I must warn that my question is based on a one time observations.
So far it has only happened once. Though this was in a development setup and my question is in regard to production deployment.
Over the last week I have had a program running on my Edison module on my own hardware with blinking LEDs indicating the program is running.
This module is connected to a power supply with a switch that I daily turn on and off.
Usually the UART connecting the serial console like this:
Edison module <--> external wires <--> Edison Breakout Board <--> USB <--> Desktop computer.
Today I noticed for the first time that the LED wasn't blinking.
This time I had the USB disconnected like this:
Edison module <--> external wires <--> Edison Breakout Board <--> USB <-X-> Desktop computer.
I reconnected the USB cable between the Expansion board and my computer and opened the console.
What I then saw was the last line in the screenshot below: "boot >".
I don't know how it ended up in this state.
The power supply has been turned off since the last time I worked on the device so it can't be that I left it in the boot prompt and forgot that I did so.
My speculation is that since the UART/Serial console cables was hanging free in the air(laying out over my desktop) they might have generated a signal that was interpreted as a key press into the module thus aborting the boot process.
If this is a plausible explanation what precations should I take when making the final product?
By default there won't be any external cables connected to the UART, only the PCB wiring to the connector.
I'm thinking of connecting UART_RX(or UART_TX whichever is the input to the module) to GND on the connector to make sure there are not accidental signals that would abort the boot process.
Here are my updates.
I managed to repeat the "stuck at boot" issue a few more times.
It only happened when the Edison Breakout Board was connected to my card and disconnected from the computer.
So it appears some signal was sent back and misrepresented.
This only happened when the system was powered down for a long time(over night), draining our capacitors didn't seem to reduce this time.
Without the breakout board connected to the card with the Edison module there were no longer any boot issues.