Browse Our Content Library

We ran Nginx Ingress for years. Then Kubernetes announced its retirement in November 2025. Because we'd adopted Gateway API, migrating to Envoy Gateway was straightforward. Here's the pattern that makes infrastructure components replaceable.

Putting all your workloads with one provider creates negotiating leverage—for them. European providers are competitively priced, not cheaper. The value is optionality: the ability to credibly discuss alternatives, which pays for itself in contract negotiations.

Data centers consume about 1.5% of global electricity and growing rapidly. When you choose where your workloads run, you're making environmental decisions. Here's how to evaluate sustainability claims, why Nordic regions and France matter, and the pragmatic European position on nuclear power.

We run production on Scaleway. Here's how we evaluated European providers, what criteria actually matter for production workloads, and why Scaleway's combination of managed Kubernetes plus essential services sets the standard for European cloud independence.

The Terraform license change taught us: widely adopted doesn't mean truly open. Kubernetes is different—CNCF-governed with over 88,000 contributors from 8,000+ companies. Your manifests work on Scaleway, OVHcloud, Hetzner, AWS, or bare metal. This is the portability foundation that makes independence possible.

When HashiCorp changed Terraform's license in August 2023, organisations that treated it as a standard got a wake-up call. Here's what we learned, why we moved to OpenTofu, and how Kubernetes demonstrates what a true standard looks like.

When the React2Shell vulnerability hit Next.js in December 2025, our architecture enabled rapid response: own your container builds, GitOps deployment, Kubernetes rollout controls. Security patch deployed in 3 hours. Independence isn't just sovereignty—it's agility.

Flux's trust model: clusters pull their own configuration from Git. No CI/CD credentials everywhere. Secrets encrypted with SOPS. Self-reconciliation. This isn't just security—it's reduced anxiety and operational overhead.

Instead of Prometheus and Grafana in every environment, we built a management cluster that monitors all clusters via remote write. Single pane of glass. Reduced resource usage. But the real win: reduced operational complexity.

Honest numbers: Scaleway is competitive with hyperscalers, not cheaper. After year-one credits expire, costs are comparable. The value isn't primarily cost savings—it's European jurisdiction, reduced lock-in, and negotiating leverage.

Don't migrate everything. Pick a pioneer workload: stateless, containerized, low-risk, representative. Learn what works, build team capability, prove the model. Here's how to choose your first migration and execute it successfully.

The 'you need a huge DevOps team' myth prevents organizations from trying. With Kubernetes, GitOps, and AI assistance, 3-6 engineers can run production infrastructure. The key: sustainable on-call rotation (1 week every 2 months) and smart tooling choices.

Most containers request 2GB, use 200MB. Prometheus queries show actual vs requested. Right-sizing saves money, improves density, reduces environmental impact.

Step-by-step: choose pioneer workload, set up OpenTofu, use Flux, encrypt secrets with SOPS, deploy to Kubernetes. Real pattern from our experience.

Claude isn't writing all our code—it's helping us understand complex systems faster. OpenTofu modules, Kubernetes manifests, debugging GitOps issues.

Management cluster queries metrics from test/prod clusters. Secured with mTLS via Gateway API. The implementation: separate Gateways, client certificates, ReferenceGrants.

Our disaster recovery isn't theoretical—we've tested it. Fresh cluster from Git: OpenTofu provisions infrastructure, Flux bootstraps, under an hour to working cluster. Drift is the enemy.

Secrets encrypted in Git, per-environment keys, Flux automatic decryption. Our actual implementation: kubeconfig files in project root, SOPS .enc.yaml files, age keys in GitHub secrets.