Bug 2051627

This bug was initially created as a copy of Bug #2046211

I am copying this bug because: 

The same bug applies to OSP 16.1

Description of problem:

The mechanism used by Leapp to identify issues during the RHEL upgrade relies on some python clases named as actors. These actors block the upgrade in case of wrong configuration, or wrong interface names ,etc ... However, RHOSP required to workaround some of these actors to allow customers continue with their upgrade in specific situations, as well as to make faster our CI testing.

For that reason, a block of code was included in tht that would allow us, if required, to remove some of these actors at the time of upgrading Overcloud nodes:
https://github.com/openstack/tripleo-heat-templates/blob/stable/train/deployment/tripleo-packages/tripleo-packages-baremetal-puppet.yaml#L408

            - name: Remove leapp actors to prevent them inhibiting the upgrade
              vars:
                actors_to_remove: { get_attr: [RoleParametersValue, value, 'upgrade_leapp_actors_to_remove']}
              file:
                path: /usr/share/leapp-repository/repositories/system_upgrade/el7toel8/actors/{{ item }}/actor.py
                state: absent
              loop: "{{ actors_to_remove }}"

This code was introduced in:
16.2: https://bugzilla.redhat.com/show_bug.cgi?id=1984875
16.1: https://bugzilla.redhat.com/show_bug.cgi?id=1984873

In version leapp-0.13.0-100 the location of these actors was changed from:

/usr/share/leapp-repository/repositories/system_upgrade/el7toel8/actors/
into
/usr/share/leapp-repository/repositories/system_upgrade/common/actors/

making the block of code referenced above irrelevant, as no file will be removed from that directory and at the same time, making fail the other mechanism put in place to allow the upgrade with unsupported drivers:

https://github.com/openstack/tripleo-heat-templates/blob/stable/train/deployment/tripleo-packages/tripleo-packages-baremetal-puppet.yaml#L404-L411

We need to change the references to the directories or come up with a more robust solution which will work even if there are directory changes in the future.

This BZ impacts to all our CI jobs, making them fail:

http://rhos-ci-logs.lab.eng.tlv2.redhat.com/logs/rcj/DFG-upgrades-ffu-16.1-from-13-latest_cdn-3cont_3hci-ipv4-ovs_dvr/249/undercloud-0/home/stack/overcloud_system_upgrade-controller-0.log.gz


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


How reproducible:

Run any FFU CI job from OSP13 to OSP16.1