That wouldn’t make sense to have the 900 mbps speed in one way and not the other. The cable is the same.
This would make perfectly sense since in the cable are 4 pairs of wires, some are used for sending and others for receiving. On gigabit ethernet every wire is used.
Simplest test would be to just “turn around” the cable and run the test again. If the Pi receives now 900 Mbit/s but can only send 135 then you know the cable is broken somehow.
2 Likes
I just wanted to say 
.
1 Like
I’ve tried with another cable, issue persists. This could be something related with VLANs, because when both devices are under the same lan tag, speeds are on the 900+ range both ways.
but this would be a challenge of the switch you are connected to?
How so? It’s connected to a switch which in its turn is connected into the USG.
but the VLAN ID you configure on your switch has no effect on DietPi because it simply doesn’t matter for the DietPi device.
It goes without saying that if you are testing for throughput, you should eliminate as many intermediate devices as possible. We were under the impression that the test takes place with direct connection and suddenly we learn that there are even multiple vlans involved. First verify the speed with a direct connection between the Pi and the laptop. Then start adding switches and routers in the path.
2 Likes