Ansible and Vagrant

The information below is applicable to both Vagrant Ansible provisioners:

  • ansible, where Ansible is executed on the Vagrant host
  • ansible_local, where Ansible is executed on the Vagrant guest

The list of common options for these two provisioners is documented in a separate documentation page.

This documentation page will not go into how to use Ansible or how to write Ansible playbooks, since Ansible is a complete deployment and configuration management system that is beyond the scope of Vagrant documentation.

To learn more about Ansible, please consult the Ansible Documentation Site.

The Playbook File

The first component of a successful Ansible provisioner setup is the Ansible playbook which contains the steps that should be run on the guest. Ansible's playbook documentation goes into great detail on how to author playbooks, and there are a number of best practices that can be applied to use Ansible's powerful features effectively.

A playbook that installs and starts (or restarts) the NTP daemon via YUM looks like:

- hosts: all
    - name: ensure ntpd is at the latest version
      yum: pkg=ntp state=latest
      - restart ntpd
    - name: restart ntpd
      service: name=ntpd state=restarted

You can of course target other operating systems that do not have YUM by changing the playbook tasks. Ansible ships with a number of modules that make running otherwise tedious tasks dead simple.

Running Ansible

The playbook option is strictly required by both Ansible provisioners (ansible and ansible_local), as illustrated in this basic Vagrantfile` configuration:

Vagrant.configure("2") do |config|

  # Use :ansible or :ansible_local to
  # select the provisioner of your choice
  config.vm.provision :ansible do |ansible|
    ansible.playbook = "playbook.yml"

Since an Ansible playbook can include many files, you may also collect the related files in a directory structure like this:

|-- Vagrantfile
|-- provisioning
|   |-- group_vars
|           |-- all
|   |-- roles
|           |-- bar
|           |-- foo
|   |-- playbook.yml

In such an arrangement, the ansible.playbook path should be adjusted accordingly:

Vagrant.configure("2") do |config|
  config.vm.provision "ansible" do |ansible|
    ansible.playbook = "provisioning/playbook.yml"

The Inventory File

When using Ansible, it needs to know on which machines a given playbook should run. It does this by way of an inventory file which lists those machines. In the context of Vagrant, there are two ways to approach working with inventory files.

Auto-Generated Inventory

The first and simplest option is to not provide one to Vagrant at all. Vagrant will generate an inventory file encompassing all of the virtual machines it manages, and use it for provisioning machines.

Example with the ansible provisioner:

# Generated by Vagrant

default ansible_ssh_host= ansible_ssh_port=2200 ansible_ssh_user='vagrant' ansible_ssh_private_key_file='/home/.../.vagrant/machines/default/virtualbox/private_key'

Note that the generated inventory file is stored as part of your local Vagrant environment in .vagrant/provisioners/ansible/inventory/vagrant_ansible_inventory.

Example with the ansible_local provisioner:

# Generated by Vagrant

default ansible_connection=local

Note that the generated inventory file is uploaded to the guest VM in a subdirectory of tmp_path, e.g. /tmp/vagrant-ansible/inventory/vagrant_ansible_local_inventory.

Host variables:

As of Vagrant 1.8.0, the host_vars option can be used to set variables for individual hosts in the generated inventory file (see also the notes on group variables below).

Vagrant.configure("2") do |config|
  config.vm.define "host1"
  config.vm.define "host2"
  config.vm.provision "ansible" do |ansible|
    ansible.playbook = "playbook.yml"
    ansible.host_vars = {
      "host1" => {"http_port" => 80,
                  "maxRequestsPerChild" => 808},
      "host2" => {"http_port" => 303,
                  "maxRequestsPerChild" => 909}

Generated inventory:

# Generated by Vagrant

host1 ansible_ssh_host=... http_port=80 maxRequestsPerChild=808
host2 ansible_ssh_host=... http_port=303 maxRequestsPerChild=909

How to generate Inventory Groups:

The groups option can be used to pass a hash of group names and group members to be included in the generated inventory file.

As of Vagrant 1.8.0, it is also possible to specify group variables, and group members as host ranges (with numeric or alphabetic patterns).

With this configuration example:

Vagrant.configure("2") do |config|

  config.vm.box = "ubuntu/trusty64"

  config.vm.define "machine1"
  config.vm.define "machine2"

  config.vm.provision "ansible" do |ansible|
    ansible.playbook = "playbook.yml"
    ansible.groups = {
      "group1" => ["machine1"],
      "group2" => ["machine2"],
      "group3" => ["machine[1:2]"],
      "group4" => ["other_node-[a:d]"], # silly group definition
      "all_groups:children" => ["group1", "group2"],
      "group1:vars" => {"variable1" => 9,
                        "variable2" => "example"}

Vagrant would generate an inventory file that might look like:

# Generated by Vagrant

machine1 ansible_ssh_host= ansible_ssh_port=2200 ansible_ssh_user='vagrant' ansible_ssh_private_key_file='/home/.../.vagrant/machines/machine1/virtualbox/private_key'
machine2 ansible_ssh_host= ansible_ssh_port=2222 ansible_ssh_user='vagrant' ansible_ssh_private_key_file='/home/.../.vagrant/machines/machine2/virtualbox/private_key'








  • Prior to Vagrant 1.7.3, the ansible_ssh_private_key_file variable was not set in generated inventory, but passed as command line argument to ansible-playbook command.
  • The generation of group variables blocks (e.g. [group1:vars]) is only possible since Vagrant 1.8.0. Note however that setting variables directly in the inventory is not the preferred practice in Ansible. If possible, group (or host) variables should be set in YAML files stored in the group_vars/ or host_vars/ directories in the playbook (or inventory) directory instead.
  • Unmanaged machines and undefined groups are not added to the inventory, to avoid useless Ansible errors (e.g. unreachable host or undefined child group)

For example, machine3 and group3 in the example below would not be added to the generated inventory file:

  ansible.groups = {
    "group1" => ["machine1"],
    "group2" => ["machine2", "machine3"],
    "all_groups:children" => ["group1", "group2", "group3"]

Static Inventory

The second option is for situations where you would like to have more control over the inventory management.

With the inventory_path option, you can reference a specific inventory resource (e.g. a static inventory file, a dynamic inventory script or even multiple inventories stored in the same directory). Vagrant will then use this inventory information instead of generating it.

A very simple inventory file for use with Vagrant might look like:

default ansible_ssh_host=

Where the above IP address is one set in your Vagrantfile:

config.vm.network :private_network, ip: ""


  • The machine names in Vagrantfile and ansible.inventory_path files should correspond, unless you use ansible.limit option to reference the correct machines.
  • The SSH host addresses (and ports) must obviously be specified twice, in Vagrantfile and ansible.inventory_path files.
  • Sharing hostnames across Vagrant host and guests might be a good idea (e.g. with some Ansible configuration task, or with a plugin like vagrant-hostmanager).

The Ansible Configuration File

Certain settings in Ansible are (only) adjustable via a configuration file, and you might want to ship such a file in your Vagrant project.

When shipping an Ansible configuration file it is good to know that:

  • as of Ansible 1.5, the lookup order is the following:
  • Ansible commands don't look for a configuration file relative to the playbook file location (e.g. in the same directory)
  • an ansible.cfg file located in the same directory as your Vagrantfile will be used by default.
  • it is also possible to reference any other location with the config_file provisioner option. In this case, Vagrant will set the ANSIBLE_CONFIG environment variable accordingly.

© 2010–2017 Mitchell Hashimoto
Licensed under the MPL 2.0 License.