Administrators are unable to execute the Day-2 action to "Update VM Class" on a Tanzu Kubernetes Cluster deployed through VMware Aria Automation. The Day-2 action immediately fails when triggered from the Aria Automation user interface.
Observed Errors:
UI Error: Update VM Class cannot be performed on <Cluster_Name> since the state of the resource has changed.
Log Error: A 404 Not Found error is returned when the Day-2 action is triggered.
Isolation testing confirms that other Day-2 actions (such as "Scale Worker Nodes") succeed, and updating the VM class natively via kubectl directly on the cluster succeeds. The issue is entirely isolated to the Aria Automation Day-2 "Update VM Class" workflow API call.
VMware Aria Automation 8.x
vSphere 8.0 Update 3+
Tanzu Kubernetes Grid
The failure is caused by an API incompatibility introduced by newer versions of vSphere.
Aria Automation 8.x hardcodes its queries to look for a specific legacy resource endpoint (VirtualMachineClassBinding) to update the VM class. However, in vSphere 8.0 Update 3+, the underlying VM Operator was updated to v1alpha2. This architectural shift made VirtualMachineClass natively namespace-scoped, completely removing the legacy VirtualMachineClassBinding resource definition from the Supervisor Cluster. Because the endpoint no longer exists, Aria Automation hits a 404 Not Found error when it attempts to execute the Day-2 action.
Additionally, the Kubernetes integration in Aria Automation (including TKG cluster management) was officially deprecated as of version 8.18 and is slated for removal. It is on a deprecation path with no planned enhancements to support the new vSphere v1alpha2 APIs.
Please utilize the immediate Workaround (Native Kubernetes Management)
Continue to use native Kubernetes commands to update the VM Class directly on the cluster.
Execute kubectl edit tanzukubernetescluster to apply the required VM class updates.
Strategic Resolution (Recommended Long-Term)
Plan a migration to VMware Cloud Foundation (VCF) 9.x. The new Cloud Consumption Interface (CCI) via the All Apps organization mode is the officially supported, forward-looking architecture for native Cluster API management.