Skip to main content

Error Handling in Ansible

When working with Ansible, it's common to encounter some errors. As a playbook grows in complexity, it's essential to understand how to handle these effectively. In this tutorial, we'll explore error handling in Ansible.

Error Handling Methods​

Fail Module​

The fail module in Ansible makes a task fail immediately. It's commonly used to make a task fail when a certain condition is met.

- name: Fail task
  fail:
    msg: "The task has failed."
  when: condition

In this example, if the condition is met, the task fails immediately and returns the message "The task has failed."

Ignore Errors​

In some cases, you might want a task to continue even if it fails. You can achieve this using the ignore_errors directive.

- name: Ignore errors
  command: /bin/false
  ignore_errors: true

In this snippet, the task will continue running even if the command /bin/false fails.

Changed When​

The changed_when statement is used to override the default condition under which Ansible considers a task as changed.

- name: Change conditions
  command: /bin/something
  register: result
  changed_when: "'FAILED' in result.stdout"

In this case, the task is considered changed only when 'FAILED' is in the stdout of the task.

Using Blocks for Error Handling​

In Ansible, a block contains a group of tasks that are logically grouped. Error handling can be applied to all tasks within a block using rescue and always clauses.

- name: Block with error handling
  block:
    - debug:
        msg: 'This is the start of the block'
    - command: /bin/false
    - debug:
        msg: 'This command will not be printed'
  rescue:
    - debug:
        msg: 'This is the rescue section'
  always:
    - debug:
        msg: 'This is always executed'

In this block, if the command /bin/false fails, the playbook jumps to the rescue section. The always section runs regardless of the outcome.

Registering Variables​

Using the register keyword, the output of a task can be saved and used in subsequent tasks. This is helpful for error handling.

- name: Register variable
  command: /bin/something
  register: result
  failed_when: "'FAILED' in result.stdout"

In this case, we're registering the output of the command to a variable 'result'. The task will fail when 'FAILED' is in the stdout.

Max Fail Percentage​

The max_fail_percentage directive allows you to specify a percentage of hosts in which if more than the specified percentage fails, the play is halted.

- hosts: webservers
  max_fail_percentage: 30
  tasks:
    - name: This is a test task
      command: /bin/something

In this example, if more than 30% of hosts fail, the play is halted.

Conclusion​

Error handling in Ansible is a crucial aspect of playbook development. Through methods such as the fail module, ignore_errors, changed_when, blocks, register, and max_fail_percentage, you can effectively manage and control the errors in your Ansible playbook.