Edka – Kubernetes Clusters on Your Own Hetzner Account

TL;DR

TL;DR: Edka is a control panel for Kubernetes on Hetzner Cloud — provisions a production-ready k3s cluster in ~2 minutes, then lets you deploy PostgreSQL, metrics, cert-manager, and more through a minimal UI.

Source and Accuracy Notes

• Official site: https://edka.io
• HN Show HN: https://news.ycombinator.com/item?id=44915164
• Apps demo: https://edka.io/apps
• Deployments demo: https://edka.io/deployments

What Is Edka?

If you have ever tried to set up Kubernetes on Hetzner Cloud, you know the drill: configure k3s, set up the Hetzner Cloud Load Balancer integration, install cert-manager, hook up metrics with metrics-server, deal with PV (persistent volume) provisioning — and that is before you even think about deploying your application.

Edka automates all of that. It is a web-based control panel that provisions a lightweight k3s-based Kubernetes cluster on Hetzner Cloud and then layers on a stack of sensible defaults: metrics-server, cert-manager, CloudNativePG for PostgreSQL, and a deployment pipeline with semantic versioning, instant rollbacks, autoscaling, and secrets management.

The author has been running Kubernetes since the early alpha days and spent four years helping small businesses cut cloud costs by moving to Hetzner. Edka is the tool they built to stop repeating the same provisioning work every time.

How It Works

Edka is organized into four layers:

Layer 1 – Cluster Provisioning

Creates a k3s-based Kubernetes cluster on Hetzner Cloud. k3s is a CNCF-certified lightweight Kubernetes distribution that is easier to manage than a full kubeadm cluster and scales well for small-to-medium workloads.

Layer 2 – Add-ons (one-click)

Pre-configured add-ons that Just Work on Hetzner without manual setup:

• metrics-server for resource metrics
• cert-manager for automatic TLS certificates via Let’s Encrypt
• Various operators, pre-wired for Hetzner Cloud Load Balancers

Layer 3 – Applications

Minimal-configuration UIs for stateful applications. The standout example is PostgreSQL: fill in a few fields, and Edka installs CloudNativePG, provisions a highly available PostgreSQL cluster with Point-In-Time Recovery (PITR), and gives you ready-to-use connection endpoints. Restore backups to any moment in time with a single click.

Layer 4 – Deployments

Connect your CI pipeline to push container images to Edka’s registry. From there Edka handles:

• Rolling deployments with semantic versioning rules
• Instant rollbacks to any previous version
• Horizontal pod autoscaling (HPA)
• Persistent volume claims
• Secrets and environment variable imports
• Public service exposure via Hetzner Load Balancers

Tech Stack

Edka itself is built with TypeScript, React, and Tailwind CSS on the frontend, with PostgreSQL, Redis, and BullMQ powering the backend job queue. Sensitive data (credentials, secrets) is encrypted at rest using Vault backed by AWS KMS.

Self-Hosting Comparison

The main alternative is handling everything by hand using k3sup, Terraform, or Ansible playbooks. That works, but every cluster ends up slightly different, and the knowledge lives in your head or scattered across a wiki.

Edka standardizes the setup so all your clusters follow the same proven pattern. The trade-off is trusting the Edka platform with your cluster configuration — if you prefer full control and reproducibility via GitOps, rolling your own with k3s + Flux or ArgoCD is still valid.

When Edka Makes Sense

• You want a production-ready Kubernetes cluster on Hetzner without spending a day configuring it
• You run multiple small-to-medium projects and want consistent cluster setups across all of them
• You need managed PostgreSQL with PITR without paying for a cloud RDS instance
• You are comfortable with k3s and just want the operational overhead reduced

Edka is currently in public beta and free to use.

FAQ

Q: Is this production-ready? A: Edka is in public beta. The author has been using similar setups for four years with small businesses, but evaluate it against your own reliability requirements before running critical workloads.

Q: How does this differ from k3s + Hetzner Terraform provider? A: The Terraform + k3s approach gives you the raw cluster. Edka adds the control plane UI, pre-configured add-ons, a managed PostgreSQL operator, and a deployment pipeline with rollback support.

Q: Can I use an existing Hetzner account? A: Yes. Edka creates resources (servers, load balancers, volumes) in your own Hetzner Cloud account. You retain full access and billing control.

Q: What happens if Edka goes away? A: You have a standard k3s cluster running in your account. You can manage it with kubectl directly, export configs, and migrate workloads to any other Kubernetes environment.

Conclusion

Edka fills a real gap: Hetzner Cloud is one of the best-priced providers for self-hosted Kubernetes, but the operational overhead of setting up a proper cluster has historically made it a project only worth it for teams with dedicated DevOps resources. Edka brings that门槛 down to the point where a solo developer or small team can have a production-ready cluster with managed PostgreSQL, TLS certificates, and a deployment pipeline running in under two minutes.

If you are already running Kubernetes on Hetzner by hand, Edka is worth evaluating as an operational layer on top. If you have been avoiding self-hosted Kubernetes on Hetzner because the setup looked daunting, Edka is the tool that removes that excuse.