I have a system of 10 nodemcu(s), connected to each other via painlessmesh.
One of them is designated to
- disconnect from the mesh
- connect with a given wifi AP
- send all the collected data via a POST request
- connect with the mesh again to keep collecting data
Firstly, any advice on this design?
Secondly, my problem is that although this system works well (everything from sharing data on the mesh to server requests) for a few hours, later on the nodes start to drop and once they do, they never register on the mesh again unless I manually power cycle them one by one.
So, I'd like to record any errors/exceptions that might be leading to this behaviour.
Is restarting the nodes at random times in the code going to help? They don't have to maintain any state variables for long-term working of the system.
Thanks and Regards,