Uplink is reported as down when adding a vmnic to a LAG on a vSphere Distributed Switch

Products

• VMware vSphere ESXi

Issue/Introduction

• Shortly after adding a vmnic to a Link Aggregation Group (LAG) on a vSphere Distributed Switch, the uplink is reported as down in the vmkernel.log:
  

  YYYY-MM-DDTHH:MM:SS.SSSZ In(182) vmkernel: cpu6:####)Team.cswitch: TeamVSLACPLAGEventCB:9111: [nsx@6876 comp="nsx-esx" subcomp="vswitch"]Received event LAG STATE, LAG lag#/-####, link DOWN, uplink /0x0, link UNKNOWN

   

• Executing the command esxcli network vswitch dvs vmware lacp stats get shows no Rx or Tx LACPDU errors for the affected vmnic:
  

  DVSwitch  LAG ID      NIC     Rx Errors  Rx LACPDUs  Tx Errors  Tx LACPDUs
  --------  ----------  ------  ---------  ----------  ---------  ----------
  ####   ####  vmnic#          0       83529          0      166078

   

• A packet capture taken at the vmnic level on the ESXi host reveals 00:00:00:00:00:00 P: 0 K: 0 ******** in the PARTNER information of the LACP packets received from the physical switch, even though the packets sent from the ESXi host contain proper information for both ACTOR and PARTNER:
  
  Packet from ESXi:
  

  ## YYYY-MM-DD HH:MM:SS.SSSS <ESXi NIC> Slow-Protocols LACP  124   v1 ACTOR <ESXi MAC Address> P: #### K: #### ****SG*A PARTNER <Physical Switch Mac Address> P: #### K: #### *F***G*A

  
  Packet from the physical switch:

  ## YYYY-MM-DD HH:MM:SS.SSSS <Physical Switch> Slow-Protocols LACP  124   v1 ACTOR <Physical Switch Mac Address> P: #### K: #### *F***G*A PARTNER 00:00:00:00:00:00 P: 0 K: 0 ********

Environment

ESXi 7.x
ESXi 8.x

Resolution

Since the physical switch is not processing the LACP information from the ESXi host, contact the physical switch vendor for further investigation.