yaxis1 / dsti-devops-2020-fall-iac

Vagrant and Ansible lab

Infrastructure as Code

The goal of this lab are to:

• Create a CentOS 7 VM using Vagrant
• Install GitLab on this VM using Ansible
• Perform health checks on the GitLab installation using Ansible

Usefull links

• Vagrant documentation
• Ansible documentation
• Ansible module index
• GitLab installation doc for CentOS 7
• GitLab Health Check doc

Prerequisites

Before you can start the lab, you have to:

1. Install Virtualbox: https://www.virtualbox.org/wiki/Downloads

2. Install Vagrant on your computer: https://www.vagrantup.com/downloads.html

3. (Optional) On Windows, ensure that Hyper-V is disabled:

   1. Open a new Powershell
   2. Run the following command:

      Disable-WindowsOptionalFeature -Online -FeatureName Microsoft-Hyper-V-All

4. Download the centos/7 Vagrant box for the Virtualbox provider:

   $ vagrant box add centos/7
   ==> box: Loading metadata for box 'centos/7'
       box: URL: https://vagrantcloud.com/centos/7
   This box can work with multiple providers! The providers that it
   can work with are listed below. Please review the list and choose
   the provider you will be working with.
   
   1) hyperv
   2) libvirt
   3) virtualbox
   4) vmware_desktop
   
   Enter your choice: 3

5. Install vagrant vbguest pluggin (enables folder mounting)

   vagrant plugin install vagrant-vbguest
   

6. Clone the lab repository to your computer:

   git clone https://github.com/adaltas/ece-devops.git
   

7. Go to the infra-as-code directory:

   cd ece-devops/infra-as-code
   

Lab

Useful information

In the lab, we will use the ansible_local provisioner, so Ansible will be automatically installed on the VM by Vagrant. You don't need it on your computer!

Resources

In the infra-as-code directory, you will find:

• A Vagrantfile that defines the VMs to be managed by Vagrant (1 CentOS 7 VM named gitlab_server in our case)
• A playbooks/ directory that contains Ansible playbooks to install GitLab and run health checks

Lab Part 1

1. Take a look at the Vagrantfile and at the YAML files playbooks/run.yml and playbooks/gitlab/install/tasks/main.yml
2. Create and provision the VM
   1. Run the vagrant up command
   2. You should end up with the following error (this is planned):

      TASK [gitlab/install : Install GitLab] *****************************************
      fatal: [gitlab_server]: FAILED! => {"changed": false, "msg": "No package matching 'gitlab-ee' found available, installed or updated", "rc": 126, "results": ["No package matching 'gitlab-ee' found available, installed or updated"]}

   3. Check that everything is ok by connecting to the VM through SSH:

      vagrant ssh gitlab_server
      

3. Complete the GitLab installation
   1. Fill the blanks in gitlab/install/tasks/main.yml based on the steps 1. and 2. of the GitLab installation doc for CentOS 7
   2. Update the playbooks on the VM using vagrant upload:

      vagrant upload playbooks /vagrant/playbooks gitlab_server
      

   3. Rerun provisioning with the command vagrant provision
4. Test your installation by connecting to http://20.20.20.2 (step 3 of the GitLab installation doc):
   1. Choose a password
   2. Login as the user root using the password

Lab Part 2

1. Read the GitLab Health Check doc
2. Run a healthcheck using curl:
   1. Connect to the VM using vagrant ssh
   2. Run the command:

      $ curl http://127.0.0.1/-/health
      GitLab OK

3. Read playbooks/roles/gitlab/healthchecks/tasks/main.yml to understand how it relates
4. Run the gitlab/healthcheck role
   1. Connect to the VM using vagrant ssh
   2. Run the playbooks using the right tag (replace TAG):

      ansible-playbook /vagrant/playbooks/run.yml --tags TAG -i /tmp/vagrant-ansible/inventory/vagrant_ansible_local_inventory
      

Note: you can also modify the Vagrantfile to change the tag you want to run your playbook with. Then, instead of running the ansible-playbook command inside the VM (after SSH), you can tell vagrant to run ansible provision again with command:

vagrant provision

5. Run the 2 other kind of health checks in the playbook (using the uri module):
   1. Readiness check.
   2. Liveness check
6. Print the results of the health checks in the console
7. (Bonus) Print a custom message with only the disfunctional(s) services in the Readiness check if there are some. To test the printing, stop redis using the command sudo gitlab-ctl stop redis on the node before running the playbook again. Tip: use the json attribute of the response.