Network problem or DNS-Problem

I’ve had a configuration problem for over a week where I can’t get any further.
I am using the Home Assistant in Docker.

The system is no longer accessible from the outside, not even the other services running on the Raspberry (behind a Fritzbox with port sharing). Also, some apps in the HA no longer work because they no longer have a connection to the Internet, even the current update from 2024.1.3 to 2024.1.5 cannot be carried out.
Outgoing pings are no longer attempted via eth0 but via a virtual interface ( veth*).

If I restart the system, everything works, but only for a few minutes.
I can provide screenshots or more information at a later date.

I hope that someone can help me.

What changed or what happened immediately before it started.

I learned from a very good infrastructure engineer 20 years ago ‘all issues are name resolution (DNS) issues until proven otherwise.’

Still true. So what changed.

1 Like

If I knew, I wouldn’t be able to answer that exactly.
I have updated the system from Raspbian Bullseye to Bookworm.
I wonder why the pings first go out correctly from eth0 and then later from veth. This is not clear to me at all.
Also why I can always reach the system internally, but not from outside (or only in the first few minutes).
A firewall is not running.

Check your routes and make sure you NetworkManager, or whatever you use for managing your network interfaces, are correctly configured and working properly.

1 Like

Are you taking updates? I’d so what happens if you roll back to the version you had before it stopped?

Maybe not a firewall but from a network perspective - the same thing.

If you can’t pinpoint an update that happened in that time then I suggest drawing yourself a diagram with your subnets and draw your comms back and forth through your expected paths. Think through each hop. If I want to get to the internet from HA to do this what’s my path? And verify every hop with what the config is.

My bet is something at the Fritzbox

I was able to narrow down the cause:
As soon as I stop the service “hassio-supervisor.service” I have full network functionality again.
The cross-check by starting the service returned the problems.
Let’s see how I get on with it.

Read about the recent problems with the latest docker upgrade and supervised ?