Access host machine running web service from container on Arch linux

[wonderplayer]

Hello,

Sorry for bringing this up, but I have an issue accessing my host machine's web server from a container.
Feels like I've looked at almost everything and nothing helped me, which leaves me questioning only if my firewall isn't configured properly. With the little knowledge of firewalls that I have, tried to expose service's ports to home and public network, but that didn't help.

So, Podman version 5.3.2. I have launched a simple web service on http://localhost:5042 and created a container for it on http://localhost:64923.
Then created Alpine container, installed curl on it and tried to call the service with no luck:

[r@r-82b1 ~]$ podman run -it --name AlpineWithCurl alpine:latest 
/ # apk add curl

/ # curl -v http://host.containers.internal:64923/
* Host host.containers.internal:64923 was resolved.
* IPv6: (none)
* IPv4: 169.254.1.2
*   Trying 169.254.1.2:64923...
* Connected to host.containers.internal (169.254.1.2) port 64923
* using HTTP/1.x
> GET / HTTP/1.1
> Host: host.containers.internal:64923
> User-Agent: curl/8.11.1
> Accept: */*
> 
* Request completely sent off
< HTTP/1.1 200 OK
< Content-Type: text/plain; charset=utf-8
< Date: Sat, 25 Jan 2025 16:44:35 GMT
< Server: Kestrel
< Transfer-Encoding: chunked
< 
* Connection #0 to host host.containers.internal left intact

/ # curl -v http://host.containers.internal:5042
* Host host.containers.internal:5042 was resolved.
* IPv6: (none)
* IPv4: 169.254.1.2
*   Trying 169.254.1.2:5042...
* connect to 169.254.1.2 port 5042 from 192.168.50.17 port 42592 failed: Connection refused
* Failed to connect to host.containers.internal port 5042 after 0 ms: Could not connect to server
* closing connection #0
curl: (7) Failed to connect to host.containers.internal port 5042 after 0 ms: Could not connect to server

/ # cat /etc/host
127.0.0.1       localhost
::1     localhost ip6-localhost ip6-loopback
ff02::1 ip6-allnodes
ff02::2 ip6-allrouters
127.0.1.1       r-82b1
169.254.1.2     host.containers.internal host.docker.internal
192.168.50.17   e7eeff0a106e AlpineWithCurl

Is there anything I could check and try out?
I've tried as well creating a custom network which changed IP address of container to be different than my host machines, and also followed instructions how to simulate slirp4netns behaviour with pastas from here and here.
And running rootful container also didn't provide desired result.
Tried same setup on Ubuntu 24.04 LTS freshly installed and updated and the result was the same, so I'm clearly doing something wrong, but can't figure out what...

Thank you!

[wonderplayer]

Update - tried with disabling firewall and no luck.
Maybe it's simply impossible...

[eriksjolund]

I have launched a simple web service on http://localhost:5042

Connecting to the host's localhost is not allowed by default.

You could use one of the pasta options --map-gw, --map-host-loopback or --tcp-ns (-T).

I wrote some examples of how to use rootless Podman with the pasta network driver to connect to the host:

Outbound TCP/UDP connections to the host's localhost

Outbound TCP/UDP connections to the host's main network interface (e.g eth0)

[wonderplayer]

It really works! Thank you very much! Will go through your pages for sure to see what else have I missed :D
It worked with --map-host-loopback and --tcp-ns options, but not with --map-gw, which I tried before as well and didn't work then either.