This article describes a use case in which customer wants to connect an MC-AE interface with "force-up" enabled to a downstream ESXi A-A mode host who does not support LACP.

The customer wants to connect an MC-AE interface to a downstream ESXi A-A mode host to achieve redundancy. The customer wants two MC-AE child ports active at the same time to load balance server traffic.

The current MC-LAG environment is being deployed with A-A mode and it works fine. However, the downstream host does not support LACP, which made MC-AE ports on both peers down.

LACP "force-up" option can be used in this use case but JTAC lab test verifies that "force-up" can only keep one MC-AE child port up and the other down. the second MC-AE port remains down no matter what.

As a result, there will be traffic blackhole and, theoretically, 50% traffic loss if the ESXi host hashed traffic to the MC-AE port who is down because ESXi host is using static LAG (without LACP) so it keeps its two child port up all the time.

JTAC checked with PLM team and it's confirmed that this is expected behavior. MC-AE with "force-up" option cannot make two child port up to connect static LAG such as ESXi host.

Customer needs to look for other solutions such as VC to connect A-A mode ESXi host with static LAG.