Ansible Canary Deployment — Complete Guide

Introduction

Roll out changes to a subset of servers first with Ansible serial, batch_size, and health verification. This guide provides practical examples, best practices, and production-ready patterns.

Quick Start

---
- name: Canary Deployment
  hosts: all
  become: true
  tasks:
    - name: Execute task
      ansible.builtin.debug:
        msg: "Implementing canary deployment"

Method 1: Basic Approach

- name: Basic implementation
  ansible.builtin.debug:
    msg: "Canary Deployment - basic method"

Method 2: Advanced Pattern

- name: Advanced implementation with error handling
  block:
    - name: Main task
      ansible.builtin.debug:
        msg: "Canary Deployment - advanced method"
  rescue:
    - name: Handle failure
      ansible.builtin.debug:
        msg: "Task failed, executing recovery"

Production Example

---
- name: Production canary deployment
  hosts: all
  become: true
  vars:
    app_name: myapp
    environment: production

  tasks:
    - name: Validate prerequisites
      ansible.builtin.assert:
        that:
          - app_name is defined
          - environment in ['dev', 'staging', 'production']

    - name: Execute canary deployment
      ansible.builtin.debug:
        msg: "Running in {{ environment }} for {{ app_name }}"
      register: result

    - name: Verify success
      ansible.builtin.assert:
        that: result is not failed
        fail_msg: "Canary Deployment failed"

Best Practices

Test in check mode first — always run --check --diff before applying
Use variables — parameterize for reuse across environments
Handle errors — use block/rescue/always for graceful failures
Idempotency — ensure repeated runs produce the same result
Documentation — add comments explaining why, not what

Troubleshooting

Issue	Cause	Fix
Task fails	Missing prerequisites	Check dependencies are met
Not idempotent	State not checked before action	Add conditional checks
Permission denied	Insufficient privileges	Use `become: true`
Timeout	Network or resource issue	Increase timeout values

Conclusion

Roll out changes to a subset of servers first with Ansible serial, batch_size, and health verification. Use check mode for validation, handle errors gracefully, and always test in a non-production environment first.