This has happened twice in the last few days, monitor goes offline, I go out to the panel, connect with the app and force it to reconnect to wifi and then it’s back online. Both times it has gone offline at exactly 3am.
None of the other ~30 devices on my network are having any problems. Sense support has not been helpful so far. Anyone seen this?
When they push firmware updates it’s usually at 3 am
Maybe you are interrupting the update so it’s tryin again the following day
Well that sure seems like the culprit, however I’m certainly not doing anything at 3am to interrupt any update. So something must be going on with my unit and this update trying to install.
I wonder if you forcing to reconnect is somehow interrupting the update. I really don’t know though and it’s a guess.
Support didn’t help?
Well, when I force it to reconnect it’s been offline for 4-5 hours already, so I don’t think a firmware update is running for 4+ hours and then being interrupted by me.
I have a ticket open again and someone is looking in to it, thanks for the tip on the firmware update time that will help my case a lot. 
If anyone else runs into this… The problem is that sense must have decided to hard code their own DNS into the firmware (1.1.1.1 and 8.8.8.8) if the monitor isn’t able to reach either of those IPs on UDP port 53 the firmware update will fail, revert back to the old firmware and fail to reconnect to the internet.
On my network I have UDP port 53 blocked and force all devices to use a local DNS server so I can track what they’re doing.
The fix was to temporarily allow the sense monitor to get out to the internet on udp port 53. This is the first IOT device I’ve run into that does not honor my local DNS over whatever is hard coded, so I hope sense is going to fix this.
There are several posts on this already.
Here is the response from sense
Same problem here. In my network port 53 was always disabled, however Sense never had a problem using local DNS. The “3AM EST issue” started on April 30th of 2019.
I’m attaching a graph on when the requests to 1.1.1.1, 8.8.8.8 and 208.67.222.222 were initiated by my sense monitor.
stats from the last 30 days:
I have allowed the above 3 IPs, hoping the the firmware upgrade will complete tonight.
I don’t think its a good idea to use hardcoded DNS servers…
and one more:
it’s from the last 24 hours (my monitor is often offline from 3AM till 11AM each day, starting April 30th):
They have removed all but 1.1.1.1
You have to contact support if you want it removed also.
Please see these links in reference to the DNS issues mentioned above.
