Workers Configuration

When creating a Cluster managed by Cluster API, you have to define how to manage your worker machines, the machines where Kubernetes nodes and your workload will be hosted.

Instead of managing each worker machine individually one by one, it is a common practice to manage groups of worker machines with a similar configuration by using MachineDeployment or MachinePool resources.

For example, if you are using the MachineDeployment resource (see MachinePools for the MachinePool alternative).

---
apiVersion: cluster.x-k8s.io/v1beta2
kind: MachineDeployment
metadata:
  name: my-cluster-md-0
spec:
  clusterName: my-cluster
  replicas: 5
  template:
    spec:
      bootstrap:
        configRef:
          apiGroup: bootstrap.cluster.x-k8s.io
          kind: KubeadmConfigTemplate
          name: my-cluster-workers
      deletion:
        nodeDeletionTimeoutSeconds: 30
      infrastructureRef:
        apiGroup: infrastructure.cluster.x-k8s.io
        kind: VSphereMachineTemplate
        name: my-cluster-worker-machine
      version: v1.34.0

Tip

ClusterClass users can configure worker options (e.g. replicas and variables) directly from the Cluster object using spec.topology.workers. See Cluster Creation.