60s is fine, you could probably go to 300s.
When you say off, the LED on the hub was not lit up at all, or was it red?
I can´t tell. It was just off the LAN. I couldn´t reach it through IP address.
Next time (if happens), when I detect a frozen hub, will call home to ask anyone to tell me the color of the LED.
When I have noticed it at home, the LED is never off, it´s always on, but unresponsive.
The color it was I can´t tell because I am daltonic!, but wife or someone will tell me.
I had done that initially and it worked but a few minutes had passed and it stopped working. went through the steps again and got the error.
The spot you marked is an app, and a tasmota device (not zigbee).
What is app 6?
Also, device 5407 below is having some sort of a problem. Is it saying there was 2513 consecutive events of not seeing the device? in how long?
Yes I was not clear.
I was referring to:
dev:127117 is a Zigbee USB repeater, one of those small USB plugs that work as repeaters
app:6 is Hubitat Dashboard
dev:1260 is a three gang wall switch flashed with Tasmota
dev:5407 is a Zigbee Motion Sensor
So the crash actually happened right after the error from the dashboard. Looks like maybe the device named "alarms" you have as an attribute tile, but the text in that attribute it too large for the dashboard. It is repeating every minute it looks like. That would be something worth chasing down and fixing, or just remove that tile from the dashboard.
If you think its that zigbee repeater all you could do is unplug it and see what happens.
I haven´t done what you suggest, but I just noticed Cloud access just got blocked.
I was able to access the hub remotely via IP address, and I tried to open Dasboards app, going to a dashboard and it wouldn´t open any way, neither Cloud dash nor local dash.
Is it possible to "reset" an app? and maybe see if it works again before it gets frozen?
No not really, only if you reboot the hub.
The Dashboard dying is probably a symptom of whatever else is going on behind the scenes with whatever crazy mix of custom apps and drivers you have going on, throwing errors constantly.
Yes, I agree.
What is weird is that it started only a couple of weeks ago, without anything special in-between.
Any ideas?
Can you help me on this one? I didn´t understand this particular comment,
"Looks like maybe the device named "alarms" you have as an attribute tile, but the text in that attribute it too large for the dashboard. It is repeating every minute it looks like."
I have a screenshot of last event, and it´s not related to Dashboard,
Thanks for your help Jeff
The top line there from 3804 look like a normal status message from a tasmota devices (can tell due to some logs below that).
The warning right below from device 3799 is more interesting, it is saying that device has been offline for 106953 minutes (74 DAYS!). Find what device that is, go to the device list and click the red "X" at the top to show the disable boxes if they are not showing, and disable that device by checking the box next to it. This will prevent the driver from executing at all, and stop the errors for now without having to remove the device (yet).
If you have any other similar errors or warning about unreachable devices, disable those also.
Many user made integrations do not play nice if the device goes offline, they may keep trying to reach the device constantly which will bog down the hub until things start crashing.
3804 is Parent app for a 4 Child Tasmota device using an old Tasmota driver from Marcus.
3799 is a Zigbee button device barely used. I just disabled it.
Will look for other devices in the same situation.
Curious. Is your network speed on your hub set for Fixed 100 or Auto?
Is set to Auto
I just woke up and found Cloud Dashboards not working again.
And after going through everything and confirming that everything else works ok, I decided to disable Hub Info Device driver, and Rebooted HE hub.
I know nobody else has complained about it, but there is nothing else I have done for my hub to start freezing but the installation of that particular device.
When/If the hub freezes again, I will report it and enable HID again.
Do you have any debug level logs on your unifi? At this point I'm still dubious it's the hub itself. If you can set it up use wire shark and mirror the port the hub is plugged into. When you find the time the hub locked to, check the wire shark logs and see what's being broadcast to that port
Can you give me some instructions?
Wire shark is fairly self explanatory when you set it up. Unsure on the unifi dream machine how to set it up but I know it can do that..
I see a lot of issues with udm and hubitat (and other devices). Usually it is because something is configured incorrectly by people who don't quite understand all that's involved in ip networking. That's not a slam against them at all mind you, just a fact. So something invariably gets borked because of an innocuous setting. Everything that is going on with your issue(s) screams jumbo frames or similar even though you have jumbo frames turned off. I really would like to see the wire shark logs if everything going in and out of the hubitat port. I'll also note I've been a network engineer and security consultant for over 35 years and I still miss stuff I have to track down and correct. Nature of the biz. Hard to spot everything at once
I believe some times when the cloud dashboards are going offline, the hub is still online, which is a new one. But yeah otherwise the case where the hub is totally offline and locked up is exactly what happens with jumbo frames, and it always seems to be this Unifi equipment. And the people always swear jumbo frames is off, and then as soon as they isolate the hub from all the other devices it works fine 
1 Like
On UniFi jumbo frames can be set globally or per switch, so make sure any switches either are following global off or are off on the switch. Also, UniFi has released a lot of updates over the past year, make sure all equipment is up to date.
1 Like
