beejiujitsu / ansible

ansible

Prerequisites

Use pyenv and pipenv (recommended)

You may want to use pyenv and pipenv to manage your Python interpreter version and your virtualenv.

Install Ansible (traditional method)

If you want to use pipenv, you can skip this step

pip

• pip install ansible (may need to run this as sudo)

Getting started

Git clone

Initialize the project (if you use pipenv)

• pipenv install - create a project, initialize the virtualenv, and install dependencies
• pipenv install --dev molecule docker - add development dependencies to the project (optionnal)

---

Directory structure

.
├── ansible.cfg
├── galaxy_roles
├── inventories
│   ├── group_vars
│   └── host_vars
├── LICENSE
├── log -> /tmp/ansible.log
├── notes
├── playbooks
├── README.md
├── requirements.yml
├── roles
└── tmp

ansible.cfg

This is a project specific configuration file to configure Ansible. View to see a more detailed explanation of the default settings provided by this project.

Read more

• Ansible.cfg - http://docs.ansible.com/ansible/intro_configuration.html

galaxy_roles

This is where you should install your third-party roles from Ansible Galaxy.

Recommended to configure requirements.yml with the third party roles you are using.

Read more

• Ansible Galaxy - http://docs.ansible.com/ansible/galaxy.html

inventories

This is where you should put the list of hosts for Ansible to connect to. It also contains directories for group_vars and host_vars.

Example:

.
├── group_vars
├── host_vars
└── production

We have a file called production, but if you have a lot of environments, instead of production, you might call it production_app_name or app_name_production, etc.

In it you might define it like so:

# inventories/production
[application]
app01.server.com
app02.server.com
app03.server.com

[database]
db01.server.com

[production:children]
application
database

production:children, indicates to Ansible this is a group of made of all the things in the groups listed below.

So its equivalent to manually writing:

# inventories/production
[production]
app01.server.com
app02.server.com
app03.server.com
db01.server.com

group_vars

This directory contains yaml files which correspond to a group name, and tells Ansible to load the variables for the hosts in that group.

Example structure:

.
└── group_vars
    ├── all
    │   └── ubuntu.yml
    ├── application
    │   ├── rails.yml
    │   └── secret
    │       └── rails.yml
    └── database
        ├── mysql.yml
        ├── postgresql.yml
        └── secret
            ├── mysql.yml
            └── postgresql.yml

• all - is a Ansible group, where it will be loaded for all hosts.
• application- is a user defined group of hosts, we defined this in our production hosts file in the previous example.
• database - is a user defined group of hosts, we defined this in our production hosts file in the previous example.

A group variable file looks like this:

# inventories/group_vars/application/rails.yml
---
rails_version: 5.0.0

So the rails_version variable will be available for all hosts listed under application (app[0-3].server.com).

NOTE: You might have noticed the secret directory, this is where you can take advantage of ansible-vault. For secret specific variables such as api keys and etc, I like to put them inside a secret folder, and encrypt the folder with ansible-vault.

Secrets are prefixed with the key secret_.

Example:

# Encrypted secrets for rails
# inventories/group_vars/application/secret/rails.yml
---
secret_rails_secret_token: 1db372dbc0a8200da1d7f9c51384fe0094c940628c497e7e519eb4977aba244cdad0ab9c4965383d2dc9244296ea6b889fe711d40dc590a4455d17e1a012db58

# Unencrypted secrets for rails
# inventories/group_vars/application/rails.yml
---
rails_version: 5.0.0
rails_secret_token: "{{ secret_rails_secret_token }}"

You can see our unencrypted variable file reference the encrypted variable file's variable in secret.

host_vars

host_vars work just like group_vars, but instead of naming it after a group name, you use the hostname. The variables only be loaded for that specific host.

Example:

.
└── host_vars
    └── app01.server.com
        └── rails.yml

Read More

• Inventory - http://docs.ansible.com/ansible/intro_inventory.html
• Variables - http://docs.ansible.com/ansible/playbooks_variables.html

---

log

By default this is just a symlink to tmp/ansible.log.

---

notes

This is where you can add your notes for things like manual configuration.

---

playbooks

This is where playbooks should be placed.

Read More

• Playbooks - http://docs.ansible.com/ansible/playbooks.html

---

requirements.yml

Edit this yml file to add your third-party roles from Github or Ansible Galaxy.

Then install your roles with this command:

ansible-galaxy install -r requirements.yml

Ansible Galaxy - requirements - http://docs.ansible.com/ansible/galaxy.html#advanced-control-over-role-requirements-files

---

roles

This is where you can add your custom Ansible roles for usage by your playbooks.

Read more

Roles - http://docs.ansible.com/ansible/playbooks_roles.html#roles

---

tmp

A tmp directory for convenience. For example, you may prefer to write logs here.

---

LICENSE

MIT