Setup the template in GIT
In this repository, you will find the template code, just copy and store it in a new template repository for testing.
This template repository looks as follows:
.
├── ansible.cfg
├── collections
│ └── requirements.yml
├── group_vars
│ └── all.yml
├── main.yml
├── prep_03_check_credentials.yml
├── README.md
├── roles
│ └── requirements.yml
├── test_01_run_collection_roles.yml
├── test_number_role_rolename.yml
└── vars.yml
Create a new repository in your project named 'test_plan_template' or something like it.
Be sure to make this a template, not modifying the code. This is where you fork the other
testing repositories from.
that's all...
Now you are almost ready to let your creativity take control..
The files in the template repository
ansible.cfg
collections/requirements.yml
group_vars/all.yml
main.yml
prep_03_check_credentials.yml
README.md
roles/requirements.yml
test_01_run_collection_roles.yml
test_101_role_name_test_name.yml
vars.yml
Fork the template
Lets start simple...
Say you want to test apache server in you development environment..
Fork (or make a copy) the test_plan_template into a new repository named
apache_testing.
Clone the new repository to your work environment, using git clone.
You now have a complete framework on your machine to test....
well, nothing yet :-)
The apache_testing repo needs to be configured for the environment
configuration
The framework uses a group_vars/all.yml file for configuration like:
collection_to_test: <namespace.collection>
collection_role_list:
- <namespace.collection.role_name>
env_version: dev
report_host: <report_webserver>
report_dir: <webserver_root>
report_file: report.csv
use_vault_role:
- role_vault
extra_roles: []
required_credentials: []
bb_name: <role_name>
You might recognize some fields, but let me describe them for you anyway:
- collection_to_test
The collection from wich the role will be run to be tested.
This can NEVER be empty or the play will fail. - env_version
The branch/tag to checkout for the role_to_test, this gives the opportunity to test different versions of
roles (later). - report_host
A virtual machine that is configured for ansible, to write a record of testing to.
This is done only when a test is performed from 'Tower' or 'Ansible Automation Platform', so when a test is
scheduled, you can trace back the job_id for this run and its result. This gives your tests tracebility
for management reporting (you could upload the file data into splunk/grafana) to create dashboards. - report_dir
The directory on the reporting server where the file will be written. - report_file
The file that holds all records of the tests that have been run.
Due to the current nature of the reporting, the code overwrites records on key values, if you don't want this
just adapt the template accordingly and updated the exsisting forks. - use_vault_role
If you are using vault, just state the vault repository here, it will be cloned, run and cleaned after the run.
Be sure to add the vault credential to the play. - extra_roles
Extra roles can be used to do configuration before testing..
For example, create a pluggable database on an empty Oracle database platform you want to test.
The role will be cloned into the roles directory, but you will have to call them from your plays yourself.
At the end of the run, these roles a cleaned up. - required_credentials
If you use custom credentials in your environment, you can insert the lookup variables here, so they will
be checked before the run starts, this prevents failed or half-way runs.
If you are happy with the configuration, save it and push it into git, so others can use it.
The base config is now complete.
You could run it now just so you know it runs without error.
Before you do, review the README.md in the apache_test repository, there you will find how to run it.