Skip to main content
AWS Compute Optimizer analyzes the CPU, memory, network and disk use of the instances in an Auto Scaling group and recommends an instance type that would fit the workload at a lower cost. nOps shows these recommendations with AWS’s estimated monthly saving.

Why it matters

Instance types for Auto Scaling groups are often chosen once and copied between environments. A better-fitting type lowers the cost of every instance the group runs.

How nOps finds it

This recommendation comes from AWS Compute Optimizer, not from nOps. Compute Optimizer looks at the group’s CloudWatch metrics over its lookback period and recommends a type that would still meet the workload’s needs. Memory is only considered when the CloudWatch agent reports it. nOps imports the recommendation daily from AWS Cost Optimization Hub, which returns one recommendation per resource (the one with the highest saving), and shows AWS’s estimated monthly saving, which already takes your discounts into account.

Before you start

AWS Compute Optimizer and AWS Cost Optimization Hub enabled for the account (or the whole organization). The CloudWatch agent on the instances makes memory part of the analysis.

How to implement

  1. Review the recommended type and check it supports what the workload needs (architecture, network and storage features).
  2. Create a new launch template version with the recommended instance type.
  3. Point the Auto Scaling group at the new version and start an instance refresh (aws autoscaling start-instance-refresh).
  4. Watch CPU, memory and latency during and after the refresh, and check your scaling policies still behave as expected.
  5. Update the launch template in Terraform or CloudFormation.

Risks and considerations

  • Smaller instances can change how scaling policies react, because each instance carries more of the load.
  • Check peaks outside the lookback window, such as month-end jobs.
  • To roll back, point the group back at the previous launch template version and refresh.