To create a new cluster, open the Kubermatic Kubernetes Platform (KKP) dashboard, choose a project, select the menu entry Clusters
and click the button Add Cluster
on the top right.
Enter a name for your cluster and click Next. Here you can also activate Audit Logging and Pod Security Policy, assign labels to your cluster, and choose a Kubernetes version.
![Overview of cluster creation](03-create-cluster-choose-name.png)
Choose the cloud provider:
![Menu to choose cloud provider](03-create-cluster-choose-provider.png)
and the datacenter region closest to you:
![Menu to choose datacenter](03-create-cluster-choose-region.png)
In the next step of the installer, enter the API token into the Provider credentials
field. If you chose DigitalOcean, your view will look like this:
![Overview of cluster settings](03-create-cluster-api-tokens.png)
If you entered a valid API token, your node settings will be pre-filled:
![Overview of cluster settings with prefilled node section](03-create-cluster-node-settings.png)
Scroll down to choose or add an SSH key. You can choose one of the keys you already created for the project, or create a new one.
You can assign labels to your nodes. You can also set node taints which is a property that allows your node to repel certain pods.
The chosen SSH key will be used for authentication for the default user (e.g. ubuntu
for Ubuntu images) on all worker nodes. When you click on Next
, you will see a summary and the cluster creation will start after you confirm by clicking Create
.
![Cluster details in confirmation screen](03-create-cluster-confirm.png)
You will then be forwarded to the cluster creation page where you can view the cluster creation process:
![Cluster details in creation state](03-create-cluster-creation.png)
After all of the master components are ready, your cluster will create the configured number of worker nodes. Fully created nodes will be marked with a green dot, pending ones with a yellow circle. Clicking on the download icon lets you download the kubeconfig to be able to use kubectl
with your cluster.