A collective of platform engineers who built and operated a multi-tenant internal developer platform powering some of Europe's largest online marketplaces — now sharing what we learned along the way.
We're a team of engineers who spent years designing, building, and operating a shared internal developer platform — a multi-tenant Kubernetes offering that let dozens of product teams build and ship with speed, security, and scale.
Behind the scenes, that platform powered marketplaces serving millions of users every day:
🛒 Leboncoin (France) · 📦 Kleinanzeigen (Germany) · 🇮🇹 Subito (Italy) · 🏠 Milanuncios (Spain)
Running one platform across many countries, brands, and workloads taught us a lot about building platforms as products, managing tenancy and isolation, and supporting genuinely diverse needs from a single foundation. Runetes is where we write it all down.
Real stories from the field — the wins, the outages, the trade-offs, and the hard-won operational wisdom. No sanitized case studies; the actual lessons from running a platform that thousands of engineers depend on.
Recent writing from the team:
- 🧰 Running Multi-Tenant Kubernetes Clusters at Scale: Our Production Toolbox — the operators, components, and patterns we rely on in production.
- 🌐 A Journey Through Kafkian SplitDNS in a Multi-Tenant Kubernetes Offering — serving one tenant's odd requirement without breaking everyone else.
- 📉 When VerticalPodAutoscaler Goes Rogue: How an Autoscaler Took Down Our Cluster — a misconfiguration that hid for months, and what it cost us.
- 🌍 How We Avoided an Outage Caused by Running Out of IPs in EKS — networking at scale is never "just networking."
- 🏗️ How We Migrated 30 Kubernetes Clusters to Terraform — bringing a fleet under infrastructure-as-code.
📚 Browse the full publication → Learnings from the PaaS
| 🏢 Platform as a Product | Treating internal teams as customers — paved roads, golden paths, and knowing when not to build. |
| 🧩 Multi-Tenancy & Isolation | Sharing infrastructure safely across teams, brands, and countries without blast radius surprises. |
| ⚙️ Operating Kubernetes at Scale | Fleets of clusters across regions and accounts — upgrades, autoscaling, and day-2 reality. |
| 🛡️ Reliability & Incidents | Post-mortems written honestly, so the next team doesn't relearn them the hard way. |
| 💸 Efficiency & Cost | Getting more out of every node — right-sizing, ARM, and doing the hard optimization once for everyone. |
| 🏗️ Infrastructure as Code | Managing large fleets declaratively, and the migrations that get you there. |
If you're scaling an internal platform, wrestling with multi-tenancy, or just curious about what platform engineering looks like at scale — you're in the right place.
- 📖 Follow us on Medium — Learnings from the PaaS
- 💬 Have a question or a war story of your own? Open a Discussion or start an issue in one of our repos.
Runetes — from ᚱ Raidho, the rune of the journey, and -netes, for the platforms we run.
Stories from the field, for everyone building the road ahead.