I've been using easyiot for a while now but have always been experiencing stability issues. It crashes for no reason and with nothing in the log.
Sometimes it can run for a week and sometimes just for a day. The process just seems to die.
I'm running 0.9b on an Rpi 2 (image downloaded from here.
I have a about 15 - 20 sensors and a few automation scripts. Nothing exceptional just some trigger scripts and one cron script with an hourly task.
pi@raspberrypi ~/easyiot $ sudo /etc/init.d/easyiot stop
start-stop-daemon: warning: failed to kill 2101: No such process
No process in pidfile '/tmp/easyiot.pid' found running; none killed.
kill: usage: kill [-s sigspec | -n signum | -sigspec] pid | jobspec ... or kill -l [sigspec]
[ ok ] easyiot stopped.
Have you tried re-flashing the pi SD card with the pi2 server? It might be corrupt. If it is not that then I have to say that the power supply is most likely the culprit here and it could easily be the cause of corruption on the SD card.
A good quality power supply is not cheap, but it never hurts to use them.
I have some really bad Chinese ones here, some are even dangerous lacking safety components and some are just repackaged supplies from God knows what equipment.
Get a decent power supply and you will not regret it, money well spent.
One thing you can try is to add a nice big electrolytic capacitor (say 1000uF 10V or above). Connect it across the 5V and GND on the pi. Make sure to get it the right way around, they only work one way (and the wrong way they can explode!). It might help until you can replace the supply.
Just a thought, if the problem would be the power supply or the SD card, shouldn't the whole Raspberry PI be unstable then?
In my case only the easyiot process stops.
I have more SD cards and also branded power supplies which I will try with. But I'm using the same kind of power supply and SD card in another PI (not running easyiot) without a single problem for over a year.
I have also enabled debug in mono to see if I can catch anything. Is there any trace logging that could be interesting to enable? I would like to see what exception is thrown or anything that can show what's causing the process to just die.
I am just another user trying to help out. Maybe you could try temporarily disabling the cron jobs and scripts and see if that makes any difference?
The randomness might still be a power issue. One way to make the pi cheap is to put in the minimum power filtering and expect a quality stable psu to be running it. Those cheap ones will be spooked by mains spikes, static discharges and even demand changes from the pi itself as it performs different tasks.
Do you have a volt meter to measure the voltage on the pi?
May be MrEioT will see something here that I can't and chime in...