Decommissioning NSX Tier-0 Gateway with Active VRFs and BGP Configurations
Article ID: 435238
Updated On:
Products
Issue/Introduction
The user needs to decommission a Tier-0 Gateway in an NSX environment. While no Tier-1 Gateways or Segments are currently attached, the Tier-0 Gateway has multiple VRF Lite instances connected with active BGP peering sessions.
Environment
VMware NSX
Cause
Standard decommissioning of a parent Tier-0 Gateway is blocked if child objects, such as VRF instances or active routing protocols, are still provisioned.
Resolution
Precautions:
- Traffic Validation: Ensure there is no active production traffic passing through the VRFs or the parent Tier-0 Gateway before starting the decommissioning.
NOTE: Confirm that all necessary maintenance window approvals and architectural sign-offs are in place for the removal of these routing components.
Configuration Backup: Perform a manual NSX backup before starting the decommissioning process.
Service Impact: Confirm no management traffic or specialized services (e.g., Load Balancing, VPN) are pinned to these VRFs.
Step-by-Step Decommissioning:
Remove VRF BGP Neighbors:
Navigate to Networking > Tier-0 Gateways.
Edit each VRF Gateway associated with the parent Tier-0.
Go to BGP > Neighbors and delete all configured neighbors.
Delete VRF Gateways:
Once BGP is cleared, delete each VRF Gateway instance.
If the deletion fails, ensure no stale Route Maps or Prefix Lists are explicitly tied to the VRF.
Clear Parent Tier-0 BGP and Routing:
Edit the parent Tier-0 Gateway.
Disable BGP and remove any static routes.
Delete BGP Neighbors on the parent Tier-0 if any exist.
Remove Uplink Interfaces:
Navigate to Interfaces on the Tier-0 Gateway.
Delete all Uplink interfaces. This includes interfaces on the parent Tier-0 used for external connectivity.
Delete the Tier-0 Gateway:
Select the Tier-0 Gateway and click Delete.
Additional Information
Feedback
