Transport nodes status down or degraded with faulty TEP event in NSX clusters

Products

• VMware NSX

Issue/Introduction

• After the deployment of a new workload domain on a NSX cluster, transport nodes (hosts) may report a status of Down or Degraded within the NSX Manager. This is typically accompanied by a faulty TEP event alarm.
• The issue is specifically observed when a Virtual Machine (VM) is connected to the overlay network, which triggers the transport node failure. Initial troubleshooting reveals that TEP-to-TEP connectivity is failing across all nodes in the newly established workload domain.

Environment

VMware NSX

Cause

The issue was caused by an incorrectly configured Uplink Profile applied to the transport nodes in the shared cluster. Specifically, the VLAN assigned to the Tunnel Endpoints (TEPs) was either missing or misconfigured in the profile. As a result, Geneve-encapsulated overlay traffic could not be correctly tagged or routed across the physical network, leading to TEP communication failures when overlay traffic was initiated.

Resolution

To resolve this issue, ensure the transport nodes are using an uplink profile with the correct TEP VLAN ID configured.

1. Log in to the NSX Manager.
2. Navigate to System > Fabric > Profiles > Uplink Profiles
3. Create a new Uplink Profile or edit the profile currently assigned to the cluster.
4. In the Transport VLAN field, enter the correct VLAN ID dedicated to TEP communication.
5. Apply the updated profile to the Transport Node Profile or individual Transport Nodes.

Additional Information

For detailed logs regarding TEP connectivity, you can review /var/run/log/nsx-syslog.log on the affected ESXi hosts to identify VTEP configuration updates. Ensure that your physical network infrastructure supports the VLAN defined in the uplink profile and that MTU settings (recommended 9000 for jumbo frames) are consistent across the path.