Containerspaces¶

A Containerspace is a container management platform that provides our customers with the tools and services they need to deploy, manage, and maintain their Kubernetes clusters, containerized applications and infrastructure.

With Rancher integration, containerspaces provides our customers with a perfect automated way to manage Kubernetes clusters on whitesky.cloud NV portal.

Our integration lets you seamlessly deploy all the components of the Kubernetes management environment in whitesky.cloud NV portal. At the same time, you stay in control of the configuration and the physical locations of the workloads.

Creating Containerspaces¶

Open Containerspaces from the navigation drawer.
Click CREATE MANAGEMENT CLUSTER in the top bar. A management cluster creation wizard will appear.
In the Basic configuration step, enter the following cluster information and click NEXT:
- Name of the cluster.
- Rancher version — pick from the available versions. The supported Kubernetes versions for the selected Rancher version are shown below the field.
- Cloudspace where RKE nodes should be provisioned.
- Service account (required) — see Service Accounts below.
- External network IP address that will be used to publicly access the RKE dashboard. The external IP of the linked cloudspace is used by default.
- Optionally, check Physical storage to use flash-based storage local to the hypervisor instead of the default software-defined storage.
In the Domain configuration step, add your domain configuration and click NEXT. Choose one of the options:
- Create new domain. Add a subdomain under one of the top-level domains registered for your customer. This will automatically create a DNS record for your subdomain. See DNS documentation for more information on the DNS feature.
- Your own domain. Provide a domain name previously registered in a registrar of your choice.
In the Certificate step, configure an SSL certificate for your domain. Choose one of the certificate options (a radio choice, not a checkbox):
- Let's Encrypt — request an SSL certificate from Let's Encrypt.
- Certificate Store — use your own certificate. SSL certificates can be added by the Customer admins under Admin > Certificate Store > Add Certificate. Then choose the cluster's size:
- Single — the management cluster runs on one virtual machine.
- Highly available — three virtual machines grouped in an anti-affinity group. Highly available increases the resource cost accordingly.
Click CREATE. A dialog window will open to confirm that the Rancher management cluster is being deployed. You can follow the deployment logs published on the portal.

Service Accounts¶

Containerspaces run their components (the management cluster, the CSI driver, and the Rancher agents) under a service account rather than a user. The service account's permissions determine what the containerspace can do against the whitesky.cloud NV portal APIs, so the containerspace inherits the access rights of the service account assigned to it.

Required permissions¶

The service account must have full access (read, create, delete, execute, and update) on the location of the cloudspace where the cluster is deployed. See Service Accounts for how to create one and attach roles.

Selecting a service account when creating a containerspace¶

When you create a management cluster, the creation wizard includes a required Service account field:

In the cluster creation wizard, select a Service account from the dropdown. The dropdown shows each service account's name and its attached roles.
If no service accounts exist yet, use the CREATE A SERVICE ACCOUNT button to create one.
Continue with the rest of the wizard as usual.

The selected service account is used to run all of the containerspace's components.

Changing the service account on an existing containerspace¶

You can change the service account on an already-deployed management cluster at any time. The platform then automatically rotates the credentials to every component of the cluster — you do not need to touch any VM, secret, or API call manually.

Open the details page of your management cluster under Containerspaces.
In the Service account row, click the pencil icon.
Select the new service account and click Update.
The dialog switches to a Credential rotation view that shows the progress of pushing the new credential to each target:
VMs in the cluster
Kubernetes clusters
Rancher agents

Each target shows its status (In progress, Succeeded, or Failed). If any target fails, you can Retry that item individually or Retry all.

Note: Changing the service account requires the new account to have full access on the cluster's location. After saving, the new credential is pushed to all cluster VMs; a brief service interruption is possible.

If a rotation is in progress or has failed, a "Credential rotation in progress or failed — view status" link appears next to the service account name on the details page so you can reopen the rotation view.

Containerspaces details page¶

Go to Containerspaces section in the navigation drawer to list your containerspaces. Click on a specific containerspace to navigate to its details page. On the details page, you can:

See the management cluster's general information.
Manage Kubernetes clusters, cluster nodes, provisioned nodes
Download the KubeConfig file of the management cluster

Creating Kubernetes cluster¶

Open the details page of your new containerspace on whitesky.cloud NV portal.
Click CREATE KUBERNETES CLUSTER at the top bar.
Enter your Master and Worker node pools information and click NEXT.
- Name of your Master/Worker nodes.
- Cloudspace where the nodes will be attached.
- Memory size of your Master/Worker VMs.
- Node Count in your node Pools.
- vCPUs used by the nodes VMs
Tip

Note: You can see that it's allowed to add multiple master node pools in case our customers want to have a georedundant K8s cluster. You can have three master node pools and deploy one node in every master node pool; that will give you a georedundant K8s cluster.
Set your cluster name and select the Kubernetes version from the dropdown, then click CREATE. Both the cluster name and the Kubernetes version are required; if either is missing you'll see an error when you click CREATE. The cluster name must use lowercase letters, digits, and hyphens only.

