I was wondering if overstressing the system could be a reason, so my questions are - how many devices are you controlling with your ST hub and how many schedules/routines or typically webCore pistons do you have? At what point does the number of devices/pistons etc become an issue?
I have approximately 80 physical devices on my network (of which 41 are Hue bulbs) and about 15 virtual devices. I try to limit myself to webCore for all my automation and I currently have about 40 pistons.
There are definitely community members who have several hundred devices, including over 150 physical devices. @johnconstantelo might have more to add.
@bamarayne also has well over 100 physical devices, and I don’t think he’s been seeing any particular problems lately.
Apparently support has recently been telling people who have more than 40 Z wave devices not to run a zwave repair if they’re having network problems. That seems weird to me, as I’ve expressed, but it is what it is. A zwave network has a hard stop at 232 devices because of the addressing scheme, but I wouldn’t expect to see specific problems below about 150 Zwave devices other than system utilities taking a long time to run. But of course SmartThings is a cloud-based platform, not pure zwave, so there are a lot of idiosyncrasies.
Separately, I know you’re already active in the thread that I’m about to link to, but for anyone who finds your thread first, there have been multiple reports this month of people who are seeing lots of different weird errors. It’s not a huge number of people, but it is more than a dozen at this point, so I don’t think it’s just individual hub failures or mesh strength issues, although it could be. Particularly since all of these reports are from people who had set ups that were working just fine and then suddenly multiple things started failing.
But for those who are having issues with multiple devices going off-line and multiple automations failing to fire with problems that just started in June 2017, you can start with the following discussion thread:
There’s a limit to zigbee, but it’s in the thousands. You can hit it if you’re doing lighting for an office building, or a sensor net for a large commercial farm or an oil field, but otherwise you don’t usually run into it.
1 - The SmartThings app is too slow, so slow it is useless to me on a regular basis. For configuring something it is fine. Getting to “things” to turn stuff on / off, check status it is slow. The new dashboard is useless.
2 - ST is missing schedules on a daily or every other day basis.
One known factor: if you have a single piston that has many devices, it may time out. Sometimes it’s better to break up devices into multiple Pistons just to stay under the timeout threshold. The folks in the core peer Assistance threads can say more about that.
I have a few pistons (controlling my Hue Lights), which control up to 20 lights. These are simple pistons turning lights on and off. These are however running reliably. The pistons which fail on a regular basis, are the ones controlling single devices like water heaters and wall heaters. My pistons are looping until it has confirmation that the device has turned on/off, which appears to be making things worse if they start going wrong. I have however been unable to determine what is going wrong first. Device communication or piston "overload? due to the action repeating all the time.
This business is frustrating. My setup has been working great for the past 24 hours, but now it has started acting up again. (Without any devices having been added or removed). Devices within 8m from the hub, can’t be reached by the hub (I turned device health back on again when I realised that many devices have stopped responding, to see if they are online). Strangely, it seems to be the same type of devices that go offline. Today the 2 Aeon motion sensors are “offline”. One is within 4m line of site of the Hub and the other downstairs - never had any issues with them. Most of the Fibaro Single Switch 2 are also not visible.
What do you guys get when you do a Z-Wave repair? Does it reach a stage where it reports anything? I’ve tried running it many times since my system started acting up, but I don’t get any messages. I left it for more than half an hour, but all I see is this: