Description of problem:

When a proxy was set according to the documentation the container pull works fine but the containers are using the proxy configuration to communicate with eachother.

This is due to podman bug RHEL-127541

How reproducible:

Every time a proxy is set

Is this issue a regression from an earlier version:

No

Steps to Reproduce:

1. Add a proxy configuration according to the Satellite documentation [1]

2. Create a new container or attach to one of the running IoP containers 

podman exec -it <container-id-or-name> bash

3. Check if the variable http_proxy was passed to the container 

$ env | grep proxy

Actual behavior:
Due to the passthrough of the proxy setting the containers use the proxy to communicate with each other. This was prevented by the upstream proxy server.

Expected behavior:
The containers should not use a proxy within the podman network.

Workaround:
Setting in /etc/containers/containers.conf:

[containers]
http_proxy=false

Links:

[1] https://docs.redhat.com/en/documentation/red_hat_satellite/6.18/html/installing_satellite_server_in_a_connected_network_environment/performing-additional-configuration-on-server_satellite#configuring-podman-to-use-an-http-proxy

Additonal Notes

For Podman to be able to pull containers, it's sufficient to set https_proxy only and http_proxy is not required.

Update documentation to only set that.

Setting http_proxy=false for containers seems to be a good idea still, as the https_proxy environment variable would still leak into the container and might still result in unwanted behaviour.