Running Karpenter and StormForge on EKS

In this video, you’ll learn how to run Karpenter and StormForge on EKS. StormForge takes care of the application stack scaling and Karpenter takes care of the Worker Node scaling.

Optimize Both Sides of the Kubernetes Equation

Karpenter and StormForge address different layers of Kubernetes resource optimization.

StormForge optimizes workloads
StormForge analyzes workload behavior and provides optimized CPU and memory resource settings, helping applications run efficiently without unnecessary overprovisioning or reliability risk.

Karpenter optimizes infrastructure
Karpenter dynamically provisions and removes nodes based on workload requirements, helping ensure your cluster has the right compute capacity when it needs it.

Together, they create a complementary optimization model: StormForge helps determine what workloads need. Karpenter helps provide the infrastructure to run them.

What You’ll See in the Demo

Follow along as the cluster is connected to StormForge and prepared for workload optimization.

• Connect an existing EKS cluster running Karpenter to StormForge
• Install StormForge using Helm
• Verify the StormForge installation
• Begin collecting workload utilization data
• Use that data to optimize CPU and memory resources
• Combine workload optimization with Karpenter’s node-level scaling

From Workload Demand to Infrastructure Capacity

The effectiveness of node-level optimization depends in part on the resource requirements of the workloads running on those nodes.

StormForge uses actual workload behavior to optimize Kubernetes CPU and memory settings. Karpenter can then use those workload requirements when determining the compute capacity to provision.

The result is an optimization loop that addresses efficiency at both layers:

StormForge → Optimize workload resources → Karpenter → Provision appropriate nodes

You can try it yourself with our free trial.