OTAA loop/debuging

I’m banging my head for this particular problem for over a week now.

I have next configuration:

lopy_node->gw(kersing packet forwarder)->gw_bridge(on gateway)->lora(app)server

When i reset the node the OTAA request gets to lora app server , app server assigns dev_addr, kersing sais that it’s been pushed to concentrator. From the logs it seems to me that the concentrator sends the packet but node doesn’t get it. But if I stop the node to try to Join the network the lora-app is still sending JoinAccept to mqtt broker.

mqtt application/#

{“applicationID”:“2”,“applicationName”:“application”,“deviceName”:“elab-bazz”,“devEUI”:“aaaaddddaaaadddd”,“devAddr”:“00767fe5”}
{“applicationID”:“2”,“applicationName”:“application”,“deviceName”:“elab-bazz”,“devEUI”:“aaaaddddaaaadddd”,“devAddr”:“01ff566d”}

kersin mp_pkt_fwd/#

Jul 02 11:53:18 gwekos03 mp_pkt_fwd[24915]: INFO: tx_start_delay=1495 (1495.500000) - (1497, bw_delay=1.500000, notch_delay=0.000000)
Jul 02 11:53:18 gwekos03 mp_pkt_fwd[24915]: src/mp_pkt_fwd.c:1671:thread_jit(): lgw_send done: count_us=1114807819
Jul 02 11:53:18 gwekos03 mp_pkt_fwd[24915]: 11:53:18 INFO: [down] for server PULL_ACK received in 1 ms
Jul 02 11:53:26 gwekos03 mp_pkt_fwd[24915]: 11:53:26 INFO: [stats] received packet with valid CRC from mote: 05060708 (fcnt=515)
Jul 02 11:53:26 gwekos03 mp_pkt_fwd[24915]: 11:53:26 INFO: [up] PUSH_ACK for server 127.0.0.1 received in 0 ms
Jul 02 11:53:26 gwekos03 mp_pkt_fwd[24915]: 11:53:26 INFO: [down] for server serv_addr[ic] PULL_RESP received - token[51:76] :slight_smile:
Jul 02 11:53:28 gwekos03 mp_pkt_fwd[24915]: 11:53:28 INFO: [down] for server PULL_ACK received in 0 ms

The process then continues to run and is also running if I stop the node. Do you have any idee how I can debug if node recieves anything from this gateway since non private (TTN) implementation works as expected in same configuration. The difference is in packet forwarder routing to loraserver.io

Are you sure that you are not receiving Joinrequest?
try a mosquitto_sub -v -t “application/#” to see on which topics you are receiving the messages

Excuse me for late reply. I was on vacation.

I recorded the flow when request is made. Downstream seems fine to me.

ADDED:

I tried to test against latency issues by installing all stack in one position:
GW+GWBRIDGE and SERVER+APP-SERVER+MQTT on other end the problem remains.

Pycom device don’t get pass the OTAA join. I’m loosing my mind on this. Over TTN as stated works as expected.

The problematic point was in my case the packet forwarder from kersing. After replacing it with lora_pkt_fwd the problem was solved and everything seems to work now - join/up/down communication.

You can try with the new version of mp_pkt_fwd
Kersing has fixed an error and I got the new version running with loraserver