I'm using LoRaWAN GWs LtAP LR8/LR8G LTE6 with ROS 7.20.1 + 7.20.2 to deliver traffic from my LoRa nodes to TTN.
According to the LoRa General Properties page I've created blacklist flters for the complete JoinEUI range (0000000000000000-FFFFFFFFFFFFFFFF) and NetID range (000000-FFFFFF) and added them to the server definition, used for delivering the data to TTN:
/iot lora joineui add joineuis=0000000000000000-FFFFFFFFFFFFFFFF logging=yes name="- block all JoinEUIs" type=blacklist
/iot lora netid add logging=yes name="- block all NetIDs" netids=000000-FFFFFF type=blacklist
/iot lora servers add address=myttnserver.eu1.cloud.thethings.industries joineui="- block all JoinEUIs" name="my TTN server" netid="- block all NetIDs" protocol=UDP
/iot lora set 0 antenna=uFL disabled=no servers="my TTN server"
At the first glance it seems that the filters work correctly. In the log I can see records like "[FLTR] 000038 in range of blacklist" (or 00000f, 000013) for NetID, and "[FLTR] 24e124c0002a0001 in range of blacklist" (or 70b3d50070000000) for JoinEUI - some of them are foreign devices. But when I'm testing my registered nodes, even if I can see matching records in the log (that a blacklist has caught them), I can still see these messages being delivered to the target TTN server, and their details reveal, that the messages were delivered by the GW under test.
Am I overlooking something essential? Do I have to explicitly state somehow, that the "filtered" messages are to be dropped?