handle_unreachable_hosts_method_3.yml

---

# in this method, the playbook will show as success in Tower, listing unreachable systems as unreachable at the end
# unreachable systems identified in the first play
# any tasks/roles can be executed in the 2nd play on reachable systems
# result: all tasks are executed on all reachable systems

- name: ping hosts
  hosts: all
  gather_facts: false
  tasks:
    - block:
        - name: ping host
          win_ping:
          ignore_unreachable: true
      always:
        - name: add reachable hosts to a new group
          add_host:
            name: "{{ inventory_hostname }}"
            group: reachable_hosts

    - debug:
        msg: "failed hosts: {{ ansible_play_hosts_all | difference(groups.reachable_hosts) }}"
      run_once: true

    # you can take any action here on behalf of failed hosts
    # remember to use (run_once: true) and (delegate_to: localhost) as appropriate
    # why delegate_to: localhost? you might ask. Imagine a scenario where all hosts failed.
    # in that case, there are no hosts to execute any tasks on, you can delegate_to: localhost for any action needed
    # for example send email, make api call to update a record somewhere etc.

- name: other play
  hosts: reachable_hosts
  tasks:
    - debug:
        msg: "{{ inventory_hostname }} is reachable"