EKS control plane parameters are now tunable!

Hey everyone, it's Shiichan! Here's an update Kubernetes operators won't want to miss!

AWS What's New aws.amazon.com

What was announced?

On AWS What's New, it was announced that Amazon EKS now supports configuring parameters for Kubernetes control plane components, including the scheduler, controller manager, and API server.

Why it matters

Until now, the EKS control plane mostly ran on its defaults. Now administrators get parameter-level control over its behavior, which matters most for teams running clusters at scale who want to tune things to fit their workload patterns.

What changes

Here's what you can now configure:

• Pod placement strategies, to improve resource utilization
• How quickly horizontal pod autoscaling (HPA) responds to demand changes
• Resource lifecycle parameters, such as event retention duration

Cluster administrators now have real control over control plane behavior beyond the defaults.

Dive Deep

The example in the announcement is a good illustration. Setting the scheduler's node resource fit strategy parameter to MostAllocated packs pods onto nodes that are already well utilized, letting you run the same workloads on fewer nodes.

The default LeastAllocated strategy does the opposite, spreading pods across nodes. You can keep that where headroom matters more than density. Whether you want to prioritize density for cost efficiency, or headroom for stability, you can now choose based on what your cluster actually needs.

This is available in all AWS Regions where Amazon EKS is available. For the full list of configurable parameters, check Control plane configuration in the Amazon EKS User Guide.

Wrap-up

• Amazon EKS now supports configuring the scheduler, controller manager, and API server
• You can tune pod placement strategy, HPA responsiveness, and event retention, among other parameters
• Example: setting the scheduler to MostAllocated packs pods densely to reduce node count, while the default LeastAllocated favors spreading pods out
• Available in all AWS Regions where Amazon EKS is available

This one's for infrastructure teams running EKS at scale who want another lever for cost efficiency or stability!