2021.7: A new entity, trigger IDs and script debugging

@ludeeus I never said that you are horrible. I said the advice you gave to run as fully root is horrible, and I am not the only one who said this. That is not a personal attack. Please learn the distinction of a criticism about a statement and a personal attack, they are not the same thing. It is completely normal to apply critical thinking to a statement made by a person without attacking that person, my statement never said anything about you, only about something you said.

3 Likes

Great Mark(bedankt)

I really appreciate the way you explain the situation. Now a days in these forums you can also get an useless comment with no help at all.

1 Like

use a template condition instead using trigger.id in ['Time Out', 'Alarm On']

1 Like

Actually, the trigger condition supports multiple :slight_smile:

- condition: trigger
  id:
    - "OMG Puppies"
    - "OMG Kittens"
2 Likes

well the doesn’t need to change anything!

The logs say:

A request from a reverse proxy was received from <my-tablet's-local-IP-in-the-same-subnet>, but your HTTP integration is not set-up for reverse proxies

Is there something in the browser’s settings that can trigger this?

Check out the breaking changes. It’s covered there and in the http docs.

I don’t think it is as I don’t do any proxying. This all happens in my local network.

Generally there is nothing in browser setting for that.
Maybe some kind of filter proxy running on the thing? Privacy filters? VPN-ish kinda things?

Whatever the case, Home Assistant get a request from that IP that identifies itself to Home Assistant as a reverse proxy.

Unsurprisingly got a notification that one of my posts was flagged for saying that the advice to run something fully as root was horrible. Let me say so again: I did not attack @ludeeus in anyway or form, I said his advice was horrible, not himself, and at least one user agreed with me on that. I am pretty sure who did it, and it’s a person known for taking criticisms personally. I don’t know in what world running something fully root is acceptable in the Linux community without the developer explaining fully why that is required when there are examples proving that it’s not required, but it’s definitely not a world based on reality, this is not a personal attack it is something widely accepted and taught about Linux and this particular person’s hubris does nobody any favors.

9 Likes

Wow, this indeed was the browser. Fully Kiosk Browser has an option Add XFF Header (PLUS) Apply X-Forwarded-For header to HTTP GET requests

The quotes are mandatory?

Or would the following also work?

- condition: trigger
  id:
    - OMG Puppies
    - OMG Kittens

Great scott, what the pineapple tree it that for an option in a browser? :grinning_face_with_smiling_eyes:

Glad you got it figured out :+1:

Without works, YAML assumes it is a string you are giving it in that case, however, general good practice to quote strings to be explicit.

For example these are 2 different things:

- 0
- "0"

:wink:

1 Like

I’m using NGINX SSL proxy. Do the changes in 2021.7 require an http: config change for trusted proxies? It wasn’t clear to me whether or not this particular config required the change mentioned in the release notes.

1 Like

I believe the answer is yes. I have Traefik with SSL provided by LetsEncrypt and it was required for me.

Yes, you’ll need to change your config. This change applies to all proxies.

Thank you. I’m running an HAOS install. Would the trusted_proxies: IPs be the same as those listed under trusted_networks?

It’s the docker container IP. I would guess it’s 172.30.33.6, however it could be anything. You can update, then try to access it from outside your network. In your logs there will be an error listing the IP.

1 Like

It’s the IP for the docker container, the easiest way is to run your container in host network mode so your DHCP server can assign it a local IP, you can then set your DHCP server to always serve the same IP to your docker container, with the IP essentially becoming static you can setup your reverse proxy (in this case nginx) to that IP and it will technically always work.