Skip to content
Cloud Provider Benchmarks

Reports · Kubernetes

Managed Kubernetes 2026: control planes, autoscaling and storage across seven clouds

Documentation comparison of managed Kubernetes on seven clouds: control plane, node autoscaling, LoadBalancer services and storage classes.

Direct answer

AWS, PortvMind Cloud, Google Cloud, Azure, DigitalOcean and Vultr document managed Kubernetes with autoscaling, load balancers and storage classes. Hetzner does not operate the control plane; customers run their own clusters against its cloud controller and CSI driver.

Executive summary

Six of the seven tracked providers document a managed Kubernetes control plane with node autoscaling, automatic LoadBalancer services and a storage class: AWS (EKS), PortvMind VKE, Google Cloud (GKE), Microsoft Azure (AKS), DigitalOcean (DOKS) and Vultr Kubernetes Engine. Hetzner documents no managed control plane. Its cloud controller, CSI driver and the upstream cluster autoscaler apply only to clusters the customer operates, and those three capabilities are therefore partial. This edition does not report cluster-creation time, node-join time or conformance results.

Key findings

  1. 01 A managed control plane is documented for six providers and absent for Hetzner.
  2. 02 Where the control plane is managed, node autoscaling, LoadBalancer integration and a storage class are also documented as available.
  3. 03 Hetzner's cloud controller, CSI driver and upstream autoscaler are partial: they exist for self-managed clusters only.
  4. 04 PortvMind VKE additionally documents the Kubernetes version chosen at creation, public and internal load balancers, and PVC backup through its Backup service.
  5. 05 Creation time, node-join time and Sonobuoy conformance are defined in the methodology.

1. Scope

This report compares managed Kubernetes as documented by AWS, PortvMind Cloud, Google Cloud, Microsoft Azure, DigitalOcean, Vultr and Hetzner. The criteria are the four used everywhere on this site: a provider-operated control plane, node autoscaling within configured limits, automatic provisioning of a load balancer for Services of type LoadBalancer, and dynamic persistent volumes through a storage class. Method: managed Kubernetes methodology. Timed measurements from that method are not in this edition.

2. Capability comparison

Managed Kubernetes capabilities, all providers
Criterion PortvMind AWS Azure Google Cloud DigitalOcean Vultr Hetzner
Kubernetes
Managed Kubernetes Provider-operated Kubernetes control plane. Yes Yes Yes Yes Yes Yes No CCM and CSI driver for self-managed clusters only.
Kubernetes node autoscaling Worker node count scales automatically within configured limits. Yes Yes Yes Yes Yes Yes Partial Through the upstream cluster autoscaler on self-managed clusters.
Kubernetes LoadBalancer integration Services of type LoadBalancer provision a provider load balancer automatically. Yes Public and internal load balancers. Yes Yes Yes Yes Yes Partial Via the Hetzner cloud-controller-manager on self-managed clusters.
Kubernetes storage classes (CSI) Persistent volumes provisioned dynamically through a provider storage class. Yes Yes Yes Yes Yes Yes Partial Via the Hetzner CSI driver on self-managed clusters.

3. Providers with a managed control plane

AWS, Google Cloud and Microsoft Azure. EKS, GKE and AKS are documented managed Kubernetes services. On the four criteria above, each is assessed yes.

DigitalOcean and Vultr. DOKS and Vultr Kubernetes Engine are documented managed services. Node autoscaling, LoadBalancer integration and a storage class are assessed yes for both.

PortvMind VKE. Assessed yes on the same four criteria. The service page also documents the version chosen at creation, public and internal load balancers, and PVC backup. Regions and list prices are in Cloud services, regions and list prices 2026.

4. Hetzner: self-managed clusters only

Hetzner is assessed no for a managed control plane. The published integration is a cloud controller manager and a CSI driver that a customer attaches to a cluster they run themselves. Node autoscaling through the upstream cluster autoscaler, LoadBalancer services through that controller, and volumes through the CSI driver are therefore partial, not a managed service. A buyer who needs the provider to operate the control plane should not treat these partial values as equivalent to EKS, GKE, AKS, DOKS, Vultr Kubernetes Engine or VKE.

5. What this edition does not measure

The methodology defines cluster creation time (API call to three Ready nodes), node-join time when a pool grows from three to six nodes, LoadBalancer address time, Sonobuoy certified conformance, and kube-burner API latency. This edition does not rank providers by speed. A timing is added once it has been repeated within tolerance.

6. Conclusion

On documented capability, six providers offer managed Kubernetes with autoscaling, load balancers and storage classes, and Hetzner does not. A later edition can add timings once they have been repeated within tolerance. Buyers comparing PortvMind VKE with the hyperscalers should read the VKE page for version selection, internal load balancers and PVC backup, which the four-row matrix does not show on its own.

Limitations

  • Assessments are documentation-based. No cluster was created for this edition.
  • Kubernetes version ranges, pod limits, CNI choices and price per control plane are not compared.
  • Hetzner partial values describe self-managed clusters, not a provider-operated control plane.

Version history

Version Date Notes
1.0 First edition. Capability comparison only; no timed runs.

Frequently asked questions

Which tracked clouds do not offer managed Kubernetes?

Hetzner. It documents a cloud controller and a CSI driver for clusters the customer operates, not a managed control plane.

Are there timing results in this report?

No. Cluster creation, node join and conformance are defined in the methodology and can appear in a later edition.

Sources

  1. [1] Amazon EKS — AWS Official documentation accessed:
  2. [2] Google Kubernetes Engine — Google Cloud Official documentation accessed:
  3. [3] Azure Kubernetes Service — Microsoft Official documentation accessed:
  4. [4] DigitalOcean Kubernetes — DigitalOcean Official documentation accessed:
  5. [5] Vultr Kubernetes Engine — Vultr Official documentation accessed:
  6. [6] Hetzner Cloud — Hetzner Official documentation accessed:
  7. [7] PortvMind VKE Kubernetes — PortvMind Provider (first-party) accessed: