Unable to increase Worker Node VM disk size from VM Class settings
search cancel

Unable to increase Worker Node VM disk size from VM Class settings

book

Article ID: 432689

calendar_today

Updated On:

Products

VMware vSphere Kubernetes Service

Issue/Introduction

Unable to increase the Worker Node disk size from the VM Class Edit Settings option within the vSphere UI. The option to expand the storage capacity available to the node is not present, preventing the expansion of the Worker Node's native disk to meet workload requirements.

Environment

VMware Kubernetes Service

Cause

The Worker Node disk size is strictly dictated by the OVA template associated with the vSphere Kubernetes releases (VKr), rather than the VM Class definition. Product architecture binds the VM disk capacities for these nodes to the initial OVA deployment specifications. The default storage footprint for a standard VKr image is hardcoded as menrtioned in Using VM Classes with VKS Clusters

Resolution

To accommodate increased workload data storage requirements, use one of the following two supported methods:

  • Use Persistent Volumes

Utilize Persistent Volumes (PVs) to handle workload data requirements. This method effectively abstracts data storage from the boot disk and mitigates the need to expand the node's native disk. This aligns with Kubernetes storage best practices by separating ephemeral node storage from persistent application data.

  • Create a Custom Image

Create a custom VKr machine image to define a larger root disk during the build process as mentioned in Building a Custom VKr Machine Image. This directly alters the base OVA footprint, permanently resolving the capacity constraint for new node deployments.