This commit moves the pulling of images, packages,
and updating config files into a non-serialized play.

The serialized play is now in charge of marking unschedulable,
draining, stopping and restarting services, and marking
schedulable.

If rpm install / container download takes 60s per host,
this will save 3 hours and 10 minutes at 200 hosts per cluster
and forks of 20 hosts.

This commit refactors some upgrade code paths.

Touched areas are:

1) Reduces usage of 'oo_all_hosts' in various places,
especially when running upgrade_control_plane.

2) Reuses common code across the various upgrade*
playbooks.

3) Moves docker upgrade checks into container_runtime_role.

4) Combines smaller playbooks and plays to reduce file sprawl.

This commit removes openshift.common.service_type
in favor of openshift_service_type.

This commit also removes r_openshift_excluder_service_type
from plays in favor of using the role's defaults.

When a package install/update fails due to network blips or other spotty
availability, retry it. If the failure is a real failure (e.g. package
is really not there) it still fails after 3 tries (Ansible default).

Currently, having openshift_node and openshift_node_upgrade
as two distinct roles has created a duplication across
handlers, templates, and some tasks.

This commit combines the roles to reduce duplication
and bugs encountered by not putting code in both places.