Since it is known that the issue is caused by a device sending unnecessary amount of Unkown Unicast traffic into the ACI Fabric, it is required to figure out which device is causing this, and see if it can be prevented. This is usually caused by devices...
When an ARP or IP packet is forwarded to the spine for a proxy lookup, and the endpoint is unknown in the fabric, a special glean packet is generated and sent to all leafs on the appropriate BD (internal) multicast group address. This triggers an ARP request from...