Rocky based deployment fails with the msg "No variable found with this name: pre_deployments_ROLE"

[Cagri Ersen]

Description
===========
Rocky based deployment fails with an error related to a non-existent variable:

Server deployments task from /var/lib/mistral/overcloud/deploy_steps_playbook.yaml lookups for a variable like pre_deployments_<ROLE> which couldn't be found and throw an error like below:

```
2019-09-24 23:44:38,983 p=6194 u=mistral | fatal: [controller-0]: FAILED! => {"msg": "No variable found with this name: pre_deployments_Controller"}
...so on
```

When I check the variables in /var/lib/mistrarl/<STACKNAME> folder, I see <ROLE>_pre_deployments variable in group_vars/<ROLE> files for example:
```
Controller_pre_deployments:
  - NetworkDeployment
  - ControllerDeployment
  - ControllerHostsDeployment
  - ControllerAllNodesDeployment
  - ControllerAllNodesValidationDeployment
  - ControllerArtifactsDeploy
```

But there is no variable like pre_deployments_Controller which could be the source of the issue imho.

Steps to reproduce
==================
- Install and prepare undercloud
- Create custom templates for a HCI environment.
- Do overcloud deployment with:
```
openstack overcloud deploy \
  --timeout 120 \
  --templates \
  -r ~/templates/roles_data.yaml \
  -n ~/templates/network_data.yaml \
  -e /tmp/templates/environments/network-isolation.yaml \
  -e /tmp/templates/environments/network-environment.yaml \
  -e ~/templates/nic-mapping.yaml \
  -e ~/templates/network.yaml \
  -e ~/templates/node-info.yaml \
  -e ~/templates/scheduler_hints_env.yaml \
  -e ~/templates/ips-from-pool-all.yaml \
  -e ~/templates/fixed-ip-vips.yaml \
  -e ~/templates/ceph-custom-config.yaml \
  -e /tmp/templates/environments/ceph-ansible/ceph-ansible.yaml \
  -e /tmp/templates/environments/ceph-ansible/ceph-rgw.yaml \
  -e /tmp/templates/environments/disable-telemetry.yaml \
  -e ~/templates/misc-settings.yaml \
  -e ~/templates/timezone.yaml
```

Expected result
===============
Ansible task named Server deployment should be passed w/o any error.

Actual result
=============
Fails with a bunch of error like below:
```
2019-09-24 23:44:38,983 p=6194 u=mistral | fatal: [controller-0]: FAILED! => {"msg": "No variable found with this name: pre_deployments_Controller"}

```
Environment
===========
1. Openstack Rocky

2. 10 baremetal node (3 controller + 7 ComputeHCI node)

[Rabi Mishra (rabi)]

Looks like this was changed recently with https://bugs.launchpad.net/tripleo/+bug/1842141.

After https://review.opendev.org/#/c/680722/ and https://review.opendev.org/#/c/680724/, pre_deployments_{{ role }} is the var rather than {{ role }}_pre_deployments.

I think tripleo-common in your undercloud mistral containers are old.

[Cagri Ersen (cagri-ersen)]

The docker images come from https://hub.docker.com/u/tripleorocky and it seems all of them include tripleo-common version is 9.6.1 (which should be 9.6.2 i think.)

As a workaround I've changed /usr/share/openstack-tripleo-heat-templates/common/deploy-steps.j2 file, pre and post deployment conditions like below:

```
with_items: "{{ '{{' }} lookup('vars', tripleo_role_name + '_pre_deployments')|default([]) {{ '}}' }}"```

```
with_items: "{{ '{{' }} lookup('vars', tripleo_role_name + '_post_deployments')|default([]) {{ '}}' }}"
```