Edge Cluster Deletion Fails in SDDC Manager with Error 500030 due to NSX-T Name Mismatch

Products

• VMware SDDC Manager
• VMware Cloud Foundation

Issue/Introduction

• An Edge Cluster deployment in SDDC Manager failed during the validation or activation phase (e.g., BGP configuration issues).
• The Edge Cluster status in SDDC Manager is stuck in ACTIVATING.
• Attempts to remove the failed cluster using the remove_edge_cluster.sh script (referenced in VCF NSX-T Edge Cluster Deployment Removal Tool) fail.
• The script successfully cleans up the edge cluster logical construct but fails to delete the Edge Transport Nodes.
• The /home/vcf/cleanup/cleaner/edge_cluster_cleaner.log contains the following error:

  httpStatus = BAD_REQUEST 
  Delete attempt failed for edge-node-## /api/v1/transport-nodes/<UUID>: 
  {'error_code': 500030, 
  'error_message': 'The object ' 
  				 'path=[/infra/sites/default/enforcement-points/default/edge-transport-node/<UUID>] ' 
  				 'cannot be deleted as either it has children or it is being ' 
  				 'referenced by other objects ' 				
                                   'path=[/infra/sites/default/enforcement-points/default/edge-clusters/<UUID>/edge-nodes/0]', 
  'httpStatus': 'BAD_REQUEST', 
  'module_name': 'Policy'} 

   

• In NSX Manager, the Edge Cluster object exists but has a different name than the one recorded in the SDDC Manager Inventory (e.g., edge-cluster-to_be_deleted-01 in NSX vs edge-cluster-to_be_deleted in SDDC Manager).

Environment

• VMware Cloud Foundation
• VMware NSX-T

Cause

• This issue occurs when the Edge Cluster name in the NSX Manager inventory does not match the Edge Cluster name in the SDDC Manager database.
• The remove_edge_cluster.sh script relies on the SDDC Manager inventory name to locate and delete the corresponding Edge Cluster object in NSX-T. If the name in NSX-T has been modified (manually renamed or automatically appended with a suffix during a partial/failed workflow), the script fails to identify and delete the Edge Cluster object.
• Consequently, when the script attempts to delete the individual Edge Transport Nodes, the operation fails with Error 500030 because those nodes are still members of the existing (but misnamed) Edge Cluster in NSX-T.

Resolution

To resolve this issue, rename the Edge Cluster in NSX Manager to match the name expected by SDDC Manager, then re-run the cleanup script.

 

Follow the below steps

1. Verify the correct name in SDDC Manager:
   
   
   1. Log in to the SDDC Manager UI.
   2. Navigate to Workload Domains > Click on the Workload Domain name > Edge Clusters
      
      Note the exact name of the stuck cluster (e.g., edge-cluster-to_be_deleted).
      
      
2. Rename the Cluster in NSX Manager:
   
   
   1. Log in to the NSX Manager UI.
   2. Navigate to System > Fabric > Nodes > Edge Clusters.
   3. Locate the target Edge Cluster. It may have a suffix (e.g., edge-cluster-to_be_deleted).
   4. Edit the Edge Cluster and rename it to match the SDDC Manager name exactly (e.g., change edge-cluster-to_be_deleted-ec to edge-cluster-to_be_deleted).
   5. Save the configuration.
      
      
3. Re-run the Cleanup Script:
   
   
   1. Download the script from KB VCF NSX-T Edge Cluster Deployment Removal Tool and copy to SDDC manager using any file transfer tool.
   2. SSH into the SDDC Manager as vcf user and su to root
   3. Navigate to the directory containing the cleanup script.
   4. Execute the script again:

      ./remove_edge_cluster.sh --cluster <edge_cluster_to_be_deleted> --user <sso_user>

       

4. Verification:
   
   
   1. Confirm the script completes successfully without error 500030.
   2. Verify the Edge Cluster and Edge Nodes are removed from NSX Manager, vCenter, and SDDC Manager.