Tip

Note: whitesky.cloud NV portal allows you to monitor the provisioning process. Once the process is finished, you can notice that the status of your Kubernetes cluster is changed from Provisioning to Active and the machine pools are changed to Running.

If you check your Kubernetes cluster details page, you can find the master and worker nodes deployed with the status Active.

Vacuum¶

whitesky.cloud NV portal allows you to remove all abandoned master and worker nodes that do not belong to a cluster anymore using a single click.

To vacuum your cluster:

Click VACUUM in the top bar of your management cluster details page.
Select the nodes that you want to delete.
Click VACUUM SELECTED NODES to continue.

Tip

Note: You must confirm this action, since the deleted nodes can't be restored.

Console¶

In the top bar, you find the CONSOLE that allows you to access your management cluster rancher dashboard.

You will need to log in to the dashboard through whitesky.cloud NV IAM, which supports two-factor authentication that makes your management cluster more secure.

The rancher dashboard allows you to manage your clusters by managing workloads, deploying applications, managing storage and more.

Cluster Nodes¶

Cluster Nodes lists all the VM nodes used within your management cluster.

Listed nodes are displayed with associated parameters like:

Parameters	Description
Health checks	Shows the status of health checks that run automatically to monitor your nodes' health.
Storage size	Storage size in GiB.
vCPUs	Number of assigned vCPUs.
Memory size	Memory size in MiB.
IP address	VM IP address.
Status	VM status whether RUNNING or HALTED.
Actions	Options that allow you to reboot the node or open the node console.

Provisioned Nodes¶

Provisioned Nodes lists all the master and worker nodes in your management cluster.

Listed nodes are displayed with associated parameters like:

Parameters	Description
Health checks	Shows the status of health checks that run automatically to monitor your nodes' health.
Cluster	Name of Kubernetes cluster where the node operates.
Virtual Machine	Shows the node VM name.
Cloudspace	Shows the Cloudspace name where the VM operates.
Actions	Options that allow you to reboot the node or open the node console.

Kube Config¶

whitesky.cloud NV portal provides the Kubernetes configuration file of the management cluster, which allows you to copy and paste it into any Kubernetes client like kubectl, K9s, ...etc.

You can find your KUBE CONFIG on the management cluster details page.

Updating the Rancher Version¶

The management cluster's own Rancher version can be upgraded separately from the workload Kubernetes cluster's version — for upgrading a Kubernetes cluster itself, see Cluster Upgrades.

On the management cluster details page, either edit the Version field directly in the details table, or use the dedicated update page.
Select a Rancher version from the dropdown. The Kubernetes versions supported by that Rancher version are shown as a hint below the field.
Click UPDATE to start the upgrade.

You can follow the upgrade's progress from the cluster's creation log (see below).

Management Cluster Creation Log¶

The management cluster details page keeps a log of every provisioning event — its initial creation, and any later transition such as a Rancher version upgrade.

Click CLUSTER CREATION LOGS in the top bar to open the most recent transition's log.
The log page shows the transition's Cluster status, Created at time, and Duration, followed by a detailed, timestamped log of each provisioning step.

While a transition is still in progress, the log refreshes automatically and shows End of execution once it completes.

Deleting containerspaces¶

To delete a containerspace

Navigate to the details page. At the top of the page, click DELETE MANAGEMENT CLUSTER.
The confirmation page lists everything that will be deleted along with the management cluster: its Kubernetes clusters, the VMs backing them, connected DNS records, and any Ingress resources (with links to their load balancers and server pools).
Type the cluster's name to enable the CONFIRM button, then click it.

Note: When you delete a management cluster, you automatically delete any Kubernetes cluster included in that management cluster, and all of the resources listed above.

Price Widgets¶

During management cluster creation, Kubernetes cluster creation, and node pool addition, a live price estimate panel is shown based on the selected configuration; the DETAILS button shows the cost breakdown.

Cluster Management and Cluster Dashboard Buttons¶

On the management cluster details page, Cluster Management and Cluster Dashboard buttons deep-link into the Rancher dashboard (machine pools page and cluster explorer respectively).

Unlimited Node Count (>1000)¶

When adding or scaling a node pool, an unlimited checkbox switches the node count input from a slider to a free-text number field, allowing counts above 1000.

Conditional Tab Visibility¶

On the management cluster details page, the Kubernetes Clusters, Provisioned Nodes, and Kube Config tabs are only shown when the cluster status is RUNNING. The Cluster Nodes tab is always visible.

Pending State Notes¶

When a cluster is in the pending state, the Node Pools and Kube Config tabs show a note: "will be available when the cluster is deployed and running."

Node Pool Nodes¶

Each node pool's details page has a Nodes section listing individual nodes (name, cloudspace, VM).

Node Pool Fields¶

The node pool details page also shows driver and hostname_prefix fields.

Per-Row Actions on Clusters List¶

The management clusters list has inline Delete and Console buttons per row (in addition to the actions on the details page).

Success Dialog¶

The cluster creation log page shows a success dialog when the deployment transition completes (polled every 20 seconds).