Containerspace Cluster Lifecycle

Containerspaces integrate with Rancher to provide managed Kubernetes clusters on the Pilvio.pro platform. This page covers the management cluster's redundancy setting, node pool management, ingress, and cluster upgrades.

For getting started with Containerspaces, see Containerspaces.

Management Cluster Type: Single vs Highly Available

When creating a Containerspace (a Rancher management cluster), you choose how the management cluster itself is deployed:

Type	Description
Single	The management cluster runs on a single virtual machine.
Highly available	The management cluster runs on three virtual machines, grouped in an anti-affinity group.

Tip

Highly available increases the resource cost of the management cluster (roughly 3x compute and storage compared to Single). This setting is about the Rancher management cluster's own resilience — it's independent from how many master node pools you configure for a Kubernetes cluster running inside it (see below).

Creating a Containerspace

Creating a Containerspace only provisions the Rancher management cluster — it does not create a Kubernetes cluster by itself.

1. Navigate to Containerspaces.
2. Click CREATE MANAGEMENT CLUSTER.
3. Configure the cluster (name, Rancher version, cloudspace, external network IP, cluster type), domain, and certificate — see Containerspaces for the full wizard walkthrough.
4. Wait for the management cluster to reach the Running state.
5. Click CREATE KUBERNETES CLUSTER on the management cluster's details page to provision an actual Kubernetes cluster inside it — this is a separate step with its own node pool and Kubernetes version configuration.

Managing Node Pools

A node pool is a group of nodes (master or worker) sharing the same configuration. The Master and Worker node pools are first configured as part of creating the Kubernetes cluster; once created, you can add further pools, scale existing ones, or delete them independently.

Adding a Node Pool

1. Navigate to the Kubernetes cluster's Node Pools page.
2. Click Create node pool.
3. Configure:
4. Name
5. Cloudspace — which cloudspace to deploy nodes into
6. Master node toggle — on for a master pool, off for a worker pool
7. Memory and vCPUs per node
8. Node count
9. Optionally, Pin vCPUs (billed at a higher rate); for master pools, Physical storage
10. Click Add node pool.

Note

There's no per-node-pool Kubernetes version or network field — the Kubernetes version is set once for the whole cluster when it's created, and nodes use the pool's cloudspace directly.

Scaling a Node Pool

1. Navigate to the Node Pools page, or the node pool's own details page.
2. Use the inline +/- controls, or open the pool and set a New size.
3. Click Scale.

Deleting a Node Pool

1. Navigate to the node pool's details page.
2. Click DELETE NODE POOL, then confirm.

All VMs in the pool are deleted and deregistered from the cluster.

Ingress on Worker Node Pools

There's no automatic, zero-configuration load-balancer controller for Kubernetes Services or Ingress resources. Instead, each worker node pool has a manual, opt-in ingress controller toggle:

• By default, an Nginx ingress controller is installed for each worker node pool at cluster creation.
• From the node pool's details page, use Install ingress controller or Uninstall ingress controller to change this after the fact.
• The node pool's details show the ingress class name, controller status, TLS issuer, and load balancer IP once installed.

Cluster Upgrades

The Kubernetes version is an inline editable field on the Kubernetes cluster's details page:

1. Navigate to the Kubernetes cluster details page.
2. Find the version field — if a newer compatible version is available, it's selectable there.
3. Choose the target version to start the upgrade.

The Rancher version of the management cluster itself can be upgraded the same way, from the management cluster's own details page.

Deleting a Containerspace

1. Navigate to the management cluster's details page.
2. Click DELETE MANAGEMENT CLUSTER.
3. The confirmation page lists everything that will be deleted with it: Kubernetes clusters, VMs, DNS records, and Ingress resources.
4. Type the cluster's name to enable the confirm button, then confirm.

This destroys the entire management cluster, including every Kubernetes cluster, node, and workload inside it.

Management clusters cannot be restored from the Recycle Bin

There is no Containerspace/Management-Cluster tab in the Recycle Bin, so a management cluster itself cannot be restored from the Recycle Bin. Only the underlying VMs may appear in the VM recycle bin.