napya_roleconversion.md

Playbook to Role Conversion

GOAL:

• Convert the playbook below into a role.
• Call your newly created role from a second playbook.

1. First, let's reset any changes we've made to any of our Ansible inventories, configs, and whatnot.

   student@bchd:~$ bash ~/px/scripts/full-setup.sh

2. Now create your new role's directory structure. We're naming the role challenge.

   student@bchd:~/.ansible/roles$ mkdir -p ~/.ansible/roles && ansible-galaxy init ~/.ansible/roles/challenge

3. Move into your new role directory.

   student@bchd:~/.ansible/roles$ cd ~/.ansible/roles/challenge

4. If you haven't yet, install the tree application, it can really help!

   student@bchd:~/.ansible/roles$ sudo apt install tree -y

5. Use the tree application to get a better visual of the role you created.

   student@bchd:~/galaxy/roles/challenge$ tree ~/.ansible/roles/challenge

6. The goal is to take the playbook below and convert it into a role structure. Here are some hints on what to do:

   • You will need to edit vars/main.yml and tasks/main.yml. Cut and paste!
   • For this basic role, you won't need any other directories besides var and tasks.
   • Remember that play keywords like hosts, connection, gather_facts, become, become_method, and the name of the play DO NOT belong in roles!
   • If you'd like to see what the inside of a role looks like, check out the Minecraft server setup role we looked at the other day!

   ---
   - name: Arista playbook that triggers failure
     hosts: arista_switches
     connection: network_cli
     become: yes
     become_method: enable
     gather_facts: no
   
     vars:
       ansible_ssh_pass: alta3
       current_os: eos  
   
     tasks: 
       - name: Back up current running config
         arista.eos.eos_config:
           backup: yes
         register: results
         when: ansible_network_os == "{{ current_os }}"
   
       - name: Block of change modifications to our switch
         block:
           - name: configure the login banner
             arista.eos.eos_banner:
               banner: login
               text: |
                 this is my login banner
                 that contains a multiline
                 string
                 if this works correctly
                 we should never see the banner because the
                 config will be rolled back by our rescue
               state: present
   
           - name: trigger a failure (to prompt rescue)
             shell: "/bin/false"
   
         rescue:
           - name: load configuration from file
             arista.eos.eos_config:
               src: "{{ results.backup_path }}"
               replace: config
   
         always:
           - name: Some task you "ALWAYS" want to perform
             debug:
               msg: "An ALWAYS task might be the mail module, to send an email to engineers or logging indicating a job just ran and its status."

7. When you feel you've successfully created the role, try and run it from a playbook! Click the dropdown if you need an example of how this is done.

Playbook calling a Role

---
- name: executing a role
  hosts: arista_switches
  connection: network_cli
  become: yes
  become_method: enable
  gather_facts: no

  roles:
    - challenge

Steps to create necessary files for the role

1. Create the vars/main.yml file and add the necessary variables:

   student@bchd~$ vim ~/.ansible/roles/challenge/vars/main.yml

   ---
   ansible_ssh_pass: alta3
   current_os: eos

2. Create the tasks/main.yml file and add the tasks:

   student@bchd~$ vim ~/.ansible/roles/challenge/tasks/main.yml

   ---
   - name: Back up current running config
     arista.eos.eos_config:
       backup: yes
     register: results
     when: ansible_network_os == "{{ current_os }}"
   
   - name: Block of change modifications to our switch
     block:
       - name: configure the login banner
         arista.eos.eos_banner:
           banner: login
           text: |
             this is my login banner
             that contains a multiline
             string
             if this works correctly
             we should never see the banner because the
             config will be rolled back by our rescue
           state: present
   
       - name: trigger a failure (to prompt rescue)
         shell: "/bin/false"
   
     rescue:
       - name: load configuration from file
         arista.eos.eos_config:
           src: "{{ results.backup_path }}"
           replace: config
   
     always:
       - name: Some task you "ALWAYS" want to perform
         debug:
           msg: "An ALWAYS task might be the mail module, to send an email to engineers or logging indicating a job just ran and its status."