I did not have the watchdog enabled at the time, but it would not have triggered. I enabled it later it did not trigger when the failure re-appeared.
The fault behaviour looks like this: The system keeps running, but the filesystem is no longer really accessible. I had an ssh session connected (and it did not disconnect) but on ‘ls’ or ‘sudo’ it can no longer find /bin/ls or /usr/bin/sudo. The watchdog does not trigger though. I guess I can configure it to test the filesystem as well. Will look into that next.
In the mean time I’m trying to find out if the problem was caused by a bad SD card or if the temperature was just too high. (it is outside in the sun in a metal enclosure)
anyway, reliable shutoff is pretty important under failure conditions.