Having issues over the last day or so with Zigbee devices becoming unresponsive and some are requiring me to add them back to the hub. Is this a known issue, the hub has the latest firmware - could that be where the issue may have come from? Seems really odd after such a long time of stability that this would happen. I don’t know of any changes that were made that would cause random Zigbee device drop offs.
Help!
When this (multiple zigbee devices acting wonky) has happened to me, independent of zigbee coordinator, the underlying factor has always been the same:
A single zigbee router (repeater) falls off the network. And by doing so, takes down the routers and end devices connected through it to the coordinator. When this has happened, I just reboot all my zigbee routers (by unpowering/re-powering them), and within 10 minutes, the mesh has recovered .....
If you want to "reset" your zigbee network, you can turn off your hub for 20m, and then turn it on again. (Shut off command, remove power at wall plug, wait 20m, plug in again). Zigbee devices will go into "Panic" mode and when the hub restarts your Zigbee devices will rebuild their neighbor tables and your problems may be resolved.
Thanks for the quick replies. I tried unplugging everything except battery powered Zigbee devices first and then I left the hub unplugged for an hour or so. Wonder if I need to break out the new hub, that seems a bit silly. How do you get in contact with support - no longer an option?
Tagging @support.
Also, you can visit http://support.hubitat.com.
Nope - don't unplug the Zigbee devices, you need to leave them plugged in/batteries in.
Just turn off the hub 20m and turn on again, taht will do the "panic" mode for the devices. They need to be on and unable to connect for >15m for desired panic mode to start so they rebuild neighbor tables when you turn on the hub.
Thanks again for the quick feedback! Unplugging everything has worked in the past, maybe it was just luck. It’s been a couple years since I had issues like this and powering down everything including the hub worked before. Seems odd it is out of the blue - haven’t had any issues in a while. Wasn’t sure if it could be related to the latest firmware, as that has been the root cause of most of the issues.
That is highly unlikely because there were no changes to the Zigbee protocol in the recent releases. However, based on the symptoms you describe, it sounds like a repeater may be going bad in your mesh, causing devices that route through it to act up. Also, you may experience severe interference caused by some changes at your location (new Bluetooth devices, new Wi-Fi devices, new router, etc).
Try it the way I suggest, and post back after the hub boots up and sits for a bit. 
If it doesn't help then @bobbyD 's suggestion of a bad repeater may seem more likely.
I’ll leave all of the powered Zigbee devices off and slowly add them back to try and figure it out. Does it make sense it can be any Zigbee device, and not the same ones that are having issues? Just noticed a ZWave bulb not responding. I know it most likely isn’t related, just found it odd.
To be clear, the only recent network/Bluetooth changes are a Meraki and a Forrinet device I’m testing - no wireless, only wired. They are on their own subnet.
Please list your Zigbee devices (make and model). Often, the community has experience with specific devices known to cause issues with a mesh network.
Powered up the hub, now it is getting tons of errors from the Zigbee/ZWave outlets. All of them are having intermittent issues. They are the Iris 3210-L. Error message from the logs for the 3210-L all show “DID NOT PARSE MESSAGE for description : catchall: …”
Random bulbs are also non responsive.
Seems odd that all of the same model device would act up at once? These have been paired with the hub since the beginning.
Yea, the Iris 3210-L outlets are very reliable Zigbee devices. Did you also pair their built-in separate Z-Wave Repeater? I wonder if one of them is going bad, causing issues on both the Zigbee and Z-Wave mesh networks? Just a guess.
I have about eight Iris 3210-L outlets paired to my C3 Hubitat hub, along with many Iris 3326-L Motion, 3320-L Contact, and 3315-L Leak sensors. These have been very stable and reliable on Zigbee channel 20 for my network.
Are you running a 2.4GHz WiFi network? Have you configured your WiFi Router/Access Points to only use WiFi channels 1, 6, and/or 11 with a 20 MHz frequency width? These WiFi channels will not interfere with Zigbee channels 15, 20, and 25. If your router is AUTO assigning the 2.4 GHz WiFi channel, it may have changed recently causing a conflict, resulting in the zigbee issues you're experiencing. Again, just another hypothesis to investigate.
Which Zigbee bulbs are you using? Only the Sengled Zigbee bulbs are recommended to be directly paired to the Hubitat hub alongside other Zigbee HA 1.2/3.0 devices. Other Zigbee bulbs are often the source of Zigbee mesh network issues as they are poor repeaters for ZHA. Sengled bulbs are NOT Zigbee repeaters, and thus they do not negatively impact a mesh network.
Yes, all of the Iris stuff has been solid on every hub they have been paired to. ZWave is also paired. I don’t use 2.4Ghz, so no issues there. Sengled bulbs are paired to the hub, other non-repeaters are paired to the hue and lightify hubs. Now the alert in the corner shows “Some devices are generating excessive events” - when does that reset - the device isn’t plugged in?
I’ve noticed when devices become non-responsive, they still show on or off with the switch - there is no change now.
You can easily identify what device is spamming your hub by checking the Logs. What device is not plugged? Unplugging Zigbee repeaters is the surest way to bring a mesh to its knees, if you don't remove them properly from the mesh first.
I just removed it now for the first time due to the excessive events. They are all having the same error over and over - how can you isolate it to just one?
Are they all unplugged? You didn't have to remove it, should have just plugged it back in. At this point, you may have deepened the mesh problems you had started with.
No, I unplugged the one device and plugged it back in. I didn’t get to the point of unplugging all of the devices, just the one. These are the same errors I was getting earlier, doesn’t seem the one device had any impact. It is back on and works fine for the moment.There is no difference in the way anything works, so it seems it may have not impacted the mesh. How do you troubleshoot when the Zigbee network is offline? It also went offline yesterday but a reboot seemed to fix it. There was nothing abnormal in the logs. One other thing to mention is that usually when a device doesn’t work, the hub generally will still change to show on or off, based on the switch status, it does not change now when having these issues. I can reset and re-add the devices, they show as already added - they have the right info on the device page - color temp and on/off for example - there is no control from the hub, Some Zigbee devices work as usual and haven’t had an issue.
Let the mesh settle for an hour or two. Start looking at the logs. Is anything power reporting excessively or retiring out (just a particular device). When you decide to remove a device (just do 1 ) do the power off for 25 mins thing to throw your mesh into panic mode. Power back up hub and wait a few hours to have things settle. Rinse and repeat. Concentrate on mains powered devices. Battery devices don't repeat.
Thanks, that pretty much describes what I have done. I’ve been focusing on the powered repeater devices only, although there are quite a few devices that have been non-responsive - looks like they are all non-repeaters and battery powered, so that’s ok. There was an Iris plug that was generating a large amount of messages in the logs, but it still worked fine - unplugged for an hour or so then I removed and re-paired and it seems fine. Devices seem to only work today when discovered and re-added - they do show that they have already been paired and work for varying amounts of time. I left the hub powered down several times for 30-60 minutes, let it sit powered up overnight but no real impact that I can tell. The Zigbee side of the Iris plugs are probably the core of the issue - all are still sending the warnings but seem to work while they have that error. ZWave side works fine. I don’t see any other errors in the logs - it doesn’t seem possible that all of the plugs would go bad at once. This network has been rock solid for a long time.
Download the Hubitat app
