Infrastructure as Code

Infrastructure as code means that networks, servers, storage and services are no longer set up by hand but described in configuration files. Those files live in a repository, are versioned and are treated like application code — with history, review and approval.

Consistency follows from that: development, test and production are created from the same description and differ only in the parameters that are meant to differ. The state of an environment is no longer the tacit knowledge of individuals but a readable file. Every change is traceable and can be reversed deliberately. If an environment fails, it is not reconstructed but recreated — in minutes rather than days.

For provisioning infrastructure we mainly use Terraform and OpenTofu, and Ansible for configuring existing systems. Containers and orchestration are covered by Docker and Kubernetes, with versioning through Git.

The choice follows your environment, not our preference. If your team already works with Puppet or Chef, we build on that rather than demanding a migration nobody asked for. What matters is the structure underneath: reusable modules, dependable state management, and a directory layout that new colleagues can find their way around.

Once infrastructure exists as code, rules can be checked by machine. Requirements on encryption, network access or permitted regions are held as policy and validated automatically on every change — before it is applied, not after.

Credentials and keys do not belong in the repository. We set up central management in which secrets are held separately from the code and their use is logged. For audits this becomes a side effect rather than a project: who changed what, when, and who approved it, is recorded in the history. Evidence accumulates while the work is being done.

The immediate benefit is time: provisioning that used to take days runs in minutes and without a round of coordination. The larger benefit is independence — your infrastructure is documented because the documentation is the infrastructure.

That lowers the risk when staff change, shortens onboarding and makes growth predictable. New sites, tenants or test environments become a question of parameters rather than a question of projects. And it is the foundation for everything else: without describable infrastructure there is no sensible automation and no dependable recovery.