This one:
Story:
I have it on a Pie… Sitting there turning on/off as needed keeping cool.
I had about 19 weeks uptime clocked on it.
Got home, the Pi was dead.
Reading the docs for the device, pressing the button for about 5 seconds will reboot it - if it has been shutdown via this device.
Leaned in, pressed the button for about 5 seconds.
The fan spun up so all good.
(It’s headless BTW)
Waited, waited and WAITED for it to be back online.
NOTHING!
Pressed the button again for 5 seconds and waited for it to shut down.
Pulled the plug, plugged it back in and it whirred back to life.
Shortly after it had booted.
Hmmmmm…
Something’s going on.
Now, let’s work out what happened - yes?
I hear this can shut the pie down if it overheats - which MAY have happened.
I DON’T KNOW
(Sorry, I JUST don’t know.)
I have a program that checks the health of the pi - voltage, throttle, temperature, etc.
Anything from them and it is logged.
Nothing in the logs.
Hmmmmmm…
Ok, I found a bug in my bit of code that if it detects overheating and it persists for … a while, it shuts down the pi.
But the shutdown command was too quick and that it had been shutdown for that reason would be lots.
But that is if MY code detects an overheat.
(Luckily the condition is broadcast and other machines would know. They too didn’t get any such message.)
So I’m stuck to what EXACTLY happened.
Alas it has now been up for 21 hours.
Any logs would be long gone.
Any suggestions to understanding better how this device shuts down the machine if it overheats?
If it uses the same (or even different) command I use.
Both commands would (you would expect) they BOTH say it is overheating with the same thresholds.
And so although it (the hat detected the overheating - why didn’t my bit of code also?)
(semi rhetorical)
Thanks.