This article provides the prerequisites and steps to convert to the Enhanced Link Aggregation Control Protocol (LACP) support mode on a vSphere Distributed Switch in vSphere.

Notes:

• Basic LACP has been phased out in vSphere since vSphere Distributed Switch version 5.1. 
• Basic LACP (LACPv1) is only supported on vSphere versions 6.5 or below. 
• Upgrading ESXi to 7.0 may result in the physical switch disabling the LAG ports on ESXi hosts using Basic LACP:
  • "Source vCenter Server has instance(s) of Distributed Virtual Switch at unsupported lacpApiVersion"

VMware vSphere 7.0.x
VMware vSphere 6.x

Prerequisites for enabling Enhanced LACP on current Distributed Switches

• Check each of the port groups on the virtual Distributed Switch (vDS) to ensure that none are overriding the default NIC teaming policy.
• If LACP was already used on the previous version of vDS, ensure that there is only one uplink port group on the vDS.
• Ensure that you are logged into the vSphere Web Client as a user with sufficient privileges to modify the vDS. Both the dvPortgroup.Modify privilege in vCenter Server and Host.Configuration.Modify privilege on the hosts are required.
  
  Note: VMware recommends scheduling a maintenance window before performing the below actions.

Workaround:
If converting to enhanced LACP fails

Converting to the enhanced LACP support from an existing LACP configuration includes several tasks for reconfiguring the Distributed Switch. The conversion may fail if there was a concurrent reconfiguration of the Distributed Switch during the conversion. For example, physical NICs from the hosts have been reassigned to different uplinks or the teaming and failover configuration of the distributed port groups is changed. This issue may also occur if some of the hosts have disconnected during the conversion. As mentioned earlier, creating a new vSphere Distributed Switch is easiest to ensure that Enhanced LACP is enabled. 

When the conversion to the enhanced LACP support fails, it completes only partially. You must check the configuration of the distributed switch and the participating hosts to identify the objects with incomplete LACP configuration and then complete the LACP configuration manually for these objects.

To resolve this issue and to complete the LACP configuration manually

1. Check the LACP configuration of the Distributed Switch and ensure a newly created LAG is present on the Distributed Switch:
   1. In the Distributed Switch, Click Manage and then click Settings.
   2. Select LACP.
   3. If there is no LAG created, click New Link Aggregation Group.
   4. Set the LAG name and number of ports. You can assign physical NICs to each of the LAG ports, which have the same function as standalone uplinks.
   5. Set the LACP mode. If all ports on the corresponding LACP port channel on the physical switch are active, you can leave the LACP mode of the LAG to Passive and vice versa.
   6. Select a load balancing policy for the LAG. All load balancing policies of LACP are supported.
   7. Set the VLAN and NetFlow policies, if applicable.
2. Check if there is an intermediate LACP teaming configuration on the distributed port groups. The newly created LAG must be set to standby to migrate physical NICs to the LAG without dropping connectivity:
   1. Navigate to a distributed port group.
   2. In the Actions menu, click Edit Settings.
   3. Select Teaming and failover.
   4. If there is no LAG in the standby list, move the newly created LAG to that list.
      
      Note: If you do not want to use a LAG to back up the traffic for all distributed port groups, revert the teaming and failover configuration to a state where standalone uplinks are active and the LAG is unused.
       
3. Check the topology of the distributed switch and migrate the standalone uplinks to the LAG ports:
   1. In the Distributed Switch, Click Manage and then click Topology. All physical NICs must be assigned to the LAG ports.
   2. If there are any NICs that remain assigned to the LAG ports, reassign those NICs to the LAG ports.
      
      Note: The LAG must remain in standby in the teaming and failover order of the distributed port groups while you reassign physical NICs to the LAG ports.

As seen in the table below, Basic LACP (LACPv1) is only supported on vSphere versions 6.5 or below.  Upgrading ESXi to 7.0 may result in the physical switch disabling the LAG ports on ESXi hosts using Basic LACP.

LACP Compatibility with vDS

vCenter Server version	vDS version compatibility	Host version	LACP Compatibility
vCenter Server 7.0	vDS 7.0/6.6/6.5	ESXi 7.0	Only supports LACP v2
vCenter Server 6.7	vDS 6.6/6.5/6.0	ESXi 6.7	Only supports LACP v2
vCenter Server 6.5	vDS 6.5/6.0/5.5/5.1/5.0	ESXi 6.5	LACPv1 supports 5.1 LACP v2 supports 5.5/6.0/6.5
vCenter Server 6.0	vDS 6.0/5.5/5.1/5.0	ESXi 6.0	LACPv1 supports 5.1 LACP v2 supports 5.5/6.0

• Limitations of LACP in VMware vSphere
• Sample configuration of EtherChannel / Link Aggregation Control Protocol (LACP) with ESXi/vCenter
• Configuring LACP on a vSphere Distributed Switch Port Group
• Host requirements for link aggregation (etherchannel or port channel) for ESXi and ESX
• Enhanced LACP Support on a vSphere Distributed Switch

Impact/Risks:
As with any configuration change, this should be done during a maintenance window. Making a change on a Distributed Switch means the change will take effect on all hosts connected to the Distributed Switch.

Note: To avoid issues, it is easiest to configure a new vSphere Distributed Switch at the appropriate level for your vCenter and ESXi versions. The new Distributed Switch will automatically have Enhanced LACP. After creating the new Distributed Switch, you can then move your current hosts over to the new switch.