If the problem comes back I would try disabling those Tile Master apps temporarily.
After a night’s sleep, hub is still working ok.
Last night I upgraded Hub Info to v3 driver, and paused one of the TM apps.
Other thing I did yesterday after last restart (around 3 pm) when I arrived home (around 6 pm) was fixing all zigbee ghosts.
I was able to pinpoint a “null” device, a Sonoff Zigbee USB Dongle-P, which for some unknown reason lost registration on the HE zigbee devices table, so I reconfigured it, and unplugged it for a couple of hours so the null device disappeared, and plugged it again.
I haven’t done anything important that I can think of, other than what I already mentioned, and now I’ll just keep waiting and watching what goes on.
Now I’m getting ready for some activity ( ) and will be back after lunch.
So far so good, after more than 30 hours the hub hasn’t failed.
I will post if anything weird happens within the next hours or days.
Thanks for your support.
Unfortunately I woke up this morning finding a frozen hub.
Important to note that I forgot that I reboot the hub every day at 3:00 pm, and now I have to review the log to see if it froze before 3:00 am reboot, or after that reboot, when it was fresh from a recent reboot.
The only thing I can remember doing last night before quitting the night was
-
Reinstall Hub Info v3, and then activating it
-
Reinstall the Energy Cost Calculator app
I just deactivated a few devices I have configured but not in use.
I’m thinking that maybe I will remove from Tuya app some devices that are configured but not in use (8 out of 16), and have that Tuya Cloud app to be not as demanding as it is now.
Will monitor the hub throughout the day.
Edit: I reboot every day at 3:00 AM!!!
Instead of removing the Tuya or other apps, just DISABLE them to troubleshoot. If you find it is the problem you can always remove it later.
I would start with disabling the Tuya cloud parent, and also the other TileMaster app. See if that make a difference.
Will do that.
I was thinking about reducing the load on the Tuya Cloud app to half, polling only 8 devices instead of 16, with fewer devices being polled every time, but yes it is a PITA because now I have them ready to be installed when needed, and if I remove them from Tuya app, I will have to go through all the configuration process when installed.
Will report later
I just came back home after being away all day.
I was verifying activity remotely and had to power cycle the hub twice while away.
Before leaving home this morning I deactivated two Tile Master tiles and Tuya Cloud app, easing the load on the hub, but it seems something else is causing hub’s freezing.
Now I suspect that what is causing the problem is either Hub Info v3 OR Energy Cost Calculator.
Here is current status:
What would be the next step trying to pinpoint the situation?
@0.5% you'd be barking up the wrong tree...
Yeah everything is below 1% of total now, I think we ruled out an app or device causing it for now.
Is the hub actually locking up totally or same as before where the dashboards are inaccessible but you can log into the hub locally?
Before power cycling the hub I verified clients list on Unifi UI, and the hub IP was gone, not accessible via IP address.
My comment about my suspicions on those two apps is because those are the only new/different things I installed since yesterday, before the two lockups, although I don’t see any other comments supporting my suspicions.
I just got my C-7 migrated to the new C-8 and I've noticed that I'm unable to get Google Home and IFTTT re-synced and getting cloud errors.
-
Google Home error: (The parameter "state" must be set in the query string.)
-
IFTTT error: (Could not connect service.)
Based on this thread, seems a bunch of Cloud issues are going on.
Since google is tied to the hardware ID you'll likely have to re do it
This morning hub is still “awake”, working normally.
Last night I reactivated Tuya IoT platform driver, and this morning I noticed a specific device linked to this driver, a Siren which I linked to Tuya, would only throw Temp and Humidity, without any control of the other siren functions such as turning on an alarm or something like that, but I realized it would throw those two values at a stupid high rate, so I removed it from Tuya app, thus also disappeared fro Tuya IoT Platform (Cloud), and so far, activity was reduced drastically.
Anyway I was using that siren only as a temperature sensor which is no big deal.
So, going back to the specific problem I’m having, I already had two lockups without any excessive activity from a device or app.
Now I’m lost here.
Is there some kind of tool that reviews and correct any error on the data base? In case any is found?
Maybe the problem has to do with a corrupt data base.
Edit: I guess activity on “TM - Temps en casa 2” will also be reduced because that device, Sirena, was used on that app. We’ll see.
Edit 2: I’m rebooting hub to start fresh on log info at this moment.
I'm no expert, but if you have a database backup, and you should since it's automatic daily to your hub, you could just do a soft reset and restore the database. Those remote 'power cycles' could be disruptive.
You didnt answer my question
Totally unrelated to this thread. For Google you need to go into the Google Home app and do a device discovery, then possibly log into hubitat and select the new hub. IFTTT is probably a similar process on the IFTTT app.
Yes I did!
“ Before power cycling the hub I verified clients list on Unifi UI, and the hub IP was gone, not accessible via IP address.”
The hub was totally frozen, and that is why I Power cycled it
Sorry I missed that part. So just a random thought, maybe try turning on this setting in Settings > Networks Settings. In the off chance the hub is alive but disconnected, this will force it to reconnect if the ethernet drops for some reason. It just sends out a ping at the interval selected and will restart the ethernet if the ping fails.
I just checked my hub and it was OFF!!!
Just Power cycled it again, and went into this option. It´s already set on 60s.
Is this the way it should be?