How Hyperkub is organised

Everything in Hyperkub — the UI, the CLI and the API — works on the same small set of resources. Learning them once explains all three.

graph TD
    A[Account] --> C[Cluster]
    A --> K[API Key]
    A --> M[Members]
    C --> P[Cluster Pool]
    P --> N[Cluster Node]
    C --> D[Cluster Domain]
    C --> R[Cluster Credential]
    C --> RG[Region]

Account

The billing and ownership boundary. Clusters belong to an account, invoices are issued against it, and members are invited into it.

Most resources carry an account_id, and list endpoints accept it as a filter.

Cluster

A Kubernetes cluster: a managed control plane, pinned to a region and a Kubernetes version. Creating one is asynchronous — the API returns immediately with status provisioning, and the cluster becomes usable when it reaches running.

Status	Meaning
provisioning	Being built. Not yet reachable.
running	Healthy and serving the Kubernetes API.
shutdown	Intentionally stopped.
error	Provisioning or runtime failure. Contact support.

Cluster pool

A group of identically-sized nodes inside a cluster. Pools are the unit of scaling: you change a pool’s node count rather than creating nodes directly.

Use more than one pool when workloads need different machine sizes — for example a small pool for web traffic and a large one for batch jobs.

Cluster node

A single virtual machine in a pool. Nodes are created and replaced by the platform. You can list them and trigger actions on individual nodes, but you normally operate at the pool level.

Cluster credential

A short-lived, downloadable kubeconfig for a cluster. Credentials are issued individually so they can be revoked individually — if a laptop is lost, delete that one credential instead of rotating access for everyone.

See Connecting with kubectl.

Cluster domain

A custom hostname pointed at a cluster’s ingress. Each cluster has a limit on how many custom domains it can have, reported as limit_max_custom_domains.

Region

Where a cluster physically runs. Regions determine latency and available machine sizes, and are chosen at creation time — a cluster cannot be moved between regions afterwards.

API key

A credential for the Hyperkub API, scoped to an account. This is separate from a cluster credential: an API key manages your clusters, a cluster credential manages what runs inside one.