When automating server provisioning with Ansible, handlers are typically used for idempotent operations like service restarts. By design, handlers only execute when notified by a task that reports changed=true. But what if you need to ensure a handler runs every time your playbook executes?

The cleanest solution is using Ansible's meta module to flush handlers:

tasks:
  - name: Force handler execution
    meta: flush_handlers
  
  - name: Explicitly trigger nginx restart
    meta: flush_handlers
    notify: nginx-restart

Another common pattern is creating a task that always reports changed:

tasks:
  - name: Force nginx restart by creating dummy change
    command: /bin/true
    changed_when: true
    notify: nginx-restart

For service management specifically, you might bypass handlers entirely:

tasks:
  - name: Always restart nginx
    service:
      name: nginx
      state: restarted

The meta approach is best when you need handler ordering preserved. The dummy task method works well when you want to conditionally force handlers. Direct service commands are simplest but lose handler benefits.

tasks:
  - name: Conditionally force handlers
    command: /bin/true
    changed_when: force_handlers | default(false)
    notify:
      - nginx-restart
      - postfix-restart

Remember that handlers:

• Run once per playbook run even if notified multiple times
• Execute in the order listed in handlers section
• Only run at the end of each play unless flushed

When working with Ansible handlers, you'll quickly discover they only trigger under specific conditions - typically when a notify directive is called from a changed task. This becomes problematic when you need guaranteed service starts or restarts during playbook execution.

1. Using the meta Module

The cleanest approach uses Ansible's meta module to explicitly flush handlers:

tasks:
  - name: Force nginx restart
    meta: flush_handlers
  
  - name: Explicitly call handler
    meta: flush_handlers
    when: true  # Always executes

2. Creating a Dummy Changed Task

You can trick Ansible into thinking a change occurred:

tasks:
  - name: Force handler execution
    command: /bin/true
    changed_when: true
    notify: nginx-restart

3. Direct Service Module Execution

For critical services, sometimes bypassing handlers entirely makes sense:

tasks:
  - name: Ensure nginx is running
    ansible.builtin.service:
      name: nginx
      state: restarted
      enabled: yes

Combine forced execution with conditionals for more control:

tasks:
  - name: Check if handler should run
    stat:
      path: /tmp/force_restart
    register: force_file
  
  - name: Force handler if file exists
    meta: flush_handlers
    when: force_file.stat.exists

• Handler execution order isn't guaranteed across multiple flushes
• Excessive forced executions may impact performance
• Debugging becomes harder when handlers run unexpectedly