GuideTanzu Kubernetes Cluster Example Deployments

Jun 23, 2023 · 2 min read · VMware Tanzu Tanzu Kubernetes Clusters yaml ·

Intro

This post gives a few examples of Tanzu Kubernetes Cluster Manifests and how to deploy them.

Prerequiries

Successfull Installation of vSphere with Tanzu Supervisor
vSphere Namespace created
Already existing VM Classes, Storages Policies and Tanzu Kubernetes Releases assigned to the Namespace.

You will find the IP to your Kubernetes API on the Namespace Option "Link to CLI Tools"

kubectl vsphere login --server=10.40.80.20

kubectl config get-contexts

kubectl config use-context <>

Get Parameters of your vSphere Namespace

VM Class
Storagge Policy
Tanzu Releases

kubectl get vmclass

kubectl get storageclass

kubectl get tanzukubernetesrelease

Cluster with 3 Control and 6 Worker Nodes

apiVersion: run.tanzu.vmware.com/v1alpha3
kind: TanzuKubernetesCluster
metadata:
 name: tkgs-cluster
 namespace: tkgs-cluster-ns   
spec:
  topology:
   controlPlane:
    replicas: 3
    vmClass: best-effort-large  
    storageClass: workload-management-storage-policy
    tkr:
      reference:
       name: v1.23.8---vmware.2-tkg.2-zshippable
   nodePools:
   - replicas: 6
     name: worker
     vmClass: best-effort-large  
     storageClass: workload-management-storage-policy

Save this file as a yaml file on your workstation.

Apply Cluster Manifest

kubectl apply -f clusterspecs.yaml

kubectl get tanzukubernetescluster

Check the Status of the Cluster. "READY" should be true after some minutes.

Edge Cluster

apiVersion: run.tanzu.vmware.com/v1alpha3
kind: TanzuKubernetesCluster
metadata:
 name: tkgs-cluster
 namespace: tkgs-cluster-ns   
spec:
  topology:
   controlPlane:
    replicas: 3
    vmClass: best-effort-small  
    storageClass: workload-management-storage-policy
    tkr:
      reference:
       name: v1.23.8---vmware.2-tkg.2-zshippable
   nodePools:
   - replicas: 3
     name: worker
     vmClass: best-effort-medium  
     storageClass: workload-management-storage-policy

Minimal Cluser for PoC

Use this cluster only for PoC/testing. Only three VMs are created in total. Later, you'll use Kubernetes Taints to enable the Control node to run user workloads. This is useful when testing workloads with a minimal VM footprint when deploying in environments with few resources available.

Concept of Kubernetes Taints In Kubernetes, taints are a way to mark a node with a special attribute that affects which pods can be scheduled onto that node. Taints are used to repel or attract pods based on certain criteria, such as node characteristics or hardware specifications.

apiVersion: run.tanzu.vmware.com/v1alpha3
kind: TanzuKubernetesCluster
metadata:
 name: tkgs-cluster
 namespace: tkgs-cluster-ns   
spec:
  topology:
   controlPlane:
    replicas: 3
    vmClass: best-effort-medium  
    storageClass: workload-management-storage-policy
    tkr:
      reference:
       name: v1.23.8---vmware.2-tkg.2-zshippable

Do not use this in a production environment After a successfull deployment of the cluster run the following commands to allow the Control Nodes to run User Workloads:

kubectl taint nodes --all node-role.kubernetes.io/master-