Tempest and hub drops offline shortly after midnight

Posted here Hub repeatedly drops offline yesterday but afraid it may have gotten lost with the rest of the content of that thread.

My unit drops offline at the same time every night for 4 days consecutive now. Has anything changed software wise to do this. It seems too consistent to be a physical malfunction.

On resetting (by removing power) this morning - i was notified my tempest was offline. Nothing has changed in my setup since install three+ months ago. Within a minute the unit was connected and communicating though with no intervention from me.

Couple of thoughts:

  1. can you check any logs in your router to see if the WiFi network is getting reset at 01:10 each morning
  2. if the hub is connecting to a bluetooth device (giving a blue status light) without a network connection that could explain the dropouts (although I wouldn’t know why it was happening so regularly). Perhaps try and identify which device it is connecting too

A long time ago my hub would disconnect like clockwork every 14 days. This was because my router went through a reboot process every two weeks and the hub would struggle to reconnect. It is now a lot better, but you might be suffering from something similar.

Thanks @peter - my router is a “rental” from my local ISP and I have no ability to view or modify the settings on it :frowning: - so not sure if I can get to any form of log without contacting them directly and going through customer support… a possibility I’m sure, but maybe there are other diagnosis or tests I can do first before attempting that.
The hub does have a blue light every time its “locked up” or offline… any suggestions on how I could identify what it may be trying to connect to? I have so many bluetooth enabled devices these days in a house/family of 4 with teens (Apple Watches, Fitbits, phones, laptops, headphones, etc) wouldn’t even know where to start (would a BLE scanner on an iphone give me any guidance?), although I can say there haven’t been any new devices in the past week. Would resetting the radio on the hub be worthwhile?

1 Like

I pushed a newer firmware to your hub. This might help a bit … though we can’t be sure.
Scanning for devices is nice but not to sure it will tell you which one is locking in the hub … but it might help track.

Best is to scan next to the hub and already see who is in range.

1 Like

thanks @eric - will monitor and see what the morning brings.


I feel you! That’s the same situation I found myself in as well. An ISP router stuck in a 14 day reboot cycle :unamused:. Fortunately they have now sent a new router and this one has been up for 18 days and no reboot.

I may be teaching you to suck eggs, but have you tried entering the IP address of your router in a web browser to see what you can see? You might find that there is some landing page/status that could help with your troubleshooting. This might help you find the IP address: https://www.howtogeek.com/233952/how-to-find-your-routers-ip-address-on-any-computer-smartphone-or-tablet/

Good call… alas… any useful information is behind a pw protected landing page… silly ISP - what do they think i’m gonna do… unleash SkyNet?

1 Like

If you look up the manual for that brand of router it might have a default password. Or if you try username:admin passwrd:admin is often a default.

Thanks but nope :), not the default. Its even a ISP custom landing page complete with corporate logo.
But to be fair, I have the unusual circumstance that my provider is a fully private local exchange with an burgeoning fiber backbone in a rural region. I cant complain too much that they know their gear and manage it properly. :wink:

Hoping the firmware will do the trick. Seems it has already forced a change to my RH sensor CI cal (i havent hit 100% rh in a while and although it may be close its now highwr than my accurite sensor nearby… had been biased low). Will know in a couple hours.

Did not go offline last night! Woohoo.

Now to monitor the RH as it pegged at 100% early last night and has remained that way. Its consistent with nearby measurements so likely a CI check was passed along with the new firmware?