snapshot-restore.yml

- name: Restore snapshot
  hosts: snapshotted
  become: true
  tasks:
    - name: Check if snapshot exists
      ansible.builtin.stat:
        path: /timeshift/snapshots/by-ansible
      register: check_status
    - name: Start snapshot restore in background
      ansible.builtin.shell:
        cmd: "yes '' | timeshift --scripted --snapshot-device /dev/mapper/{{ inventory_hostname }}--vg-root --restore --snapshot by-ansible --yes"
        executable: /bin/bash
      # start timeshift restore in background, as it will reboot the system by itself
      async: 1200 # 20 minutes should be enough
      poll: 0 # if the command fails, we will get stuck here
      when: check_status.stat.exists
      changed_when: true # we always change something here
    - name: Wait for system reboot # wait for the restore and reboot the system
      ansible.builtin.reboot:
        reboot_command: "sleep 1" # we will not trigger the reboot ourselves, as timeshift will do it
        msg: "Reboot by timeshift every moment now..."
        connect_timeout: 10 # to allow faster reconnects
      when: check_status.stat.exists