How to Migrate NSX Edge Node Cluster to a Different Compute Cluster and Storage in VMware Cloud Foundation (VCF)
search cancel

How to Migrate NSX Edge Node Cluster to a Different Compute Cluster and Storage in VMware Cloud Foundation (VCF)

book

Article ID: 443589

calendar_today

Updated On:

Products

VMware NSX

Issue/Introduction

  • Need to move an existing NSX Edge Node cluster to a new compute cluster within the same workload domain.
  • The destination cluster uses a different storage platform (e.g., migrating from vSAN to External Storage).
  • Goal is to maintain existing management IP addresses for the Edge nodes.

Environment

VCF 5.2

 

Cause

  • Standard vSphere vMotion and Storage vMotion of Edge VMs are not supported for relocation between compute clusters managed by an SDDC Manager.
  • SDDC Manager inventory and metadata require the use of built-in workflows to maintain synchronization.

Resolution

To relocate the Edge cluster, use the Expand then Shrink method via SDDC Manager.

Option 1: Sequential Replacement (Supports IP Reuse)

This method allows you to reuse the same management IPs by replacing nodes one by one.

  1. Maintenance Mode: Place the first Edge Node in the existing cluster into maintenance mode.
  2. Shrink Cluster: From the SDDC Manager UI, use the Shrink Edge Cluster workflow to delete this specific Edge Node. This releases its management IP back into the SDDC inventory.
  3. Expand Cluster: Use the Expand Edge Cluster workflow. Select the destination compute cluster and new storage. Deploy a replacement node using the IP address released in step 2.
  4. Repeat: Once the first new node is active and healthy, repeat the process for the remaining nodes in the cluster.
  5. Perform a Domain Sync. See Reference #5 below.

Option 2: Parallel Deployment (Minimal Risk)

This method involves deploying a full set of new nodes before removing the old ones.

  1. Expand Cluster: Use the Expand Edge Cluster workflow to deploy two (or more) new replacement Edge nodes in the destination cluster/storage using temporary/new IP addresses.
  2. Service Migration: Migrate T0/T1 services from the old nodes to the new nodes.
  3. Validation: Verify all edges are active, synced, and healthy in the NSX and SDDC Manager UI.
  4. Shrink Cluster: Use the Shrink Edge Cluster workflow to delete the original Edge VMs.
  5. (Optional) Revert IPs: Once the old nodes are deleted, their original IPs can be re-added to the SDDC inventory and applied to the new nodes if desired (requires a second replace/redeploy cycle).
  6. Perform a Domain Sync. See Reference #5 below.

Additional Information

References:

  1. Migrate NSX edge nodes in a VCF Environment (392093) — Overview of the stretch/shrink workaround.
  2. NSX Edge vMotion best practices (317145) — General considerations for Edge node movement.
  3. Add Edge nodes to an NSX Edge Cluster in VCF — VCF Tech doc for adding Edge Nodes to a Cluster
  4. Remove Edge nodes from an NSX Edge Cluster in VCF  — VCF Tech doc for adding Edge Nodes to a Cluster
  5. Synchronize Inventory versions afterr out-of-band upgrade in a VCF Environment — Domain Sync