The Ansible ecosystem
Summary: Ansible is not an isolated tool. Around it revolves a rich ecosystem: Galaxy to share roles, AWX and Ansible Automation Platform for the enterprise interface, Collections to organize modules, Molecule for testing. This lesson lays out the complete map, then honestly compares Ansible to Terraform, Puppet and Chef to know when to choose what.
1. Overview of the ecosystem
The recommended learning order:
- Ansible Core (basic playbooks and modules)
- Galaxy (downloading community roles)
- Collections (understanding the modern organization)
- AWX or AAP (interface and multi-team management)
- Molecule (professional tests)
This course gives you the conceptual understanding of all these elements. Practical mastery is the subject of the Premium Ansible Course.
2. Ansible Galaxy — the world catalog
Ansible Galaxy is the public registry managed by Red Hat. It is the equivalent of Docker Hub or npm for Ansible.
2026 figures:
- More than 30,000 roles published.
- Several hundred official collections.
- Several tens of millions of downloads per month.
Best practice: always prefer roles with many stars, an active community and a clear license. The official collections (Cisco, VMware, Amazon…) are the safest choice for specialized needs.
3. Collections — the modern organization since 2020
Historically, all Ansible modules were grouped inside Ansible Core — several thousand modules maintained in the same repository. The project was becoming unmanageable.
In 2020, Red Hat introduced Collections: a new way of organizing modules, roles and plugins.
Major benefit: each publisher (Cisco, VMware, Red Hat, Amazon…) maintains its collection at its own pace. Ansible Core stays lightweight, the release cycles are independent.
In 2026, 99% of new Ansible developments go through Collections. Ansible Core itself has been reduced to a minimal engine and a handful of essential modules.
4. AWX and Ansible Automation Platform — the web interfaces
Launching a playbook from the command line works for a solo dev. For a team, you need:
- A web interface to launch the playbooks.
- Rights management (who can launch what on which server).
- A history of past executions.
- Scheduling (running a playbook every night at 3 a.m.).
- Git integration (automatically fetching the playbooks from a repository).
That is the role of AWX and Ansible Automation Platform.
Key features of AWX and AAP:
Practical advice: to learn, install AWX on a small machine or in a minikube cluster. It is free, complete and representative of what large teams use.
5. Molecule — testing your roles like real code
In 2016 appeared Molecule, a tool that automatically tests Ansible roles in ephemeral environments (Docker containers, virtual machines, cloud).
Molecule integrated into a CI pipeline (GitHub Actions, GitLab CI) makes your roles as reliable as a real software project. It is an essential practice in teams that take Ansible seriously.
6. Ansible vs Terraform — the confusion that must be cleared up
This is THE most frequent interview question. These two tools are often used together, but do different things.
Ansible vs Terraform comparison table:
| Criterion | Terraform | Ansible |
|---|---|---|
| Main objective | Provision the infrastructure | Configure and deploy |
| Language | HCL (its own declarative language) | YAML |
| Operating model | Purely declarative (state file) | Idempotent imperative |
| Typical use case | Create 10 AWS machines, a load balancer, a DNS | Install nginx on these 10 machines, configure TLS |
| Multi-cloud support | Excellent (200+ providers) | Good via cloud collections |
| Change management | Sees what changed via a state file | Applies the desired state at each run |
| Main publisher | HashiCorp (IBM acquisition 2024) | Red Hat / IBM |
| Complementarity | Often used BEFORE Ansible | Often used AFTER Terraform |
The dominant 2026 pattern:
- Terraform creates the base: networks, virtual machines, DNS, load balancers.
- Ansible configures what runs: packages, services, applications, secrets.
Can you do provisioning with Ansible? Yes, but less well than Terraform. Ansible has AWS/GCP/Azure modules, but Terraform remains much more complete and robust for that task.
Can you do configuration with Terraform? Yes via remote-exec or cloud-init, but very limited. Ansible is infinitely more powerful for detailed configuration.
Verdict: each on its own turf, the two together.
7. Ansible vs Puppet, Chef, SaltStack
Let us briefly revisit the table from lesson 4, this time with the ecosystem angle.
Recommendation for 2026:
- You are starting a project → Ansible without hesitation.
- You already have Puppet in place → maintain it as long as it works, without investing further.
- You have Chef → plan a migration to Ansible or Puppet.
- You have SaltStack → evaluate according to internal skills; otherwise, migrate to Ansible.
8. Where to find quality Ansible content
Once you understand the ecosystem, knowing where to look becomes essential.
Advice: the book "Ansible for DevOps" by Jeff Geerling is the absolute classic — translated into several languages, updated every year. If you must read a single book about Ansible, it is that one.
Remember in 30 seconds
- Ansible Galaxy = world catalog of shared roles and collections.
- Collections = modern organization of modules by publisher, standard since 2020.
- AWX = open source web interface. Ansible Automation Platform = Red Hat commercial version.
- Molecule = testing your roles automatically in an ephemeral container or VM.
- Ansible vs Terraform: Terraform provisions, Ansible configures. The two together in the modern pattern.
- Ansible is growing, Puppet stable, Chef declining, SaltStack in a niche.
Next: Real use cases: provisioning, configuration, patch, security →