bug: Ping service: Is creating pings with same port

[patrickelectric]

Bug description

Steps to reproduce

Connect two pings and a ping360

Primary pain point(s)

The UI does not make sense and it's not possible to know which sensor is being used

Prerequisites

• I have checked to make sure that a similar request has not already been filed or fixed.

[Williangalvani]

that is interesting. I wonder if bridges is stopping and that is causing the port to be identified as unused.

Can you share the log file?

[patrickelectric]

It was a report from manufacture, they are going to get in touch

[ES-Alexander]

Also raised in this forum post - I've asked for a log.

Since it's async, I suspect the driver is likely starting to detect the next device before the previous device is counted as registered in the OS. @Williangalvani not sure if you've got some idea how to fix/avoid that.

[ES-Alexander]

This has been raised again, but I've also realised a possible cause, and thought of some ideas for potentially fixing it:

... my suspicion is that the asynchronous port searching means both device connections find the same UDP port before either of them is fully connected to it (in a race condition), and then they end up sharing the connection.

That may be preventable with a well-placed sleep, or perhaps by making the port connection process raise an error if multiple connections are made to the same port, and/or treatable by adding duplicate-port detection to the ping service and making it restart relevant device connections when port sharing is detected.

[ES-Alexander]

It may also be worth just storing a set of known used ports as a blacklist, since I imagine Python's variables will update faster than a port connection does. It's more important to have unique ports than to always have the same / consistent ports, so if we use a blacklist + port connection checking then that may be sufficient.