This folder provides some basic setup for the examples used in the book. This setup uses Ansible to quickly create virtual machines with some important tools we'll be using to explore containers and Kubernetes. You will need access to Linux to run Ansible; this could be a cloud machine or Windows using the Windows Subsystem for Linux (WSL).
This setup directory supports two different methods for virtual machines: Amazon Web Services (AWS), or VirtualBox using Vagrant. AWS does not require any local resources, but you'll pay for the cloud resources you use. Vagrant uses your local computer to run virtual machines, so you'll need to have at least 16GB of RAM and 6 processor cores (the more memory and CPUs, the better).
NOTE for AWS The AWS instances created by these playbooks are not
terminated automatically. Remember to terminate the instances using the
aws-teardown.yaml playbook!
To use this Ansible configuration, you must first install Ansible. This can
be done using the Python package installer, pip.
How you install pip will depend on your operating system and distribution.
For example, on Ubuntu:
sudo apt update
sudo apt install -y python3-pip
Install Ansible with pip:
pip3 install ansible jmespath
This command also installs the jmespath Python module, which is needed
for some JSON query processing done as part of the Kubernetes cluster
initialization.
If you already use Ansible and have it installed, note that these examples were tested with Ansible 2.10.6. They are unlikely to work with any version of Ansible prior to 2.10.
You can choose to run sudo in front of pip3 commands to install for all
users, but be consistent or Ansible may have problems finding the Python
modules you install.
Also, if you installed Ansible as a regular user, you may need to add
$HOME/.local/bin to your PATH in order to be able to run Ansible
commands.
There are some additional setup steps to use AWS.
First, run a few more commands that will fetch some dependencies our Ansible configuration needs:
ansible-galaxy collection install -r collections/requirements.yaml
pip3 install -r requirements.txt
pip3 install -r collections/ansible_collections/amazon/aws/requirements.txt
You will also need to configure AWS (e.g. using aws configure) per the
usual Amazon Web Services instructions. This configuration is used by Ansible
to create resources on your behalf.
Finally, you need to make sure ssh-keygen is installed on your system as we
will use it to manage SSH keys to communicate with our instances in AWS. It
is typically installed by default on most Linux distributions.
Review the AWS variables in roles/aws-instances/defaults/main.yaml to see if
any changes are required. In particular, you might want to change the
aws_region to a region that is closest to you. If you do change the region,
you must also change the region in ec2-inventory/aws_ec2.yaml to match the
region you selected.
Once you've completed the setup, you are ready to use AWS to create virtual machines for each chapter. Just change into the correct directory for that chapter and run:
ansible-playbook aws-setup.yaml
Note that creation will take a while as we need to wait for the instance to finish initializing before we can collect the SSH host key.
NOTE Ansible will warn that the instance did not start before timeout. This is expected and the task to fetch host keys has a retry in it to wait for instance startup. Also note that if AWS is a little slow in assigning a public IP to an instance, Ansible may fail the first time. This is not an issue; just run it again and it should succeed.
To make sure any AWS instances you've created are cleaned up, just run this command in this directory or any chapter directory:
ansible-playbook aws-teardown.yaml
This command will delete any EC2 instances with the env tag of
aws_ansible_k8s, so it will clean up any instances from all chapters.
The aws-teardown.yaml playbook leaves some AWS resources intact to save
time when moving from chapter to chapter. These resources don't have any
ongoing associated costs, but if you use the AWS console you may notice
them.
If you want to completely clean up all the examples from this book, you can run the delete script in this directory to clean up these remaining resources:
ansible-playbook aws-delete.yaml
Each chapter contains a Vagrantfile (or, where the chapter explores multiple
different options, multiple directories each holding a Vagrantfile). This
allows the use of a separate set of virtual machines for each chapter to
avoid the setup from one chapter conflicting with the next.
To get started with Vagrant, we need a hypervisor. The Vagrantfile used
for each chapter is designed for VirtualBox, the default hypervisor
for Vagrant.
Installation of VirtualBox will be different for each OS or distribution. For example, on an Ubuntu system, the following command (run as root) should work:
apt install virtualbox libarchive-tools
In general, follow the instructions on the VirtualBox site.
Once VirtualBox is installed, the next step is to install Vagrant. For some distributions, Vagrant is available as a package in the standard repositories. However, it is best to review the instructions on the Vagrant site and install using the Hashicorp repositories if possible, as the version in the standard repositories may not have all the necessary features. For example, Fedora 34 users have reported trouble with both Vagrant 2.2.7 and with Vagrant 2.2.19 when installed from the Fedora repos.
Thanks to Xander Soldaat and to Abe Hanoka for their feedback regarding using Vagrant on Fedora with these examples.
Once VirtualBox and Vagrant are installed and working, and Ansible is
installed and set up per the instructions above, you can simply run
vagrant up in any directory with a Vagrantfile. This will automatically
apply the primary automation for that chapter. See the README.md file in
each chapter for more information.