Updating to 2.3.9 right now to see if this helps, thanks for mentioning this @gopher.ny,
I did end up implementing the setup described a few posts up from hubitrep, except instead of a notification I have Rule Machine make an API call to my network switch to simply turn the LAN port off, wait 30 seconds, and turn it back on. It's fine as a workaround but I am definitely looking forward to scrapping all of it to return to the hub just doing its thing reliably and invisibly!
Just updated to the latest firmware and it still doing it.
I setup a rule to ping and fetch the cloud url every 50 mins and notify me when it went down. It lost cloud connection but doesn't fail the ping or URL get. I went to test a ping and it can ping the cloud no problem. In fact it can ping anything and it still talks to echo.
The time is correct but if I hit sync to browser it reconnects to the cloud. Or if I reboot.
And as I said I have 2 hubs the other is having no problem. I also upgraded it to the latest version.
Its so strange is that the other hub just doesn't have this problem.
I've tracked this to the internet going down for a few seconds. This hub just wont reconnect to the cloud. But the other one doesn't have any problems.
I resolved an issue with cloud, Alexa skill authentication and webcore. This was resolved by rebooting my Orbi router. I had no problems with the 50 other devices on my network.
Stopping by to update that 2.3.9.135 did not eliminate the issue. Now that I have Rule Machine checking the inbound connection and automatically triggering a network reset (simulated cable disconnect, reconnect) it's less inconvenient. Mostly I just receive the notification that my scripts had to intervene.
For others who now seem to be experiencing similar problems (saw another thread as well). Rebooting routers, hubs, resetting the time, disconnect/reconnect of the network cable - none of these are fixes, they're just workarounds until it eventually happens again.
At the time I did mean 2.3.9.135. I updated the hub to 2.3.9.143 at the time of my reply, but it's too early to tell if the minor revision changes anything related to the inbound cloud proxy.
Last update 2.3.9.142 didn't fix it im updating to 2.3.9.143
How did you reset the network by rule. My test in rule never detects the cloud going down it can always see the cloud.
I've been on 2.3.9.143 for a few days now and it appears the cloud connection needed its reset earlier today - so it's not likely resolved yet.
As for the automated workaround - I implemented hubitrep's suggestion from earlier in this thread (May 1) although in addition to sending a notification I implemented a custom script on a different box to use my network's API to disable the LAN port, wait 30 seconds, and then re-enable. That part is pretty specific to me/my network setup but you could do what hubitrep outlined and perhaps add an action to reboot the hub. I didn't want to do a full reboot so I didn't look into that.
Still going down however sometimes after hrs it will now reconnect. But everythime it goes down you can reset the time from browser even tho its correct and it goes back online. Latest version as of 6/15
I’m running a C8 (up to date on firmware) and started running into this issue a couple of weeks ago. I have the C8 connected to a simple 5 port switch, which is in turn connected to the router. So far I’ve setup a rule to reboot at 4am to ensure the issue is fixed at least once a day while I’m away, but I’m trialing the workaround with the virtual switch.
I also have 2 C-5 version 2.3.9.158, the one in my home have the same problem as you explain but sync the clock to browser don't work for me but if I reboot hubitat it fix it , the cloud connexion work for a day or more and fail again after a while. The hubitat at my cottage never had this problem. This cloud connexion lost start around 2 months ago. As you explain in the hubitat network I am able to ping or tracert with many name as portal.hubitat.com google .com etc so DNS is working it is only the cloud connexion wich is not connecting.
Did the rebuild the database on a reboot have work to fix it?
as a workaround here is a rather complicated rule..
let me explain it uses the hub information driver that hits a maker api cloud url to check if the cloud is still up.. It also uses the same driver that monitors the dns.
It only notifies you and reboots (after you have had a message and possibility to intervene) if the DNS is UP but the cloud is DOWN for a given amount of time.
The dns check is used to signify if the internet is down or not.. The intent is not to get in a reboot loop if the internet is down for an extended amount of time. So the dns up is used to sense that the internet is still up but only the cloud connection is down.