My Home Assistant was working great, but then I started adding ESPHome integrations.
I ended up re-creating the container a few times, Docker Engine crashed, and I had to do a Docker reinstall.
Once I reinstalled Docker, my internet started dropping out (I was using a cheap WiFi USB dongle at the time). I noticed that there were a few new vEthernet switches all of a sudden that kept switching on and off when my internet would drop. So, I pulled Cat6 cable through the attic to my office as I read that there are issues with vEthernet and DockerNAT with WiFi connections (DockerNAT was not in my Hyper-V switch manager so I created it). Once I hooked up the Cat6 cable, my internet issues went away.
But, now I can’t use my DuckDNS URL to connect anymore and I keep getting errors and running into roadblocks.
As of right now, I can reach HA on localhost:8123 on my host machine.
I can also connect via (ethernet IPv4 address):8123 on any device on the network.
But, when I use my Duckdns URL to connect via iOS HA app (or the web browser on my local machine), I can’t get through.
I have port forwarding setup on my gateway (Xfinity standard gateway/router combo) and everything was working great before the Docker Engine update/reinstall and those vEthernet switches threw everything off.
Not sure where to continue with my troubleshooting. I am relatively new to Docker containers and am not sure what info y’all would need to see to determine any issues.
Extremely frustrated as I figured out my ESPhome problems, but then my setup took a few steps backwards now that I can’t connect to my domain URL.
Your ethernet port will have another MAC address than your WiFi port, which means your routers DHCP service will hand out another IP address to your HA host.
With a different IP your NGinx will point to the wrong IP, camera where no machine will be listening, and since DuckDND point to NGinx, then that will fail too.
In other words go to your router and assign the previous HA IP address to the MAC address of your HA ethernet port.
So, if I’m understanding correctly: since I switched to the ethernet connection, the MAC address is now different than it was with the WiFi and is causing issues with the port forwards.
Having issues removing the old WiFi adapter’s reserved IP address in my router, but I’ll tinker around with it.
Out of curiosity, why doesn’t rebuilding the container bypass these issues?
The MAC address is based on the hardware, so the docker can not really do much there.
You could say that the MAC address is similar to you customer number at your phone company and the phone company then assign your costumer number a phone number (IP address).
If you now cancel the contract with the phone company and wait a month and then sign up for a new contract, then you will get a new customer number and a new phone number will be assigned to that.
You port forwarding is just like the yellow pages.
If you did not tell them that you have a new phone number, then the yellow pages will just point towards the old number, which is not answered by you anymore.
When you go into your router and change the DHCP assignment, then it is equal to contacting the phone company and getting the old number assigned to the new customer number.