What does DigitalOcean charge for the DOKS control plane?

The standard DOKS control plane is free; you pay only for the worker Droplets, persistent volumes, and load balancers. DigitalOcean also offers an HA control plane upgrade at additional cost, which provides multi-zone redundancy for the API server and etcd — appropriate for production clusters where control plane availability is non-negotiable.

How does autoscaling work on DOKS worker pools?

Each worker pool has a min and max node count. The DOKS cluster-autoscaler watches for unschedulable pods and adds nodes within the bounds when capacity is needed. It removes nodes when they have been underutilized for the cool-down period (default 10 minutes) and their pods can be rescheduled. Setting min equal to max effectively disables autoscaling for that pool.

DO Kubernetes Cluster Builder

ComputeDigitalOcean

Build DigitalOcean Kubernetes (DOKS) cluster configurations with node pools and auto-scaling.

Last verified: May 2026

DigitalOcean Kubernetes (DOKS) Configuration

Build DigitalOcean Kubernetes (DOKS) cluster configurations with node pools, auto-scaling, HA control plane, and maintenance windows.

Required Fields

nameregionversionnode_pools

Generated Output

Output will appear here...

See It in Action

Your team's first DOKS cluster has been running everything on a single auto-scaling pool of `s-2vcpu-4gb` nodes. The CoreDNS pods keep getting evicted when the cluster autoscaler reclaims nodes, causing intermittent DNS failures across the cluster. You use the builder to add a second pool — three fixed `s-4vcpu-8gb` nodes with taints that limit it to system pods — and rerun Terraform. CoreDNS settles on the dedicated pool, the workload pool autoscales freely, and the DNS flapping stops.

What This Tool Does

DigitalOcean Kubernetes (DOKS) is a managed control plane with worker pools you size and configure. The DO Kubernetes Cluster Builder generates a complete cluster spec — region, version, networking (VPC), worker pools with autoscaling bounds, surge upgrade settings, and maintenance window — and validates the combination against current DOKS constraints. The output is Terraform-ready and matches the parameters expected by `digitalocean_kubernetes_cluster`.

Technical Details

The builder accepts cluster name, region, Kubernetes version, VPC selection, maintenance window, and one or more worker pool specs (size, count or min/max for autoscaling, optional taints and labels). It validates the version against currently-supported DOKS versions, checks the chosen Droplet size against DOKS-eligible sizes, and emits a `digitalocean_kubernetes_cluster` Terraform block including one or more `node_pool` blocks.

Common Use Cases

1Standing up a fresh DOKS cluster with autoscaling worker pools sized for a known initial workload and a clear upper bound.
2Adding a second worker pool optimized for a specific workload (GPU, memory-heavy batch jobs) to an existing cluster.
3Generating a reproducible cluster spec for an internal platform team that needs to spin up DOKS clusters on demand.
4Comparing the cost of different worker pool configurations (5x small vs. 2x large) before committing to a cluster shape.

Common Questions

What does DigitalOcean charge for the DOKS control plane?: The standard DOKS control plane is free; you pay only for the worker Droplets, persistent volumes, and load balancers. DigitalOcean also offers an HA control plane upgrade at additional cost, which provides multi-zone redundancy for the API server and etcd — appropriate for production clusters where control plane availability is non-negotiable.
How does autoscaling work on DOKS worker pools?: Each worker pool has a min and max node count. The DOKS cluster-autoscaler watches for unschedulable pods and adds nodes within the bounds when capacity is needed. It removes nodes when they have been underutilized for the cool-down period (default 10 minutes) and their pods can be rescheduled. Setting min equal to max effectively disables autoscaling for that pool.

Expert Tips

TIP

Use multiple worker pools to separate workload types. A small pool of cheap nodes for system pods (DNS, ingress controller) and a separate autoscaling pool for application workloads gives you predictable system performance and tunable elasticity.

TIP

Enable surge upgrade for production clusters. Without it, node upgrades cordon-and-drain one at a time, which can stall under PodDisruptionBudgets and leave the cluster in a half-upgraded state. Surge upgrade brings up new nodes first and gracefully migrates workloads.

Was this tool helpful?

Disclaimer: This tool runs entirely in your browser. No data is sent to our servers. Always verify outputs before using them in production. AWS, Azure, and GCP are trademarks of their respective owners.