Why it matters
Instances are often sized for a peak that never comes or chosen before the workload was known. Rightsizing removes capacity you pay for but do not use.How nOps finds it
This recommendation comes from AWS Compute Optimizer, not from nOps. Compute Optimizer looks at the instance’s CloudWatch metrics over its lookback period (14 days by default) and recommends a type that would still meet the workload’s needs. Memory is only considered when the CloudWatch agent reports it.Before you start
AWS Compute Optimizer enabled for the account. Installing the CloudWatch agent makes memory part of the analysis, which makes the recommendation more reliable.How to implement
- Review the recommended instance type and confirm it supports what the workload needs (architecture, network and storage features).
- Test the new type outside production if you can.
- Stop the instance, change its instance type (Actions → Instance settings → Change instance type), then start it.
- For instances in an Auto Scaling group, change the launch template instead.
- Update Terraform or CloudFormation so a later deployment does not change it back.
Risks and considerations
- Changing the instance type needs a stop and start, so plan for downtime or do it one instance at a time.
- A smaller type has less headroom: check peaks outside the lookback window, such as month-end jobs.
- Reserved Instances or Savings Plans bought for the old type may cover the new one differently.
FAQs
Why doesn't the recommendation consider memory?
Why doesn't the recommendation consider memory?
EC2 does not report memory to CloudWatch on its own. Install the CloudWatch agent so Compute Optimizer can include it.