Description of problem:

In order to solve a ceph upgrade issue, the docker2podman Ansible playbook was backported to ceph-ansible 3 (https://bugzilla.redhat.com/show_bug.cgi?id=1853457) this allows us to adapt the FFU procedure for ceph-enabled environments keeping ceph-ansible 3 in the Undercloud until the whole FFU procedure has finished. And only, once the converge step has run, start with the ceph upgrade step (including ceph-ansible update to ceph-ansible 4).

For this to work, a set of changes is required in the tripleo-heat-templates, otherwise the ceph_systemd step (which belongs to the FFU procedure) will fail during the ceph-ansible repo validations.

Currently, there isn't a way to workaround this, as the fail_without_ceph_ansible parameter has the value hardcoded: https://github.com/openstack/tripleo-heat-templates/blob/stable/train/deployment/ceph-ansible/ceph-base.yaml#L634
being impossible to override the value during the FFU process. An alternative would be to skip the validations during the ceph_systemd step, but that could cover up some other issues which are tackled by this validations.

Version-Release number of selected component (if applicable):


How reproducible:


Steps to Reproduce:
1. Run the Undercloud upgrade adding ceph-ansible in to_keep and making sure no ceph repos are configured after leapp.
2. Run overcloud upgrade prepare and avoid passing any value for CephAnsibleRepo
3. Run containers prepare step
4. Run ceph_systemd step for the first controller, it will fail in "TASK [Fail if ceph-ansible doesn't belong to the specified repo] ***************"




Actual results:

ceph_systemd step fails to run

Expected results:

ceph_systemd step works.


Additional info: