AnsibleByExample

Ansible azure.azcollection Example: Provision Azure ML and Storage Resources

Beginner
2024-01-013By Luca Berton

Lesson 1 of 15 in Foundations

Ansible Foundations to Production — full curriculum
  1. 1. Introduction to Ansible and Automation
  2. 2. Installing and Configuring Ansible
  3. 3. Running Ad-Hoc Commands
  4. 4. Writing Your First Playbook
  5. 5. Understanding and Using Modules
  6. 6. Variables and Facts
  7. 7. Conditionals and Loops
  8. 8. Handlers and Idempotency
  9. 9. Structuring Projects with Roles
  10. 10. Inventory Management
  11. 11. Using Ansible Vault for Secrets
  12. 12. Debugging and Troubleshooting
  13. 13. Real-World Automation Examples
  14. 14. Final Project: Deploy a Web Server
  15. 15. Bonus - What's Next with Ansible?
View path overview →

At Red Hat Tech Day Netherlands 2026 (3 June 2026, Bunnik), Red Hat announced 12 new content collections for AAP 2.7, including an expanded azure.azcollection covering Azure ML, App Configuration, Front Door, Storage, and Arc/HCI. Below is a runnable playbook that provisions a resource group, a Storage account, and an Azure ML workspace with Ansible.

Example Playbook

---
- name: Provision Azure ML and Storage resources
  hosts: localhost
  connection: local
  gather_facts: false

  vars:
    resource_group: rg-ansible-ml-demo
    location: westeurope
    storage_account_name: ansiblemlstorage01
    ml_workspace_name: ansible-ml-workspace

  collections:
    - azure.azcollection

  tasks:
    - name: Create resource group
      azure.azcollection.azure_rm_resourcegroup:
        name: "{{ resource_group }}"
        location: "{{ location }}"
        tags:
          project: ansible-demo
          managed_by: ansible

    - name: Create Storage account for ML workspace
      azure.azcollection.azure_rm_storageaccount:
        resource_group: "{{ resource_group }}"
        name: "{{ storage_account_name }}"
        account_type: Standard_LRS
        kind: StorageV2
        tags:
          project: ansible-demo

    - name: Create Azure ML workspace
      azure.azcollection.azure_rm_mlworkspace:
        resource_group: "{{ resource_group }}"
        name: "{{ ml_workspace_name }}"
        location: "{{ location }}"
        storage_account: "{{ storage_account_name }}"
        sku: Basic
        tags:
          project: ansible-demo

    - name: Report provisioned resources
      ansible.builtin.debug:
        msg: >
          Provisioned resource group {{ resource_group }},
          storage account {{ storage_account_name }},
          and ML workspace {{ ml_workspace_name }} in {{ location }}.

What This Does

The play runs locally and calls three azure.azcollection modules in sequence. First, azure_rm_resourcegroup creates the container resource group in westeurope. Second, azure_rm_storageaccount provisions a StorageV2 account that the ML workspace uses for artifacts, datasets, and logs. Third, azure_rm_mlworkspace creates the Azure ML workspace itself, wiring in the storage account created in the previous task. Each module is idempotent — re-running the playbook against existing resources reports no change instead of failing or duplicating infrastructure.

This mirrors the kind of Efficiency and Scale gains Red Hat highlighted for the expanded collection: a single playbook stitches together Storage and ML provisioning that would otherwise require separate ARM templates, Bicep files, or manual portal clicks.

Notes / Gotchas

  • Authentication: set AZURE_SUBSCRIPTION_ID, AZURE_CLIENT_ID, AZURE_SECRET, and AZURE_TENANT as environment variables, or pass them as module parameters. On AAP 2.7, use a Microsoft Azure Resource Manager credential type instead of raw env vars.
  • Naming constraints: Storage account names must be globally unique, lowercase, and 3–24 alphanumeric characters — ansiblemlstorage01 is illustrative only and will collide if reused verbatim.
  • Install the collection before running: ansible-galaxy collection install azure.azcollection, plus the Python SDK dependencies (azure-mgmt-storage, azure-mgmt-machinelearningservices, etc.) listed in the collection's requirements.txt.
  • Region availability: not every Azure region supports Azure ML workspaces; verify location against current Azure ML region availability before running in production.
  • Check mode: run with --check --diff first against a scratch resource group to confirm the plan before provisioning real ML and Storage spend.

Takeaway

With the AAP 2.7 refresh of azure.azcollection, provisioning Azure ML workspaces alongside their backing Storage accounts is a straightforward, idempotent Ansible playbook rather than a multi-tool workflow.

References

All Tutorials