Can Home Assistant observer help to restart ha core remotely?

Configuration
1

Probably due to the DST change, a remote HA setup is not accessible.
Thanks to another system on the remote network, I could make some checks - unfortunately, I did not setup ssh to homeassistant yet:

  • The observer (http://homeassistant.local:4357/) is accessible, and it reports everything is ok;
  • The port 443 served by nginx is responding, but the service it is connecting to is not responding;
    • This is the same as for the public https:// access .
  • Port 5355 is also open - seems to have something to do with multicast.

Does the observer have a specific url to allow more actions?

2

SSH access maybe?

3

I haven’t made the time to add ssh access yet to that system :slightly_frowning_face: .

4

Hello there,

As I’m having exactly the same symptoms right now and my SSH is not running either, I’m wondering how you solved this problem eventually? Last time I managed to resolve this problem by a HA core restart through SSH.

Kind regards,

Csaba

5

I’ve added SSH access.

# A freely chosen name for the connection
Host ha-location
    # Host name or IP for the "remote" Home Assistant
    HostName ha.example.com
    # Public SSH port
    # (I use a different public port - my NAT forwards that port to port 22 on HA).
    # (the example below is the default though.)
    Port 22
    User root
    # Box - I enable this when I want to connect to the Fiber Network Box UI
    # LocalForward 3080 192.168.1.1:80
    # HomeAssistant - Method 1 - use the local DNS name
    LocalForward 3124 homeassistant.local:8123
   # HomeAssistant - Method 2 - use the IP address
    LocalForward 3123 192.168.1.75:8123

With the above, http://localhost:3124 and http://localhost:3123 will connect you to the remote HA UI.

One can likewise configure access to the observer and other devices on the remote network.

My latest issue was that my network operator is chaning the default configurations from IPV4 to CGNAT IPV6 - my dynamic DNS provider does not allow setting IPV6 addresses and I lost remote access. I requested my operator to return to IPV4 full stack, which restored that.

So that could not be fixed remotely because HA was still up and running, but behind an IPV6 address.

1 Like