Role fails with check handler when installing on OS with IPv6 disabled

[ndjilinski]

Bug Overview

When the installation is happening on the OS with IPv6 disabled, the role stops on the config check handler:

`
RUNNING HANDLER [nginxinc.nginx : (Handler) Check NGINX] *********************************************************************************************
fatal: [host]: FAILED! => {"changed": false, "cmd": ["nginx", "-t"], "delta": "0:00:00.011838", "end": "2025-05-01 05:38:43.724383", "msg": "non-zero return code", "rc": 1, "start": "2025-05-01 05:38:43.712545", "stderr": "nginx: the configuration file /etc/nginx/nginx.conf syntax is ok\nnginx: [emerg] socket() [::]:80 failed (97: Address family not supported by protocol)\nnginx: configuration file /etc/nginx/nginx.conf test failed", "stderr_lines": ["nginx: the configuration file /etc/nginx/nginx.conf syntax is ok", "nginx: [emerg] socket() [::]:80 failed (97: Address family not supported by protocol)", "nginx: configuration file /etc/nginx/nginx.conf test failed"], "stdout": "", "stdout_lines": []}
...ignoring

RUNNING HANDLER [nginxinc.nginx : (Handler) Print NGINX error if syntax check fails] *****************************************************************
fatal: [host]: FAILED! => {
"config_check['stderr_lines']": [
"nginx: the configuration file /etc/nginx/nginx.conf syntax is ok",
"nginx: [emerg] socket() [::]:80 failed (97: Address family not supported by protocol)",
"nginx: configuration file /etc/nginx/nginx.conf test failed"
],
"failed_when_result": true
}
`

Expected Behavior

I understand the nginx.conf is coming with the package, but, the role - at least - should have the trigger to "disable" the check. The role already has the control over the service enablement and start. and the nginx_config can override the nginx.conf.

Steps to Reproduce the Bug

In the target host - disable IPv6 and call the playbook with included nginx role - using those vars:

nginx_manage_repo: false
nginx_install_from: os_repository
nginx_install_epel_release: false
nginx_start: false

Environment Details

• Target deployment platforms: Linux VM on VMware
• Target OSs: RHEL9 (IPv6 disable)
• Host OS (where you are running Ansible from): RHEL9
• Version of the NGINX Ansible role (or specific commit):
• Version of Ansible: 2.18.5
• How is Ansible being managed: CLI
• Version of Jinja2 (if you are using any templating capability): 3.1.6

Additional Context

No response

[alessfg]

Is there any specific reason why you are installing NGINX from the RHEL repo instead of using the NGINX repo (by using nginx_install_from: nginx_repository)? This seems to be partly due to the default NGINX config that ships in the package shipped by RHEL instead of us (the default NGINX config does not listen on IPv6).

That being said, there is probably no harm in making the config check optional when installing NGINX from a OS/distro repository. I will also pass along the issue here to the RHEL team to see if they can change their defaults in the future.