The necessary variables are pulled from `regluit-provisioning/group_vars/production/vars.yml` which in turn pulls certain secret values from `vault.yml`.
The variables are split into two files to still allow for searching references in playbook tasks.
To add or view secret values, you must decrypt the file first: `$ ansible-vault decrypt vault.yml` however **always remember to encrypt secret files before pushing to git**. This is done in a similar manner: `$ ansible-vault encrypt vault.yml`.
Ansible also allows for overriding variables from the command line when running playbooks.
This is useful for ad-hoc playbook runs without editing var files.
For example, deploying code from another branch can be done as so:
Currently we are using a static inventory file `hosts` to define target server hosts and groups.
This means that the `hosts` file must be manually updated to reflect things such as DNS changes or additional hosts being added.
In the future, the static inventory file may be replaced with a dynamic inventory solution, such as ansible's [ec2 inventory script](http://docs.ansible.com/ansible/latest/user_guide/intro_dynamic_inventory.html#example-aws-ec2-external-inventory-script)
One important aspect of the `hosts` file is that it defines the groups which a host or hosts are a part of.
Currently, there is one prod host called `regluit-prod` which is a member of the `production` group, and another called `regluit-ondeck` in the `ondeck` group intended to be a build target that can be swapped in to production.
These designations are important, as the `setup-prod` playbook specifically targets the `regluit-prod` host, and only that host will inherit the variables in `group_vars/production/`.
