Skip to content
Edka infrastructure automation platform Edka infrastructure automation platform Edka
SYS.DOCS // DOCS

Create Your First Cluster

This guide will help you set up your first Edka Kubernetes cluster.

Prerequisites

Section titled “Prerequisites”
  • A Hetzner Cloud account. If you don’t have one, see our Prerequisites page for detailed instructions.
  • A Hetzner Cloud API token with read/write permissions. Check our Prerequisites page for detailed instructions.

Optional:

  • kubectl (Kubernetes CLI - only necessary for advanced operations)
  • Headlamp - a Kubernetes IDE (recommended for managing advanced operations with your cluster visually)
  • Lens - a Kubernetes IDE (recommended for managing advanced operations with your cluster visually)

If you don’t have these yet, see our Prerequisites page for detailed instructions.

Note: Kubernetes knowledge is not required to use Edka, but it can be helpful for advanced operations.

Step 1: Create an Edka Account

Section titled “Step 1: Create an Edka Account”
  1. Go to console.edka.io and click Create Account
  2. Sign up using your email or Google account
  3. Verify your email address

Step 2: Prepare Your Hetzner Token

Section titled “Step 2: Prepare Your Hetzner Token”
  1. Generate a Hetzner API token with read/write permissions (see prerequisites)
  2. Decide how you want to use it:
    • Store a token per cluster (recommended)
    • Use a temporary token during creation (not stored)
    • Use a global token integration. Useful for quickly creating a cluster without always having to provide a token.

Edka uses a unique token per cluster (recommended). You can add or update the token later in Clusters > your cluster > Settings > General. After the cluster is created, only the cluster owner and organization owners/admins can update or remove the stored cluster token.

Note: The account-level integration in Settings > Integrations > Hetzner (Global) is mostly recommended for testing and development purposes, but it can be used for production clusters as well.

Step 3: Create Your First Cluster

Section titled “Step 3: Create Your First Cluster”
  1. In the Edka dashboard, navigate to Clusters
  2. Choose a name for your cluster (e.g., “clusterone”)
  3. Click Create Cluster

Provision Cluster

Edka configures an embedded etcd datastore automatically for every cluster.

Optional settings:

  • Cluster location (e.g. Nuremberg)
  • Kubernetes version (e.g., 1.36.0-k3s1)
  • Instance type for control plane and node pools (e.g., CX23)
  • Set a highly available control plane (recommended for production)
  • Use a Load Balancer in front of your Kubernetes API
  • Encrypt Kubernetes secrets at REST
  • Enable deletion protection to prevent accidental cluster removal
  • Schedule Workloads on Control Plane
  • Embedded Registry Mirror
  • Choosing the Operating System. Debian 13 (Trixie) is the default. Currently supported:
    • Debian 13 (Trixie)
    • Ubuntu 24.04 LTS
    • Rocky Linux 10
    • Fedora 43
    • Alma Linux 10
  • Customize the network settings in the Network tab
  • Choose between Flannel or Cilium CNI and enable CNI Encryption
  • Restrict the access to instances SSH and Kubernetes API to your own IPs
  • For clusters larger than 100 instances, you must use Host Firewall (Cilium). This enables Cilium host firewall and forces public networking with Cilium encryption
  • Set the Private Network Subnet
  • Enable NAT Gateway when private-network workloads need managed outbound egress or a stable Floating IPv4 source address
  • Enable cluster backups and optionally set a custom AGE key for backup encryption
  • Choose the number of instances and node pools in the Node Pools tab
  • Set extra node label and taints.
  • Provide a cluster token (stored per cluster), use a temporary token (not stored) or use a token stored per account

Edka will provision your Kubernetes cluster in about 2 minutes. You’ll see the progress and events as the process completes.

Cluster events are always accessible in the Cluster Events tab.

Cluster Progress

Step 4: Access Your Cluster

Section titled “Step 4: Access Your Cluster”
  1. In the Clusters dashboard, open your cluster Overview or go to Settings > Access
  2. Under Kubeconfig, click Download Kubeconfig. Each user gets their own kubeconfig and downloads are audited.
  3. If you need a new kubeconfig, use Rotate Credentials in Access to invalidate the old one and download a fresh version. If your access is revoked, contact the account owner.

Set the KUBECONFIG environment variable to point to the downloaded file:

Admin note: Owners and admins can manage team kubeconfigs in Settings > Access under Team Access. Use Rotate to invalidate a user’s kubeconfig or Revoke to remove access.

Terminal window
export KUBECONFIG=/path/to/your/kubeconfig.yaml

Verify the connection to your cluster:

Terminal window
kubectl get nodes

You should see a list of nodes in your cluster:

Terminal window
NAME                         STATUS   ROLES                       AGE     VERSION
one-master1                  Ready    control plane,etcd,master   2m30s   v1.35.1-k3s1
one-pool-default-worker1     Ready    worker                      1m45s   v1.35.1-k3s1
one-pool-default-worker2     Ready    worker                      1m45s   v1.35.1-k3s1

Alternatively, you can use Lens to manage your cluster visually. Just add the kubeconfig file to Lens and connect to your cluster.

Next Steps

Section titled “Next Steps”

Now that you have your first cluster running, let’s set up some add-ons to transform it into a performant and production ready platform: