Deploying and Exposing Basic Windows Workloads

This topic describes deploying Windows worker-based Kubernetes clusters in Tanzu Kubernetes Grid Integrated Edition (TKGI).

Overview

In Tanzu Kubernetes Grid Integrated Edition, you can deploy Windows-based workloads to Kubernetes clusters on vSphere with NSX-T. Additionally, TKGI provides beta support for deploying Windows-based workloads to Kubernetes clusters on vSphere without NSX-T.

To deploy a new Windows-based workload to a new pod, do the following:

1. Access Your Windows-Based Cluster
2. Configure a Pod Deployment Manifest
3. Deploy the Pod
4. Configure a Service Manifest
5. Deploy the Service

Warning: Support for Windows-based Kubernetes clusters is enabled for TKGI on vSphere with NSX-T and as a beta feature for TKGI on vSphere without NSX-T.

Do not enable this feature if you are using TKGI with Google Cloud Platform (GCP), Azure, or Amazon Web Services (AWS).

Prerequisites

You can deploy Windows workloads to Windows-based clusters only.

You must configure the Tanzu Kubernetes Grid Integrated Edition tile to support Windows-based clusters before you can use Windows-based clusters in TKGI. For instructions on configuring the Tanzu Kubernetes Grid Integrated Edition tile, see Configuring Windows Worker-Based Clusters.

Access Your Windows-Based Cluster

Your command line must have access to your Windows-based cluster to deploy Windows VMs and workloads to the cluster.

1. To determine which of your existing clusters is Windows-based, use the following command:

   tkgi clusters
   

   For example:

    $ tkgi clusters
   
   Name                Plan Name             UUID                                  Status     Action
    windows-k8s         Plan-11-Windows-Beta  881543kd-64fg-7826-hea6-3h7g1o04kh0e  succeeded  Create
    second-windows-k8s  Plan-11-Windows-Beta  951547dl-67kg-9631-bju8-7h9s3o98br0q  succeeded  Create
    

   Only clusters configured on Plans 11, 12, or 13 are Windows-based.

2. To access your Windows-based cluster, run the following command:

   tkgi get-credentials CLUSTER-NAME
   

   Where CLUSTER-NAME is the name of your Windows-based cluster.

   For example:

   $ tkgi get-credentials windows-k8s
    Fetching credentials for cluster windows-k8s.
    Context set for cluster windows-k8s.
   
   You can now switch between clusters by using:
    $kubectl config use-context <cluster-name>
    

   The tkgi get-credentials command creates a local kubeconfig, allowing you to manage the cluster from the command line. For more information about the tkgi get-credentials command, see Retrieving Cluster Credentials and Configuration.

3. To verify you have established access to the correct cluster, run the following command:

   kubectl cluster-info
   

4. (Optional) To review the existing pods in the cluster, run the following command:

   kubectl get pods
   

Deploy a Windows Worker Pod

A pod deployment manifest file configures the VMs deployed to a pod.

Configure a Pod Deployment Manifest

You must create a Windows worker deployment manifest before deploying your new Windows worker pod.

1. To create a Windows worker deployment manifest, create a new YAML file containing the following:

   ---
   apiVersion: apps/v1
   kind: Deployment
   metadata:
     labels:
         app: POD-NAME
     name: POD-NAME
   spec:
     replicas: 1
     selector:
       matchLabels:
         app: POD-NAME
     template:
       metadata:
         labels:
           app: POD-NAME
         name: POD-NAME
       spec:
         containers:
         - name: CONTAINER-NAME
           image: CONTAINER-FILE:VERSION
           env:
           - name: PORT
             value: "80"
           ports:
           - name: http
             containerPort: 80
         nodeSelector:
           kubernetes.io/os: windows
         tolerations:
         - key: "windows"
           operator: "Equal"
           value: "2019"
           effect: "NoSchedule"
   

   Where:

   • POD-NAME is the name of your pod.
   • CONTAINER-NAME is the internal name of your container.
   • CONTAINER-FILE is the filename of your container.
   • VERSION is the version of the image to use, for example latest.

   For example:

   ---
   apiVersion: apps/v1
   kind: Deployment
   metadata:
     labels:
         app: win-webserver
     name: win-webserver
   spec:
     replicas: 1
     selector:
       matchLabels:
         app: win-webserver
     template:
       metadata:
         labels:
           app: win-webserver
         name: win-webserver
       spec:
         containers:
         - name: windowswebserver
           image: stefanscherer/webserver-windows:0.4.0
           env:
           - name: PORT
             value: "80"
           ports:
           - name: http
             containerPort: 80
         nodeSelector:
           kubernetes.io/os: windows
         tolerations:
         - key: "windows"
           operator: "Equal"
           value: "2019"
           effect: "NoSchedule"
   

Deploy the Pod

1. To deploy a new Windows worker pod, run the following command:

   kubectl apply -f POD-CONFIG-FILE
   

   Where POD-CONFIG-FILE is the filename of the Windows worker deployment manifest created above.

