RDE Stuck at 1%

Products

• VMware Cloud Director

Issue/Introduction

• When creating a Kubernetes cluster using VMware Cloud Director Container Service Extension (CSE), the VM provisioning fails leaving the RDE stuck at 1% with no proper failure handling or UI/API indication.
• It is not possible to delete the failed created cluster from the VMware Cloud Director (VCD) UI.
• When checking the entity via  API call, the entityState is set to "PRE_CREATED". For additional information on the entity GET API call visit Resolve Entity

Environment

VMware Cloud Director 10.5.1

Cause

If the VM provisioning fails, the entityState is set to entityState is set to "PRE_CREATED" and the cluster cannot be and deleted it because it must be in either "RESOLVED" or "RESOLUTION_ERROR" state.

Resolution

Currently there is no resolution.

Workaround

1. Delete existing vApp associated with the failed cluster (they should have the same name as the cluster, if any vApp remained). To delete the vApp refer to the documentation Delete a vApp in the VMware Cloud Director Tenant Portal
2. Double check to ensure the Load Balancer components (virtual service, LB pool members) are not present for the cluster to delete, refer to the documentation Managing NSX Advanced Load Balancing on an NSX Edge Gateway in VMware Cloud Director. They are prefixed with the cluster name
3. In the VCD UI, click on the cluster and take note of the Cluster ID. Check the documentation View Tanzu Kubernetes Grid Cluster Information
4. Execute the Resolve Defined Entity API requests to change the entityState to "RESOLVED" using the Cluster ID noted from Step 3.
   
   POST https://{api_host}/cloudapi/1.0.0/entities/{id}/resolve


5. Delete the cluster running the Delete Defined Entity API request using the Cluster ID noted from Step 3 
   
   DELETE https://{api_host}/cloudapi/1.0.0/entities/{id}

Additional Information

Subscribe to this knowledge article to get updates on this issue