Warm pools maintain pre-initialized EC2 instances that have completed OS and software setup, allowing rapid transition to active use upon demand spikes. This accelerates node provisioning for workloads with burst traffic or lengthy initialization, improving responsiveness and reducing cold-start delays in managed Kubernetes environments.
Users can configure warm pools with instances either stopped (cost-efficient but slower startup) or running (faster startup), and can enable reuse of instances on scale-in. This integration works out of the box with Cluster Autoscaler, offering seamless operational improvement across most AWS regions providing EKS.
Source & further reading
Amazon EKS managed node groups now support EC2 Auto Scaling warm pools ↗Source published 2026-04-08. This report was prepared with AI assistance from the linked primary-source material. Claims and limitations should be checked against the original announcement or research.