Create Lab Machines Using Vagrant

For some time I’ve been interested in Ansible, which is Red Hat’s lightweight, infrastructure automation tool. I found an Ansible tutorial on YouTube from LearnLinuxTV. Jay, the host, uses several VirtualBox VMs as his test lab for Ansible. I decided to go a slightly different route.

I used Vagrant to create and minimally provision a lab environment that I could use for the tutorial. After some trial and error, the Vagrantfile below is my result.

# -*- mode: ruby -*-
# vi: set ft=ruby :

# Vagranyfile API/syntax version
VAGRANTFILE_API_VERSION = "2"

Vagrant.require_version ">= 2.2.0"

Vagrant.configure(VAGRANTFILE_API_VERSION) do |config|

  servers=[
      {
        :hostname => "lab01",
        :box      => "generic/ubuntu1804",
        :boxurl   => "generic/ubuntu1804",
        :ip       => "10.1.1.11"
      },
      {
        :hostname => "lab02",
        :box      => "generic/ubuntu1804",
        :boxurl   => "generic/ubuntu1804",
        :ip       => "10.1.1.12"
      },
      {
        :hostname => "lab03",
        :box      => "generic/ubuntu1804",
        :boxurl   => "generic/ubuntu1804",
        :ip       => "10.1.1.13"
      },
      {
        :hostname => "lab04",
        :box      => "generic/centos7",
        :boxurl   => "generic/centos7",
        :ip       => "10.1.1.14"
      }

    ]

  servers.each do |machine|
    config.vm.define machine[:hostname] do |node|
      node.vm.hostname = machine[:hostname]
      node.vm.box      = machine[:box]
      node.vm.box_url  = machine[:boxurl]
      node.vm.network :private_network, ip: machine[:ip]

      node.vm.provider :libvirt do |libvirt|
        libvirt.memory = 512
        libvirt.cpus = 1
      end
    end
  end

  ansible_pub = File.read(File.join(Dir.home, ".ssh", "ansible.pub"))

  config.vm.provision :shell,
        :inline => "echo 'appending SSH public key to ~vagrant/.ssh/authorized_keys' && echo '#{ansible_pub }' >> /home/vagrant/.ssh/authorized_keys && chmod 600 /home/vagrant/.ssh/authorized_keys"

  config.ssh.insert_key = false
end

The servers array has entries for each lab machine to be created. Each machine has its own hostname, its own IP address, and a Vagrant box image and URL. I’m using a private network for the IP addresses, and I matched the IP address with the host name. lab01 has IP address 10.1.1.11, lab02 is .12, and so on.

The servers.each loop creates each VM, giving it 512MB of RAM and a single CPU. I’m using QEMU/KVM as my virtualization layer, so I added the vagrant-libvirt plugin.

After the VMs are created, I’m using an inline shell provisioner to add an ssh key to the VM. I generated a new key pair just for Ansible, and this inline script puts the public half of that key pair on the VM.

In Ansible, I added a line to the ansible.cfg to specify vagrant as the ssh user.

With this setup I can quickly and painlessly create 4 VMs, using two different operating systems, ready for use in my tutorial. I’m enjoying learning about Ansible, but I had fun creating this lab environment using Vagrant too.

The GitHub repository for my lab environment is called lab is has an MIT license. Help yourself.