Posting here as well (I also posted in the Hubitat forum).
I’m in this unfortunate group of people that is experiencing this problem.
Reading the entire thread, it’s still not clear what is happening or why. My blues have been working fine for almost a year, so why all of a sudden is the keep alive thing causing trouble? Was it an update to the driver (and if so, why wasn’t a new driver pushed) or it is a problem with Hubitat?
The latter seems more likely, as Hubitat has frequent updates whereas the driver updates for the blues are relatively infrequent, and FW updates for the blues even more so.
While I’m interested in knowing the root cause, right now I just want to get back up and running. I’ve already rebuilt my mesh three times, including resetting my zigbee radio, numerous hub reboots and hanging garlic all over the house to ward off evil spirits. I’m at the end of my rope.
Even the binding between the blues and my Hue downlights, which I though was rock-solid and would work as long as there is AC power, doesn’t work. This is why I went with the expensive combination of blues and Hue lights in the first place. WTH is going on here? On one hand it seems like a problem with the blues since direct binding doesn’t work, but the fact that every last one is dropping off the network kinda points to the hub. Some sort of perfect storm, where multiple issues are at play?
I responded in the Hubitat community, but to me the signs are that the problem is related to a Hubitat firmware update, but our team has been working on a workaround in case the Hubitat folks can’t figure it out.
For the binding issue I think it was mentioned, but when you reset the network the bindings need to be recreated unfortunately. They should continue to work even when the hub can’t control them.
You don’t have a way to use Z2M or ZHA to manually update the firmware do you? I hate to have you go through that process on so many devices, but I don’t think there is a manual way to update firmware on Hubitat. TBH, it has been so long since the problem was reported in the Hubitat community I was hoping they would have some kind of diagnostics, but I understand that we are a smaller company and priority probably isn’t in looking into an issue like this with a small footprint of users.
What I didn’t realize at the time was that only the zigbee part of the blues was not functional. The blues that are not in smart dimmer mode still worked fine (as far as controlling the load) so I expected the bindings on the ones in smart bulb mode to work as well.
I really hope that isn’t the case. It’s hard to read the Hubitat guys sometimes, but your products are well known and used by a lot of people (or so it seems), not some crap someone buys from Alibaba hoping it will work.
I air-gapped my 6 fan switches and my 45 or so 2-1s are still working fine for about 36 hours now.
As I mentioned over in the Hubitat forum, the only thing I did that I can recall is to update the fan switch driver to the latest version, maybe 4 days or so ago. That seems to roughly coincide with the beginning of my troubles.
Anything I can do to help track down the cause of this? Should I roll back the driver to the previous version?
Afraid not. I don’t even know what Z2M or ZHA stands for .
Curious if you power fan switches on if it goes away.
I loved one of my fan switches to the garage hub and it’s been 16 hours with no issues on that hub (2- 2N1, 1-fan sw, 1- outdoor motion). Not much on that hub though. Driver is latest as well. I can copy driver from original hub over to test older driver too.
I powered one fan switch on last evening to clear out the humidity in our bathroom after taking a shower. It was only on for about 20 minutes, then I air-gapped it again.
I do have another hub that I could pair the fan switches with. At least if that mesh went south it wouldn’t take the whole house down with it.
I don’t mind testing, but I’d like some direction (what to try, what log entries to look for, etc.). Anything besides random shots in the dark.
They didn’t, but I only had the fan switch powered up for about 20 minutes. I thought it had something to do with a keep-alive message that only happened after 12 hours.
What fixed FW are you referring to? FW on the 2-1’s or on the fan switches?
I’ve been running FW 2.15 on all my 2-1’s for several months with zero issues, and they seem to be running fine now. And the 6 fan switches seemed to work just fine with whatever FW was installed at the factory.
What seemed to trigger off the mass mesh dropoffs was updating the driver for the fan switches. I still don’t understand how the fan switches caused the 2-1’s to drop off the mesh. Far as I can tell, none of the Hue lights or anything else dropped off the mesh. Just all the Inovelli switches.
Correct. Depending on when you last airgapped the 2-1s, it could be 1-12 hours for the next cluster message.
I have 2.17 running on a few switches to test the new firmware to help correct the issue only a few of us have seen. We can’t pinpoint the issue. Is it Hubitat? Is it Inovelli? Or both? What I do know is we sniffed my zigbee network and do not see the hub respond to the switches when it broadcast the keep alive cluster request. There’s literally no reply.
I noticed my issue when I was updating the canopy module and it just stopped responding to commands. I thought it just needed a factory reset as there was new parameters defined. Unfortunately none of my switches were responding and led to my initial post on Hubitat community.
Since only a few of us have the issue, it’s hard to correlate the problem since there’s a wide mix of hardware. I’m solely running C7 hub. I have 99% of Inovelli devices. Others have C8 and a mix of other hardware.
Ok, but something is missing here. I am running about 45 2-1’s on 2.15. They’re sending out keep-alives, right? Everything has been running fine for close to 40 hours now, without any fan switches online. Mine have been running fine for close to a year now.
Strange as it is, my problem seems to have started when the fan switch driver was updated. But there’s no obvious connection between that and the keep-alive mechanism going dead on the hub. Or is there?
If it was me, I would turn in the fan switches and see if you run into disconnects in the next 12-24 hours.
The only change I know in 2.17 was the keep alive cluster is enabled/disabled depending on if the hub supports it. There’s probably other changes in the firmware, but I was trying to restore my network and aide in troubleshooting it.
I’ve had 2.15 running since July 23 in all my 2-1s. I’ve had the beta and released fan switches for about a year with no zigbee issues either. Literally never had an issue since September 22 until Jan 24. That includes all the beta firmware jumps and driver changes.
Edit: The plot thickens… Actually, I had only air-gapped 5 of the 6 fan switches, one has been on the whole time. And yet everything has been working fine.
This makes no sense whatsoever. The other 5 are now on.
Man, this is wild. I really hope we can nail down what the deal here is between Hubitat and the Blues… I have a feeling that would help nip potential future similar versions of this issue (with other flavors of the Blues) in the bud.
Well, almost 48 hours since enabling the fan switches, everything seems to be working fine. Same devices, same drivers, same Hub FW…
Only issue is that my C8-Pro has been having some very lengthy UI freezes starting sometime yesterday. Only cleared up with a reboot this morning. No evidence that it’s related to the zigbee issue, but disconcerting. My hub load between devices and apps is 1.1%, for whatever that’s worth. No reason why the UI ought to lock up for extended periods at a time.
Wish I had some info that was useful. I feel like the whole system is a ticking time bomb.
You mean, aside from hanging garlic around the house and the exorcism?
The last time I got them back online, I air-gapped two of them and when they came back online on their own, I just turned off the main to my house for about 15 seconds. It was easier to reset the clock on my stove and microwave then airgap 50 switches.
The hub is on a UPS so it wasn’t affected and I intentionally didn’t reboot it. The switches all rejoined the mesh on their own. This was two days ago, and things are still working ok. Makes no sense.
Now, I have to go and perform the required daily burnt offering.
.