Service
Infrastructure as Code
If your server fails tomorrow, how long would it take to rebuild it exactly as it was? With Infrastructure as Code, the answer is: the time it takes to rerun a script. Your whole configuration is described in versioned, reviewed and reproducible files, not in an administrator's head.
What we set up
- Server configuration with Ansible: packages, services, users, hardening. Applied identically on every machine.
- Drift detection: every night, every hour or on demand, the servers are compared with the code. Any difference, whether it comes from a contractor, an employee or you, triggers an alert (see our approach).
- Provisioning with OpenTofu, the open-source fork of Terraform maintained by the Linux Foundation: virtual machines, networks, DNS, on your premises or at a European host.
- Everything in Git: every change is tracked, reviewed and can be undone. You know who changed what, when and why.
- Encrypted secrets in the repository with SOPS, or in Vault or OpenBao.
- Deployment through a pipeline: changes go through CI, like code (see DevSecOps).
What you gain
- Disaster recovery: infrastructure that is rebuilt, instead of repaired by hand.
- No drift: no "special" server that nobody dares to touch any more.
- Audit: the configuration is the documentation. For ISO 27001 or NIS2, the evidence is in the Git history.
- Independence: no dependency on one person or one provider, us included.
Our method
- Inventory of what exists: what runs, and how.
- The infrastructure described as code, service by service, starting with the most critical.
- Validation: a rebuild on a test machine, identical to production.
- Handover: documentation and training for your teams.