Enable HA always for upgrade job

In Ussuri we enabled pacemaker by default so when we landed a change[0]
in quickstart to handle this logic, it broke the upgrade job because
the ussuri job uses train initially and gets a non-ha standalone which
it tries to upgrade to HA. This results in an incorrect network config.
Since really the exepectation is that we'd always be upgrade HA to HA,
let's test that instead.

[0] https://review.opendev.org/#/c/725782/
Closes-Bug: #1895822

Change-Id: I78c1a0cf68534e574b14ad505404139d93983324

config/general_config/featureset056.yml

@@ -19,6 +19,13 @@ standalone_container_cli: >-
   podman
   {%- endif -%}
 
+# NOTE(mwhahaha): we only care about ha -> ha upgrades. We switched to
+# pacemaker enabled in Ussuri, so the upgrade from Train -> Ussuri is non-HA
+# to HA. We should just test HA always in the upgrade
+standalone_control_virtual_ip: 192.168.24.3
+standalone_custom_env_files:
+  - /usr/share/openstack-tripleo-heat-templates/environments/docker-ha.yaml
+
 standalone_upgrade: true
 # validation