Ansible

Ansible for DevOps & SRE (2025)

Ansible is a leading open-source automation tool for managing cloud and on-premises infrastructure. Its agentless, YAML-based approach makes it ideal for DevOps and SRE teams working across AWS, Azure, GCP, Linux, NixOS, and WSL environments.

Why Use Ansible in DevOps & SRE?

  • Cloud Automation: Provision and configure resources on AWS, Azure, and GCP using official modules.

  • Idempotent Deployments: Ensure consistent, repeatable infrastructure changes.

  • Agentless: No software required on managed nodes (uses SSH/WinRM).

  • Integration: Works with Terraform, CI/CD (GitHub Actions, Azure Pipelines, GitLab CI), and Kubernetes.

  • Extensible: Huge module ecosystem for cloud, OS, containers, and more.

Real-Life Examples

1. Multi-Cloud VM Provisioning (AWS & Azure)

- hosts: localhost
  connection: local
  gather_facts: no
  tasks:
    - name: Launch AWS EC2 instance
      amazon.aws.ec2_instance:
        name: devops-ec2
        key_name: my-key
        instance_type: t3.micro
        image_id: ami-0abcdef1234567890
        region: eu-west-1
        tags:
          Owner: devops
      register: aws_result

    - name: Create Azure VM
      azure.azcollection.azure_vm:
        resource_group: devops-rg
        name: devops-vm
        vm_size: Standard_B1s
        admin_username: azureuser
        ssh_password_enabled: false
        ssh_public_keys:
          - path: /home/azureuser/.ssh/authorized_keys
            key_data: "{{ lookup('file', '~/.ssh/id_rsa.pub') }}"
        image:
          offer: UbuntuServer
          publisher: Canonical
          sku: 20.04-LTS
          version: latest
        location: westeurope
      register: azure_result

2. Automated Patch Management (Linux)

- hosts: linux_servers
  become: yes
  tasks:
    - name: Update all packages
      apt:
        upgrade: dist
        update_cache: yes

3. Kubernetes Manifest Deployment

- hosts: localhost
  tasks:
    - name: Apply Kubernetes manifests
      kubernetes.core.k8s:
        state: present
        definition: "{{ lookup('file', 'deployment.yaml') }}"

4. Integrating with GitHub Actions

# .github/workflows/ansible-deploy.yml
name: Ansible Deploy
on: [push]
jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v2
      - name: Install Ansible
        run: pip install ansible
      - name: Run Playbook
        run: ansible-playbook -i inventory playbook.yml

5. LLM Integration for Change Summaries

import openai

openai.api_key = 'sk-...'
summary = openai.ChatCompletion.create(
    model='gpt-4',
    messages=[
        {"role": "system", "content": "Summarize this Ansible deployment log for SREs."},
        {"role": "user", "content": open('ansible.log').read()}
    ]
)
print(summary['choices'][0]['message']['content'])

Best Practices (2025)

  • Use roles and playbooks for modular, reusable code

  • Store secrets in Ansible Vault or cloud secret managers

  • Integrate Ansible runs with CI/CD pipelines

  • Test playbooks with Molecule

  • Use tags for targeted runs

  • Prefer official cloud modules for AWS, Azure, GCP

  • Document all playbooks and roles

Common Pitfalls

  • Hardcoding credentials in playbooks

  • Not using idempotent modules (avoid shell/command when possible)

  • Ignoring error handling (use ignore_errors judiciously)

  • Not validating playbooks before production runs

  • Overusing become without need

References

Ansible Joke: Why did the SRE break up with Ansible? Too many unresolved variables!

PreviouskubectlNextPuppet

Last updated