Getting Started

Log in to the PCD UI and click on the Kubernetes menu entry:

If you have logged in for the first time, you will see a getting-started page. Click on the Create New Cluster button to start creating your first cluster.

The cluster creation process provides you with a form to provide the details of your Kubernetes cluster. Let's look at those details:

• Cluster Architecture

  • This section provides details on the resource capacity available in your tenant, which will help you estimate the number of Kubernetes Worker node VMs you can create.

• Server, Storage, and Network

  • This section lets you configure your node groups. Node groups are logical entities that let you create a template for a set of Kubernetes worker nodes. Each node group can be configured to have a defined number of nodes with Networking and Storage properties.

    • Note: This release of PCD uses the default Storage configured for the VMs in the PCD. Storage configuration will be enabled in the future.

  • Auto-scaling: PCD deploys a Kubernetes Cluster auto-scalar for each cluster in PCD management plane. You can enable/disable and configure the maximum number of nodes to which the node group can scale.

• Configure your Cluster

  • Choose the Kubernetes Version for your cluster.

  • Choose which add-ons you want to enable or disable. PCD Kubernetes supports 2 types of add-ons

    • Core Add-ons

      • The add-ons are critical to the functionality of your Kubernetes cluster, including CNI, CoreDNS, and more.

    • Other Add-ons

      • These are add-ons supported and managed by PCD to extend Kubernetes capabilities and provide services such as load balancing.

      • This release of PCD Kubernetes Clusters supports only MetalLB.

      • Note: PCD only deploys MetalLB controllers on the Kubernetes Cluster, and you still have to configure MetalLB after the cluster is created.

Provide a name for the cluster, then click the Deploy Cluster Button. It usually takes 5-10 minutes to bring the complete cluster up, but it can take longer based on the number of node groups and worker nodes configured.

You can select a cluster from the Clusters list to dive deeper into its details.