Managed Kubernetes Reference

Provider kinds, API endpoints, and provider-specific option blocks for Ankra-managed cloud Kubernetes (DOKS, UKS, GKE, OVHcloud MKS, AKS, EKS).

Reference material for Managed Kubernetes.

---

Provider kinds

The {provider} path segment in every endpoint below is one of:

Provider kind	Service	Location field
doks	DigitalOcean Kubernetes	region
uks	UpCloud Managed Kubernetes	zone
gke	Google Kubernetes Engine	location (zone or region)
ovh_mks	OVHcloud Managed Kubernetes	region
aks	Azure Kubernetes Service	location
eks	Amazon EKS	region

The CLI accepts these values via --provider (ovh-mks and mks are accepted as aliases for ovh_mks).

---

API Reference

All endpoints are bearer-token authenticated under /org/clusters/managed/{provider}.

Endpoint	Method	Description
/org/clusters/managed/{provider}/options	GET	Live regions, versions, sizes, plans, and capabilities for a credential_id
/org/clusters/managed/{provider}/discover	GET	Discover existing clusters at the provider for a credential_id
/org/clusters/managed/{provider}/preflight	POST	Run provider preflight checks without creating
/org/clusters/managed/{provider}/import	POST	Adopt a discovered cluster into Ankra
/org/clusters/managed/{provider}	POST	Create a managed cluster
/org/clusters/managed/{provider}/{cluster_id}	DELETE	Deprovision a managed cluster (?force=true to bypass idle checks)
/org/clusters/managed/{provider}/{cluster_id}/node-pools	GET	List node pools
/org/clusters/managed/{provider}/{cluster_id}/node-pools	POST	Add a node pool
/org/clusters/managed/{provider}/{cluster_id}/node-pools/{name}/scale	POST	Scale a node pool to a fixed count
/org/clusters/managed/{provider}/{cluster_id}/node-pools/{name}	PATCH	Update a node pool (count and/or autoscaling bounds)
/org/clusters/managed/{provider}/{cluster_id}/node-pools/{name}	DELETE	Delete a node pool
/org/clusters/managed/{provider}/{cluster_id}/node-pools/{name}/replace	POST	Replace a pool’s machine size blue/green (create new pool, wait ready, delete old)
/org/clusters/managed/{provider}/{cluster_id}/upgrades	GET	List available Kubernetes upgrade targets
/org/clusters/managed/{provider}/{cluster_id}/upgrade	POST	Upgrade the Kubernetes version
/org/clusters/managed/{provider}/{cluster_id}/stop	POST	Stop the cluster (providers advertising supports_stop_start; AKS today)
/org/clusters/managed/{provider}/{cluster_id}/start	POST	Start a stopped cluster

The ankra cluster managed CLI covers create, delete, node-pool add|scale|update|delete, upgrade, and stop/start. Live options, discovery, import, preflight, and pool size replacement are portal/API only.

Managed node pool machine sizes are immutable at every provider, so changing a pool’s size is a blue/green replacement: Ankra creates a pool with the new size (copying count, labels, taints, and autoscaling), waits until it is ready, then deletes the old pool. Expect transient double capacity while both pools exist; workloads are rescheduled by Kubernetes when the old pool is removed. Whether a provider supports stop/start is advertised per provider as supports_stop_start in the /options capabilities.

---

Provider-specific option blocks

Provider	Block	Fields
DOKS	top level	ha - highly available control plane
UKS	top level	cluster_plan (development / production); GPU pools take gpu_plan.storage_size
GKE	gke	release_channel (RAPID, REGULAR, STABLE), mode (standard or autopilot)
OVH MKS	ovh_mks	plan (free, standard), private_network_id, anti_affinity, monthly_billed, update_policy
AKS	aks	resource_group, network_plugin (azure, kubenet), private_cluster, sku_tier (Free, Standard, Premium)
EKS	eks	subnet_ids, security_group_ids, enable_control_plane_logging