I like the concept of Device Health, but I continue to have false offline notifications in the mobile app for the same 2 zwave switches (using the stock ST DTH, not mine) that always work via the mobile app.
In the beginning, Device Health seemed to work just fine, but lately it’s reporting more offline devices that aren’t really offline. ST support hasn’t been helpful, and only offer the typical “delete device and start over” answer; which I did do, but it did not help at all.
Yes, I have it OFF on my remote site (in case it causes something bad & I’m not there to fix it) and ON at the other site where I am – just too see what happens. So far, it hasn’t done much for me…
(But I also run Simple Device Viewer and it tells me so much more about my devices.)
There are definitely some people who have found that turning device health off resolves some message transmission problems.
I have wondered if a device is being reported as off-line if the system will still try to use it as a repeater. I would think that it would, but some of the problems that people report being fixed just by turning device health off do have that kind of symptom.
So I don’t know exactly what happens differently in the platform when a device is marked as off-line, but there do seem to be some ripple effects.
It would be nice is SmartThings development explained what criteria is used to determine if a device is offline.
It is possible that device health is buggy.
Or it is possible that device health is finding an underlying marginal comms issue
Right now I am going by the assumption that it is correctly identifying something, like maybe communication to the one device that I have this is being marked as unavailable has some kind of marginal communication problems.
In my case toggling the device on and off in the iPhone app temporarily “repairs” it.
Suppose device health was actually detecting some kind of underlying communication issue with the device(s)?
So in the case you mention, it sounds like this “bad” device was the only route to one or more other devices.
But if comms to that device is in fact marginal than using it as a repeater is problematic.
It seems that since this feature was released less of my devices are being marked as unavailable, in the last few days only 1 specific device gets marked unavailable.
This could be explained if they have tweaked device health by relaxing tolerances for dropped messages or something similar.
In a mesh network, it’s fine for some devices to be unavailable from time to time. Messages should just then choose an alternate route, of course assuming that an alternate route is available. Mesh is just very different then continuous connection topologies.
In essence it’s like reporting that a phone number which is busy is out of order, when really those are two very different states.
I don’t know the exact algorithms being used for device health, but judging from what is being reported, in some cases it’s too quick to jump from “didn’t answer” to “off-line.”
Yes, I had to turn it off. Device Health was showing a couple of dozen z-wave switches and outlets as offline when they were not. Turning Device Health off resolved all of those.
Tried it twice, once soon after launch and gave lots of false alarms, tried again recently and sadly no different. Far more reliable to rely on the great Simple Device Viewer smartapp or other device monitoring apps out there.
That’s my exact situation, and frustrating when those zwave devices say offline but still work! I like Device Health for my zigbee devices because it’s very accurately reporting those offline when that happens, which has been rare lately.
