The clear_host_errors module will clear fatal and fail errors. For example, let's say you have a playbook that uses the shell module to ping managed nodes(e.g. target systems).

---
- hosts: all
  tasks:
    - shell: "ping -c1 {{ inventory_hostname }}"
      register: out

    - debug:
        msg: "{{ out.rc }}"

The following should be produced. Notice that the shell TASK returned that server2.example.com was unreachable, and then server2.example.com was not processed in the subsequent debug task.

PLAY [all]

TASK [Gathering Facts]
ok: [server1.example.com]
ok: [server2.example.com]

TASK [shell]
changed: [server1.example.com]
fatal: [server2.example.com]" UNREACHABLE! => {"msg": "Failed to connect to the host via ssh: ssh: connect to host server1.example.com port 22: Connection refused", "unreachable": true}

TASK [debug]
ok: [server1.example.com] => {
    "msg": "0"
}

PLAY RECAP
server1.example.com : ok=2  changed=1  unreacable=0  failed=0
server1.example.com : ok=2  changed=1  unreacable=0  failed=0

The clear_host_errors module will change this behavior, so that server2.example.com is processed in the subsequent debug task.

---
- hosts: all
  tasks:
    - shell: "ping -c1 {{ inventory_hostname }}"
      register: out

    - meta: clear_host_errors

    - debug:
        msg: "{{ out.rc }}"