Bug 2213895

Description of problem:

Currently, the Red Hat branded RHCS 6 docs tell our users to specify `--extra-vars "ceph_origin=rhcs"` when using cephadm-andible.

This feels needlessly complicated as it can also be handled via inventory.

It would be superb if
- the Red Hat RPM included a /usr/share/cephadm-ansible/group_vars/all/foo.yml that has a line `ceph_origin: rhcs`
- a IBM RPM would include one that points to IBM's ceph_origin
- community packages (e.g. those available to CentOS Stream users) would use the upstream default (meaning no addition of group_vars/all/foo.yml)

If this packaging change request is accepted and implemented, then adjust all RHCS docs relevant to cephadm-ansible which contain ansible-playbook ... --extra-vars "ceph_origin=rhcs" (removing the extra vars in the docs)

Version-Release number of selected component (if applicable):
Red Hat Ceph Storage 6 Installation Guide, state as of 2023-06-09

How reproducible:
always

Steps to Reproduce:
1. read the RHCS installation manual (applies to RHCS 5 and 6 docs)
2. wonder why the docs include extra vars
3. create a /usr/share/cephadm-ansible/group_vars/all/ceph.yml containing `ceph_origin: rhcs` on the machine I use to cephadm-ansible
4. file a bug so that we can maybe have this handled in the productized RPMs for all other RHCS users that make use of cephadm-ansible installed via RPM

Additional info: selected component unclassified because I failed to find a component cephadm-ansible but this must b e changed in the RPM before we can change the docs