This is regarding Switch statistics in an Omada Wireless network.  Wired Stats are not useful due to counting multipliers.  Looking at statistics for Traffic Distribution, the Wired traffic greatly outnumber the actual payload traffic, which is all wireless.  Almost 10:1!  This network has 3 switches supporting 9 AP's.  There are two StarLink terminals at opposite ends of this system.  One SL terminal is colocated with the ER7206 GW and OC200 controller, but the other SL terminal is at the opposite end of the property and is transported by VLAN through all three switches to get to the GW WAN port.  That means that SL access traffic will flow through each switch possibly 2X and is double counted on it's way back out to the AP's and clients.   Is this normal?  Is there a better way to count traffic to avoid this multiplier effect?  If we actually had wired clients, how to isolate them and count only the ports they are connected to a switch?  In the cut 'n paste below, the Wireless Traffic is the payload.  There are no wired end-user clients, only the GW, 3 switches and AP's are wired.   I don't believe that a SL terminal could even support the wired traffic indicated in the graph.  Thank you!