VM communication fails after applying or editing a transport node profile due to missing uplink mapping

Products

• VMware NSX

Issue/Introduction

After you applied a transport node profile to a cluster, or you edited a transport node profile attached to a cluster, you see these symptoms.

• VMs on the hosts can not communicate outside the hosts.
• Existing NSX Edge nodes in the same cluster may lose connectivity (tunnels showing Down).
• Running nsxdp-cli vswitch instance list on an affected ESXi host shows the Uplink for the VDS port as void instead of the expected LAG or physical uplink.
  [root@esx:~] nsxdp-cli vswitch instance list
DvsPortset-N (<VDS Name>)   <VDS UUID>
Total Ports:9216 Available:9150
  Client                         PortID          DVPortID                             MAC                  Uplink
...
  <VM name>.ethN                 NNNNNNNN        <UUID>                               <vNIC MAC>           void
• Teaming policy query shows empty activeUplinkPorts.
  [root@esx:~] nsxdp-cli vswitch teaming ls_policy query --dvs-alias <VDS Name>
Name:                   DvsPortset-N-TeamLSP
loadBalance:            FO_EXPLICIT
linkCriteria:           LINK_STATE DRV_PRESENT
flags:                  BEST_EFFORT SHOTGUN NOTIFY_SWITCH
activeUplinkPorts:
standbyUplinkPorts:
  
  • The 2 commands are found in commands/dump-vswitch-info.py.txt of ESXi support bundles.

Environment

VMware NSX

Cause

This issue occurs when the Transport Node Profile (TNP) applied to the cluster is missing required Uplink Mappings in its Teaming Policy.

The NSX configuration delivered to the ESXi hosts will be incomplete, and the virtual switches are unable to map traffic to an uplink, resulting in the void uplink and VM communication failure.

Resolution

1. Log in to the NSX Manager UI.
2. Navigate to System > Fabric > Hosts > Transport Node Profile.
3. Identify and Edit the Transport Node Profile associated with the affected cluster.
4. Open Host Switch and edit the affected VDS.
5. Scroll to the Teaming Policy Uplink Mapping section.
6. Edit the mapping to ensure your logical uplinks (e.g., Uplink-1, Uplink-2, or Lag-3) are correctly mapped to the corresponding VDS Uplinks.
   • Example: Map Lag-3 to VDS Uplink lag-3.
7. Click ADD and APPLY to close the Host Switch window, and click SAVE.
8. Saving the profile will trigger a synchronization task to update the ESXi hosts in the cluster.
9. Verify that the uplinks on the ESXi hosts are no longer in a void state by running nsxdp-cli vswitch instance list 
[root@esx:~] nsxdp-cli vswitch instance list
DvsPortset-N (<VDS Name>)   <VDS UUID>
Total Ports:9216 Available:9150
  Client                         PortID          DVPortID                             MAC                  Uplink
...
  <VM name>.ethN                 NNNNNNNN        <UUID>                               <vNIC MAC>           lag-3

Additional Information

Transport Node Profiles