NSX Edge HA Status "DOWN" due to missing VLAN tagging on TOR switchports

Products

• VMware NSX

Issue/Introduction

In the NSX UI, the Tier-0 gateway displays an HA status of DOWN for Edge Node.

Symptoms include:

• Edge HA status appearing as DOWN in the NSX Manager interface.
  

• Logging into Edge01 CLI and running "get bgp neighbor summary" shows the BGP neighbors configured:
  

• BGP neighbors failing to respond to ICMP pings from the Edge CLI.
  

• Connectivity is restored only when the Edge VM is migrated to a specific ESXi host.

Environment

VMware Cloud Foundation (VCF), NSX-T / NSX 4.x

Cause

Physical upstream Top-of-Rack (TOR) switchports are missing required VLAN tagging for Edge Tier-0 uplink segments on specific ESXi hosts.

Resolution

1. Identify Missing VLANs: Verify the specific VLANs assigned to the Edge Uplink Segments.

2. Audit Physical Switchports: Inspect the configuration of all physical TOR switchports connected to the ESXi hosts within the Edge Cluster.

3. Correct Trunk Configuration: Ensure all required Edge VLANs are tagged (allowed) on every physical uplink port for each ESXi host in the cluster.

4. Validate Connectivity:

   • Execute get bgp neighbor summary from the Edge CLI to confirm BGP state is Established.
   • "ping <bgp_neighbor>" from T0_SR_VRF_ID

 

Additional Information

NOTE:

Please review the following KB below for more information around how to troubleshoot NSX BGP connectivity: 339392