869/TCP   53s

  • Check application Pods

    kubectl get pods -n ns1

    Output example 

    PS C:\Users\Administrator\Documents> kubectl get pods -n ns1
NAME                                 READY   STATUS    RESTARTS   AGE
apache-deployment-58bf9564f6-57zdv   1/1     Running   0          28s
apache-deployment-58bf9564f6-fmll8   1/1     Running   0          28s

Access the application (LB + 2 Pods apache)

curl http://10.150.0.9
Output example 

PS C:\Users\Administrator\Documents> curl http://10.150.0.9
<h1>It works - Pod: apache-deployment-58bf9564f6-fmll8</h1>
 
PS C:\Users\Administrator\Documents> curl http://10.150.0.9
<h1>It works - Pod: apache-deployment-58bf9564f6-57zdv</h1>

The external load balancer (NSX) load balances (round-robin mode) to the Worker Nodes. Then, each Worker Node load balances to the Apache pods (random). So, to validate the load balancing to the different Apache pods, run the curl command multiple times.