Digital sovereignty is not a feature – it is an architectural principle. After transitioning our toolchain to open source, the next step was to strategically rethink the infrastructure behind the tools.
In this post, we showcase how our modern IT architecture looks – container-based, automated, scalable, and fully under our control.
Why Infrastructure Is the Key to Freedom
Many companies replace individual tools but remain structurally dependent:
- on hyperscaling cloud providers,
- on opaque black-box services,
- on rigid hosting models.
Our goal was clear: To create a foundation that offers us maximum control, flexibility, and future-proofing.
Our Architecture at a Glance
We rely on a modular infrastructure that aligns with DevOps principles – built on open standards and fully documentable.
The Core Components:
| Area | Solution / Technology | Why We Use It |
|---|---|---|
| Orchestration & Containerization | Kubernetes | Standardized, scalable, Cloud-native ready |
| Hosting | Bare Metal & Kubernetes | Full control, no vendor lock-in |
| Configuration Management | Ansible | Automated provisioning and updates |
| CI/CD | GitLab CI | GitOps-compatible, self-hosted |
| Monitoring & Logging | VictoriaMetrics, Grafana, VictoriaLogs | High performance, metric-driven alerting |
| Storage & Backup | Velero, Restic, MinIO | Kubernetes-native, S3-compatible, automated |
| Secrets Management | Vault, Infisical | API-driven, Zero Trust ready |
| Authentication | Keycloak, Authentic | Central identity, OpenID Connect, multi-factor possible |
How We Work: Automated and Traceable
GitOps Instead of Click Configuration
All changes to the infrastructure are versioned – no manual clicking around, no opacity. This brings:
- Reproducibility (from testing to production),
- Rollback Capability in case of errors,
- Auditability.
Continuous Delivery – But Controlled
Not everything has to go live automatically immediately. We rely on staged pipelines with manual gates for security-critical deployments.
Benefits We Are Already Experiencing
- Cost Savings by eliminating external PaaS models
- Better Performance, as we operate infrastructure closer to the use case
- Faster Innovation, because we manage dependencies ourselves
- Security by Design – from root policy to container
And above all: We have regained oversight – and control.
Challenges and Solutions
Of course, not everything went smoothly. A few learnings from the transformation:
| Challenge | Solution |
|---|---|
| Resource consumption of Kubernetes | Fine-tuning of nodes and workloads |
| Complexity in monitoring | Standardized dashboards & alert rules |
| Storage redundancy | MinIO + automated backup with Velero |
| Internal skill gaps | Training & pairing in DevOps teams |
Next Step: Efficiently Automating Operations
In part 4 of our blog series, we will show:
- How we efficiently manage operations and maintenance with Infrastructure as Code
- Which automations truly save time
- And why Self-Service IT is not a pipe dream, but a lived reality
Exchange Encouraged
Our infrastructure is no secret – it is an invitation to dialogue. For those on similar paths or looking to start: Let's talk.
What architectural decisions have you made?
What works for you – and what doesn't?
We look forward to the exchange.