raff_k8s_cluster

_{Updated August 25, 2026} Manages a managed Kubernetes cluster. Provisioning takes 10–20 minutes; the resource waits until the cluster is running (create timeout 30 minutes, delete 20). The cluster’s first pool is defined inline as default_pool; additional pools use raff_k8s_node_pool. On a subscription account the create charges the saved payment method for the monthly subscription automatically; pay-as-you-go accounts bill hourly.

Example

data "raff_k8s_node_plans" "all" {}
data "raff_k8s_versions" "all" {}

resource "raff_k8s_cluster" "prod" {
  name       = "prod-cluster"
  ha_enabled = true

  default_pool {
    name       = "default-pool"
    plan_id    = data.raff_k8s_node_plans.all.plans[0].id
    node_count = 3
  }

  traefik_enabled = true
}

# Write the kubeconfig for kubectl / the kubernetes and helm providers
resource "local_sensitive_file" "kubeconfig" {
  content         = raff_k8s_cluster.prod.kubeconfig
  filename        = "${path.module}/kubeconfig.yaml"
  file_permission = "0600"
}

Argument reference

Argument	Type	Required?	Description
`name`	string	Yes	Cluster name (1–63 lowercase letters, digits or hyphens, unique per account). Renames in place
`default_pool`	block	Yes	The cluster’s first node pool: `name`, `plan_id` (from `raff_k8s_node_plans`), `node_count` (2–20)
`k8s_version_id`	int	No	Kubernetes version ID from `raff_k8s_versions`. Defaults to the platform default. A version newer than the base image is reached automatically during creation
`ha_enabled`	bool	No	HA control plane (3 masters + redundant gateway, flat $30/month). Upgradable in place `false` → `true`; downgrading is not supported
`traefik_enabled`	bool	No	Install the Traefik ingress controller
`metallb_enabled`	bool	No	Install MetalLB for `LoadBalancer` services
`storage_node_count`	int	No	Dedicated Longhorn storage nodes (0, 2 or 3). 0 disables in-cluster block storage
`storage_node_disk_gb`	int	No	Data disk per storage node in GB (10–1000). Required when `storage_node_count` > 0
`cluster_cidr`	string	No	Pod network CIDR. Defaults to `10.42.0.0/16`
`service_cidr`	string	No	Service network CIDR. Defaults to `10.43.0.0/16`

Attribute reference (computed)

Attribute	Description
`id`	Cluster UUID
`cluster_id`	Short cluster ID used in URLs, the API endpoint hostname, and `raff_k8s_node_pool.cluster_id`
`api_endpoint`	`https://<cluster-id>.k8s.raffusercloud.com:6443`
`kubeconfig`	Admin kubeconfig YAML (sensitive) — feed it to the `kubernetes`/`helm` providers or write it to a file
`k8s_version`	The running Kubernetes version (full, e.g. `v1.35.7+rke2r1`)
`status`	Cluster status
`price_per_month`	Monthly price of the committed node set

Import

terraform import raff_k8s_cluster.prod <cluster-id>

Getting Started

Guides

REST API

CLI

Terraform

raff_k8s_cluster

Example

Argument reference

Attribute reference (computed)

Import

​Example

​Argument reference

​Attribute reference (computed)

​Import

Example

Argument reference

Attribute reference (computed)

Import