Getting Started
Compute Optimizer delivers intuitive and actionable recommendations to help you choose the optimal AWS resources for your workloads.
Page topics
Steps
Open allStep 1 - Choose Compute Optimizer in the AWS Management Console
 
                        Sign in to the AWS Management Console and click on [Compute Optimizer]. Click on “Get started.”  
                      
 
                     Step 2 - Opt in for Compute Optimizer
 
                        On the account setup page, review the default data settings and click “opt in.” Once you opt in, a service-linked role will be created automatically in your account.  
                      
 
                     Step 3 - Automatically generate recommendations
 
                        After you opt in, Compute Optimizer starts to scan your AWS infrastructure and generates recommendations. It may take up to 24 hours for Compute Optimizer to deliver recommendations for all supported AWS resources.  
                      
 
                     Step 4 - Define your recommendation preferences
 
                        Click on 'General' or 'Rightsizing' under Preferences on the left navigation bar to define and enable recommendation preferences, such as External Metrics Ingestion or Rightsizing recommendation preferences, based on your workloads' requirements.  
                      
 
                     Step 5 - Review recommendations
 
                        Click on “EC2 instances,” “EC2 Auto Scaling groups,” “EBS volumes,” “ECS services on Fargate ,” “Lambda functions”, or “RDS databases” on the left side to view all your rightsizing recommendations. You can also discover a list of your idle resources and their savings opportunities by clicking “Idle resources” and explore your license cost savings opportunities by clicking on “Licenses.”  
                      
 
                     Step 6 - Visualize a what-if scenario
 
                        Click “view detail” to visualize a what-if scenario. This helps you understand how the recommended instance would have performed on the recommended instance type.