aws AWS What's New ·

Amazon ECS enables faster service auto scaling with high-resolution metrics

infraawsgaengineer
feature patch

Amazon ECS service auto scaling now supports high-resolution (20-second) metrics, significantly reducing scale-out and provisioning times. This enhancement allows for reduced baseline capacity and lower compute costs while maintaining service reliability. The feature is available for CPU and memory utilization target tracking policies and is accessible across all AWS regions and compute options.

  • Faster ECS service auto scaling with 20-second metric resolution
  • Benefits of faster scaling and configuration
  • Availability and pricing
Enhancements (1)
  • Faster ECS service auto scaling with 20-second metric resolution

    Amazon ECS service auto scaling now leverages high-resolution metrics (20-second intervals) for CPU and memory utilization target tracking policies, improving scale-out and provisioning times. Benchmarking shows a 76% reduction in scale-out trigger time and a 72% reduction in total scale and provision time.

Notes (2)
  • Benefits of faster scaling and configuration

    The enhanced scaling capabilities enable lower compute costs and maintained service reliability. Users can configure 20-second resolution via the AWS Console, CLI, CloudFormation, or AWS SDKs when setting up target tracking policies.

  • Availability and pricing

    This feature is available in all AWS commercial and AWS GovCloud (US) regions for AWS Fargate, Amazon ECS Managed Instances, and Amazon EC2. Standard CloudWatch charges apply for high-resolution metrics.

Read the original announcement →

https://aws.amazon.com/about-aws/whats-new/2026/06/amazon-ecs-faster-autoscaling/