2. To confirm the status of the new pod, and the creation of new Windows worker nodes, run the following commands:

   kubectl get pods
   kubectl get nodes -o wide
   

   For example:

    $ kubectl apply -f first-k8s.yml
    deployment.extensions/win-webserver created
   
   $ kubectl get pods  -o wide
    NAME                             READY  STATUS   RESTARTS  AGE  IP            NODE                                  NOMINATED NODE   READINESS GATES
    win-webserver-795g866cd7-58oct   1/1    Running  0         88s  10.200.42.4   0983934a-6d69-8e5g-g3k1-98r8r56l345j  <none>           <none>
   
   $ kubectl get nodes -o wide
    NAME                                    STATUS   ROLES    AGE    VERSION    INTERNAL-IP   EXTERNAL-IP   OS-IMAGE                          KERNEL-VERSION      CONTAINER-RUNTIME
    0983934a-6d69-8e5g-g3k1-98r8r56l345j    Ready    <none>   19d    v1.14.1    10.85.41.118  10.85.41.118  Windows Server 2019 Datacenter    10.0.17763.503      docker://18.9.0
    6d69934a-7d43-9g3g-h4d1-54r9r97l395j    Ready    <none>   19d    v1.14.1    10.85.41.115  10.85.41.115  Ubuntu 16.04.6 LTS                4.15.0-50-generic   docker://18.9.0
    7636d69a-2e75-5l0g-k6m1-76r3r37l729k    Ready    <none>   19d    v1.14.1    10.85.41.117  10.85.41.117  Windows Server 2019 Datacenter    10.0.17763.503      docker://18.9.0
    406d694a-9g96-2d3g-f3j1-32r1r44l342x    Ready    <none>   19d    v1.14.1    10.85.41.116  10.85.41.116  Windows Server 2019 Datacenter    10.0.17763.503      docker://18.9.0
    

   In the preceding example a new pod is created, and creation and status of the new pod and new nodes verified.

Note: The ping command does not work reliably for Windows workers. For more information, see Pinging Windows Workers Does Not Work in Release Notes.

Deploy a Service to a Windows Worker Pod

A service deployment manifest file configures your service, defining how your service will run and how it will be exposed.

Configure a Service Manifest

You must create a Windows service deployment manifest before deploying your Windows worker workload.

1. To create a Windows service deployment manifest, create a new YAML file containing the following:

   ---
   apiVersion: v1
   kind: Service
   metadata:
     name: APP-NAME
     labels:
       app: APP-NAME
   spec:
     ports:
       # the port that this service should serve on
     - port: 80
       targetPort: 80
     selector:
       app: APP-NAME
     type: LoadBalancer
   

   Where APP-NAME is the name of your Windows service.

   For example:

   ---
   apiVersion: v1
   kind: Service
   metadata:
     name: win-webserver
     labels:
       app: win-webserver
   spec:
     ports:
       # the port that this service should serve on
     - port: 80
       targetPort: 80
     selector:
       app: win-webserver
     type: LoadBalancer
   
   

Deploy the Service

1. To expose your service on a LoadBalancer, run the following command:

   kubectl apply -f SERVICE-CONFIG-FILE
   

   Where SERVICE-CONFIG-FILE is the filename of your Windows service deployment manifest created above.

   For example:

    $ kubectl get services
    NAME          TYPE          CLUSTER-IP      EXTERNAL-IP    PORT(S)    AGE
    kubernetes    ClusterIP     10.100.200.1    <none>         443/TCP    20d
   
   $ kubectl apply -f first-k8s-service.yml
    service/win-webserver created
   
   $ kubectl get services
    NAME          TYPE         CLUSTER-IP        EXTERNAL-IP  PORT(S)        AGE
    kubernetes    ClusterIP    10.100.200.1      <none>       443/TCP        20d
    win-webserver LoadBalancer 10.100.200.221    <none>	    80:32073/TCP   5s
   
   
    $ curl 10.85.41.118:32073
    <pre>
    <a href=“License.txt”>License.txt</a>
    <a href=“ProgramData/”>ProgramData/</a>
    <a href=“Users/”>Users/</a>
    <a href=“WcSandboxState/”>WcSandboxState/</a>
    <a href=“Windows/”>Windows/</a>
    <a href=“var/”>var/</a>
    <a href=“webserver.exe”>webserver.exe</a>
    </pre>
    

   In the preceding example a new service is created, verified, and validated.