I am automating things with Node RED, to save some power draw from the 12v battery, like location of car, WiFi connection, etc.
When the flow decides to power off the LTE modem, Server v3 connection drops and is re-established.
Why? If WiFi is connected and set as the main provider?
Module started presenting a weird behavior suddenly.
Ping would fluctuate on and off, demonstrating something alike a boot loop or network service boot loop. Removed the module from the vehicle and did a factory reset (actual already did several and flashed the latest edge to both OTA partitions).
I can repro every single time, that after adding Server v3 configurations and starting the service I lose access to the module...
Ping goes back again to the same behavior (replies, time out, and replies again) and in my Cloud MQTT service I can see a constant set of messages of server.v3.connect...
Factory reset, configuring everything (WiFi, Server v2, etc.) works. As soon as I configure and start MQTT V3 service, hell breaks lose.
Is the module ready to be sent to the bin (not even 8 months...)?
The MQTT client (server v3) is still considered experimental.
There's an issue of the client sending many metrics with a high frequency, especially with vehicles that provide a lot of detail. See github issues.
I assume that induces the packet dropping effect you see. MQTT is a very inefficient protocol, so may overload LwIP if a lot of messages need to be sent.
Regarding the network reconnect scheme, that also seems to be a server v3 specific behaviour, I cannot reproduce the effect with server v2. You could try log level debug or verbose to find more detail.