[ English | русский | Deutsch | English (United Kingdom) | español | Indonesia ]

Extending OpenStack-Ansible with additional Ansible content

Including OpenStack-Ansible in your project

Including the openstack-ansible repository within another project can be done in several ways:

  • A git submodule pointed to a released tag.

  • A script to automatically perform a git checkout of OpenStack-Ansible.

When including OpenStack-Ansible in a project, consider using a parallel directory structure as shown in the ansible.cfg files section.

Also note that copying files into directories such as env.d or conf.d should be handled via some sort of script within the extension project.

Including OpenStack-Ansible with your Ansible structure

You can create your own playbook, variable, and role structure while still including the OpenStack-Ansible roles and libraries by setting environment variables or by adjusting /usr/local/bin/openstack-ansible.rc.

The relevant environment variables for OpenStack-Ansible are as follows:


This variable should point to /etc/ansible/plugins/library. Doing so allows roles and playbooks to access OpenStack-Ansible’s included Ansible modules.


This variable should point to /etc/ansible/roles by default. This allows Ansible to properly look up any OpenStack-Ansible roles that extension roles may reference.


This variable should point to openstack-ansible/inventory/dynamic_inventory.py. With this setting, extensions have access to the same dynamic inventory that OpenStack-Ansible uses.

The paths to the openstack-ansible top level directory can be relative in this file.

Consider this directory structure:

|- custom_stuff
|  |
|  |- playbooks
|- openstack-ansible
|  |
|  |- playbooks

The environment variables set would use ../openstack-ansible/playbooks/<directory>.

Adding new or overriding roles in your OpenStack-Ansible installation

By default OpenStack-Ansible uses its ansible-role-requirements file to fetch the roles it requires for the installation process.

The roles will be fetched into the standard ANSIBLE_ROLES_PATH, which defaults to /etc/ansible/roles.

ANSIBLE_ROLE_FILE is an environment variable pointing to the location of a YAML file which ansible-galaxy can consume, specifying which roles to download and install. The default value for this is ansible-role-requirements.yml.

To completely override the ansible-role-requirement file you can define the environment variable ANSIBLE_ROLE_FILE before running the bootstrap-ansible.sh script. With this approach it is now the responsibility of the deployer to maintain appropriate versions pins of the ansible roles if an upgrade is required.

If you want to extend or just partially override content of the ansible-role-requirements.yml file you can create a new config file which path defaults to /etc/openstack_deploy/user-role-requirements.yml. This path can be overriden with another environment variable USER_ROLE_FILE which is expected to be relative to OSA_CONFIG_DIR (/etc/openstack_deploy) folder.

This file is in the same format as ansible-role-requirements.yml and can be used to add new roles or selectively override existing ones. New roles listed in user-role-requirements.yml will be merged with those in ansible-role-requirements.yml, and roles with matching name key will override those in ansible-role-requirements.yml. In case when src key is not defined bootstrap script will skip cloning such roles.

It is easy for a deployer to keep this file under their own version control and out of the openstack-ansible tree.

Adding new or overriding collections in your OpenStack-Ansible installation

Alike to roles, collections for installation are stored in ansible-collection-requirements file. Path to this file can be overriden through ANSIBLE_COLLECTION_FILE environmental variable.

The Victoria release of openstack-ansible adds an optional new config file which defaults to /etc/openstack_deploy/user-collection-requirements.yml.

It should be in the native format of the ansible-galaxy requirements file and can be used to add new collections to the deploy host or override versions or source for collections defined in ansible-collection-requirements.

user-collection-requirements will be merged with ansible-collection-requirements using collection name as a key. In case source is not defined in user-collection-requirements, collection installation will be skipped. This way you can skip installation of unwanted collections.

You can override location of the user-collection-requirements.yml by setting USER_COLLECTION_FILE environment variable before running the bootstrap-ansible.sh script. Though it is expected to be relative to OSA_CONFIG_DIR (/etc/openstack_deploy) folder.

Calling extra playbooks during the deployment

If you install some additional deployment functionality as either a collection or a git repository on the deploy host, it is possible to automatically include extra playbooks at certain points during the deployment.

The points where a hook exists to call an external playbook are as follows:

  • pre_setup_hosts_hook

  • post_setup_hosts_hook

  • pre_setup_infrastructure_hook

  • post_setup_infrastructure_hook

  • pre_setup_openstack_hook

  • post_setup_openstack_hook

The hook variables should be configured in a suitable user_variables file. An example calling a playbook from a collection (installed using user-collection-requirements.yml) :

pre_setup_hosts_hook: custom.collection.playbook

Installing extra playbooks using collections, and referencing the playbook with its FQCN is the most robust approach to including additional user defined playbooks.

Installing extra Python packages inside Ansible virtualenv

Some Ansible collections may require presence of specific Python libraries inside execution environment. In order to accomplish that deployer can create /etc/openstack_deploy/user-ansible-venv-requirements.txt file with a list of Python libraries that should be installed inside virtual environment along with Ansible during bootstrap-ansible.sh execution.

You can override the default path to user-ansible-venv-requirements.txt file with USER_ANSIBLE_REQUIREMENTS_FILE environment variable before running the bootstrap-ansible.sh script.

Defining environment variables for deployment

Throughout the documentation we talk a lot about different environment variables that control behaviour of OpenStack-Ansible and Ansible iteself.

Starting with the Zed release a user.rc file can be placed in OSA_CONFIG_DIR (/etc/openstack_deploy) folder and contain any environment variable definitions that might be needed to change the default behaviour or any arbitrary Ansible configuration parameter. These environment variables are general purpose and are not limited to those understood by Ansible.

The path to this file can be changed by setting the OSA_USER_RC variable, but the OSA_CONFIG_DIR and OSA_USER_RC variables cannot re-defined or controlled through the user.rc file.