Determining why new Kubernetes worker node(s) are rolled out in VKS cluster

Products

• VCF Private AI Services

Issue/Introduction

New worker nodes are being deployed in the guest cluster and are replacing the existing worker nodes.

Capi controller logs confirm the rollout:

####-##-##T##:#:##.#########Z stderr F I0312 10:31:10.658226 1 controller.go:362] "Reconcile KubeadmControlPlane" controller="kubeadmcontrolplane" controllerGroup="controlplane.cluster.x-k8s.io" controllerKind="KubeadmControlPlane" KubeadmControlPlane="######/#####-#####" namespace="######" name="#####-#####" reconcileID="########-####-####-####-############" Cluster="######/#####" 
####-##-##T##:#:##.#########Z stderr F I0312 10:41:34.444583 1 controller.go:464] "Rolling out Control Plane machines: Machine #####-#####-##### needs rollout: Machine KubeadmConfig InitConfiguration or JoinConfiguration are outdated: diff: ... machine-agent-configuration-v11 → machine-agent-configuration-v12" 
####-##-##T##:#:##.#########Z stderr F I0312 10:41:34.484460 1 scale.go:107] "Machine created (scale up)" controller="kubeadmcontrolplane" ... Machine="#####-#####-#####"

Environment

vSphere 8

vSphere 9

Cause

The VKS cluster is currently undergoing a rolling update. During this process, new worker nodes are provisioned and existing nodes are gradually replaced to apply the requested configuration or version change.

Resolution

Once a rolling update has been initiated, it must be allowed to complete for the configuration change to be fully applied. The process will automatically replace the existing worker nodes and return the cluster to a stable state once all nodes have been updated.

Additional Information

Understanding the Rolling Update Model for VKS Clusters