Relocate and Remove an NSX Edge Node from an NSX Edge Cluster in NSX-T 4.0.1.1 and later versions.
search cancel

Relocate and Remove an NSX Edge Node from an NSX Edge Cluster in NSX-T 4.0.1.1 and later versions.

book

Article ID: 395026

calendar_today

Updated On:

Products

VMware NSX

Issue/Introduction

Removing an edge from and edge cluster fails as it is being referenced by a logical router.

The error observed in the UI is as follows:

Transport node delete Failed: The object path=[/infra/sites/default/enforcement-points/default/edge-transport-node/########-####-####-####-############] cannot be deleted as either it has children or it is being referenced by other objects path=[/infra/sites/default/enforcement-points/default/edge-cluster/########-####-####-####-############/edge-nodes/0 

Environment

VMware NSX 4.0.1.1 and later.

Cause

If the documentation meant for versions prior to NSX-T 4.0.1.1 are followed in an environment running 4.0.1.1 and later relocation and removal will not work, therefore the new API method is required.

 

Resolution

The process to relocate and remove differs in version prior to 4.0.1.1.

Steps for NSX-T 4.0.1.1 and later:

Relocate and Remove Edge Node – NSX 4.0.1.1+

Requirements Before Using the API

  • The Edge node must not have manually allocated service configurations.

  •  The target Edge cluster must contain at least two healthy Edge nodes for relocation.

  • Layer 2 bridging must not be configured on standby nodes.

  •  For HA, it’s recommended to have more than two healthy nodes.

What Gets Relocated Automatically

When conditions are met, the following auto-allocated services will be relocated:

  • Tier-0 and Tier-1 Service Routers

  • DHCP server

  • Metadata proxy

  • L2 Forwarder

Steps to Relocate and Remove an Edge Node

1. Get the Edge Cluster Member Index

Run this API call to retrieve the member_index of the Edge node you want to remove:

GET https://<nsx-manager-IP>/policy/api/v1/edge-clusters/<edge-cluster-id>

{    
  "deployment_type": "VIRTUAL_MACHINE",
  "members": [
      {
          "member_index": 11,
          "transport_node_id": "21a19cbf-eaba-4a59-b18d-ff71fe5d76aa",
          "display_name": "edgeVm1New"
      },
      {
          "member_index": 13,
          "transport_node_id": "740cf97d-892b-47bb-97e7-889d92252e80",
          "display_name": "edgeVm2New"
      },
      {
          "member_index": 14,
          "transport_node_id": "cd5ab447-a36a-4bc3-94ff-0a4eea9fb2ad",
          "display_name": "edgeVm3New"
      }
  ],
 

  • Example output shows each member with:

    • member_index

    • display_name

    • transport_node_id

Identify the member_index of the Edge node to remove. Assume that you want to relocate the service configurations for the Edge node named edgeVm1New, then its member_index value is 11

2. Run the Relocate and Remove API

Use the POST method with the member_index value:

POST https://<nsx-manager-IP>/api/v1/edge-clusters/<edge-cluster-id>?action=relocate_and_remove

{
  "member_index": 11
}

  • This triggers:

    • Maintenance mode for the Edge node

    • Automatic relocation of all eligible service configurations

    • Removal of the Edge node from the cluster after transfer completes

Caveats and Error Handling

  •  The API will fail silently if:

    • The node has manual allocations

    • The cluster lacks two healthy standby nodes

  •  If relocation fails after a success response, you’ll get an alarm:
    Edge Cluster Member Relocate Failure

Recommended Action:

  • Check Edge cluster capacity and health

  • Scale up cluster if needed

  • Re-run the relocate/remove API call

For additional information please refer to this document: Relocate and Remove an NSX Edge Node from an NSX Edge Cluster

 

Prior to 4.0.1.1 the Edge removal process is as follows:

Before You Begin

  • Tier-0 configurations (e.g., SR, DHCP, metadata proxy) must be manually relocated to a standby Edge node.

  • Tier-1 configurations can be:

    • Auto-allocated → use Standby Relocation feature.

    • Manually allocated → must be manually moved to another Edge node.

Steps for Tier-0 Configurations

  1. Log into NSX Manager with admin credentials.

  2. Navigate to Networking → Tier-0 Gateways.

  3. Edit the Tier-0 Gateway:

    • Go to the Interfaces section.

    • Edit interfaces bound to the current Edge node.

    • Replace the Edge node with a new one that has matching VLAN connectivity.

    • Save changes to move Tier-0 SR, DHCP, and metadata proxy configs.

Steps for Auto-Allocated Tier-1 Configurations

  1. (Optional) Adjust BFD timers (e.g., 500 ms) and apply updated Edge Cluster Profile for faster failover.

  2. Navigate to Networking → Tier-1 Gateways.

  3. Edit the Tier-1 Gateway:

    • Enable Standby Relocation.

    • Ensure another healthy Edge node exists in the cluster.

  4. Navigate to System → Fabric → Profiles → Edge Cluster Profiles.

  5. Edit the profile:

    • Set the Standby Relocation Threshold (e.g., 30 mins).

  6. Power off the NSX Edge node (do not put it into maintenance mode).

  7. Wait for the threshold period to allow automatic failover of Tier-1 services.

  8. After relocation, go to the Edge Cluster settings, remove the powered-off Edge node, and click Save.

Important Notes

  • Tier-0 configurations are never auto-relocated — manual changes are required.

  • Manually allocated Tier-1 configs also need manual reassignment.

  • Standby relocation works only for auto-allocated Tier-1 services, and requires a healthy standby Edge.

 

For more detailed information please refer to the document affixed to this KB below:

Remove NSX Edge Nodes from an Edge Cluster

 

Additional Information

Powered by Wolken Software