I have setup IFTTT to monitor the online status of a couple of cores in my home and log to a spreadsheet.
I have matching “applets” for each core, one for online and one for offline detection. I wanted to check the reliability of my broadband over time, plus how reliable one core is (it’s on a homemade circuit which might have unreliable power)
I would expect to see offline/online detection in pairs, e.g. a network interruption causes an temporary offline glitch.
That is not what I see.
I frequently get “online” notifications with no matching “offline” notification.
Is the status detection just inherently not reliable?
There is no real offline detection, it’s just a missing online signal which indicates a probable offline status, but for that the device has to be offline long enough to have the cloud “miss” the device.
The description of the trigger says “… when the device changes state”.
So how should I interpret multiple “online” entries in my spreadsheet for one particular core?
Is it that when the core was checked, it was online, but at some (unknown) time since the previous time it was checked, IFTTT knows that it had gone offline and recovered?
Based on my experience, I get an “online” notification from IFTTT shortly after the device connects.
As ScruffR pointed out, the cloud doesn’t get an actual “offline” message since the device is in-fact offline and can’t communicate.
So you wont know exactly when the device went offline, but you will know approximately when it came online.
You could publish a value to ThingSpeak (or other) every minute, then check that log to know the exact “gaps” in connectivity.
I’ve recently started using do this and found it to be unreliable as well. I have mind set to send an email and keep getting pairs of offline/online for no obvious reason, sent in the same minute. I wish the trigger could be offline for X minutes to make it more reliable as well as to ignore any manual resets.