@marki can you clarify from the iperf which side is the DC (is it 10.x??) and which side is the remote? Remember that the iperf client sends the large transfer
I've updated the UDP test part. And also the final question because I don't understand what is being dropped where and by whom and why and why TCP doesn't simply handle the fact that there is (real) 100Mb at one end, and only 10Mbps at the other.