join refused: DevNonce already used by this device about iotedge-lorawan-starterkit

[ronniesa]

Are you using 1.0 or 1.01?
Can you try first to restart Redis from the Azure portal and right after the gateway and then turn on the device. If this does not work can you try deleting the device from IoT Hub, restart Redis and the Gateway.

Please report back which action is correcting the problem it could be a bug or the device reusing the DevNonce when it should not.

The DevNonce is a random key that the device generates for the join and it must be always new. So in the code we compare it to the last successful one. It is stored as reported property in the twins.

[gerfen]

@ronniesa Sorry for the delay in responding - I've been traveling. I'm currently using 1.0 - waiting for the next release so I can move to the Edge 1.07 release. The error eventually cleared but I was not able to test with both gateways at play. I am returning next week to the customer site and can do another test to see if I am still having the same issue. I am hopeful that I will not since I was able to associate all of the devices (over 100 to the correct gateway via the twins metadata in Azure).

I will keep you posted just as soon as I have more information.

[gerfen]

@ronniesa I'm on site today testing some class c and class a devices. I'm seeing the same issue. Restarting Redis and then the test gateway resolved the issue but if I restart my devices afterwards I get back into the same state where by the production gateway is stealing the device from my test gateway.

Please let me know if there is any other info I can provide. FWIW, I will be able to replicate this situation in my test lab next week so I can gather log files as required.

[gerfen]

I've got some more data... One of my class A devices reconnected after about 20 minutes but I'm seeing an invalid frame counter error in the log file which appears to preventing it from receiving C2D messages:

2019-09-12 16:59:57.641 0004A30B00240C03: device in cache
2019-09-12 16:59:57.641 0004A30B00240C03: device in cache
2019-09-12 16:59:57.642 0004A30B00240C03: converted 16bit FCnt 6 to 32bit FCnt 6
2019-09-12 16:59:57.644 0004A30B00240C03: valid frame counter, msg: 6 server: 5
2019-09-12 16:59:57.644 0004A30B00240C03: decoding with: http://decoder/api/DecodeStregaSmartValve port: 4
2019-09-12 16:59:57.651 0004A30B00240C03: sending message {"time":null,"tmms":0,"tmst":421448932,"freq":903.5,"chan":6,"rfch":1,"stat":1,"modu":"LORA","datr":"SF10BW125","codr":"4/5","rssi":-95,"lsnr":-15.0,"size":18,"data":{"battery":3616,"enclosureIsOpen":false,"valveIsOpen":false},"port":4,"fcnt":6,"rawdata":"MzYxNjA=","eui":"0004A30B00240C03","gatewayid":"evolution-gateway","edgets":1568307597639} to hub
2019-09-12 16:59:57.676 0004A30B00240C03: message '{"battery":3616,"enclosureIsOpen":false,"valveIsOpen":false}' sent to hub
2019-09-12 16:59:57.676 0004A30B00240C03: checking cloud to device message for 00:00:00.5626434
2019-09-12 16:59:58.239 0004A30B00240C03: done checking cloud to device message, found no message
2019-09-12 16:59:58.240 0004A30B00240C03: processing time: 00:00:00.6008252
2019-09-12 16:59:58.240 0004A30B00240C03: converted 16bit FCnt 6 to 32bit FCnt 6
2019-09-12 16:59:58.240 0004A30B00240C03: invalid frame counter, message ignored, msg: 6 server: 6
2019-09-12 16:59:58.240 0004A30B00240C03: processing time: 00:00:00.6011640
2019-09-12 16:59:58.503 Physical dataUp {"rxpk":[{"tmst":422318324,"chan":2,"rfch":0,"freq":902.700000,"stat":1,"modu":"LORA","datr":"SF10BW125","codr":"4/5","lsnr":8.8,"rssi":-89,"size":24,"data":"QC+TkwKAlq0GCqjeKOoyrCxu17ZAN6g7"}]}

I'm not really sure how to resolve that issue?

[ronniesa]

Version 1.0.2 should solve this issue about the devnonce too. After you updgrade please restart redis.

[gerfen]

@ronniesa Thanks for the support on this issue. I meant to close it months ago. Closing now.