AWS Cost Optimization

To decrease costs without breaking your infrastructure, you can follow these steps from easy to hard:

1. Analyze and optimize EC2 instances:

   • Use AWS Cost Explorer to identify underutilized EC2 instances.
   • Consider rightsizing instances by switching to smaller instance types that still meet your performance needs.
   • Implement Auto Scaling to automatically adjust the number of instances based on demand.

2. Optimize EBS volumes:

   • Review and delete unnecessary or unused EBS volumes.
   • Consider changing the volume type to a more cost-effective option if performance allows.

3. Optimize S3 storage:

   • Implement lifecycle policies to automatically move infrequently accessed data to cheaper storage classes like S3 Standard-IA or Glacier.
   • Delete unnecessary or outdated objects.

4. Review and optimize data transfer costs:

   • Analyze your data transfer patterns and consider using Amazon CloudFront to reduce data transfer costs.
   • Optimize your application to minimize unnecessary data transfers.

5. Leverage cost-effective pricing options:

   • Consider using Reserved Instances or Savings Plans for EC2 and RDS to get discounted rates for long-term commitments.
   • Utilize Spot Instances for non-critical, interruptible workloads to save on EC2 costs.

6. Implement governance and monitoring:

   • Set up AWS Budgets to track and alert on spending.
   • Use AWS Cost Anomaly Detection to identify unusual spending patterns.
   • Implement tagging strategies to better track and allocate costs across different projects or departments.

7. Optimize database usage:

   • Review and optimize your RDS instances, considering factors like instance size, storage type, and multi-AZ deployments.
   • Consider using Amazon Aurora Serverless for variable workloads to automatically scale capacity and reduce costs during idle periods.

8. Explore serverless options:

   • Evaluate if any of your workloads can be migrated to serverless services like AWS Lambda, which can be more cost-effective for certain use cases.

Remember to carefully test any changes to ensure they don't negatively impact your infrastructure's performance or reliability. Continuously monitor and iterate on your cost optimization efforts as your workloads and AWS usage evolve.
Sources
Community | AWS cost optimization: Top tips and best practices to reduce your bill
AWS Cost Optimization | AWS Cloud Financial Management
AWS Cost Optimization | AWS Cloud Financial Management

A lot of this will depend on your need for availability and/or uptime. Does EC2 or ECS need to running 24x7x365? If not, scheduling the instance to stop and start as needed will provide significant savings and should be your first step. Using tools like trusted advisor or compute optimizer may help in understanding if the instances you're running are optimized. If you are able to move to a different instance type, or especially a lower size, you can see significant cost savings. Only after you've done those two would I start to consider a savings plan. As was pointed out already in these responses, a savings plan is a commitment, minimally 1 year. So if you choose to go down that path then ensure your infrastructure will be up and running for a year. For EC2 alone, you'll save the most in terms of percentage off on-demand pricing with an EC2 Instance savings plan, but a generic compute savings plan may apply to your ECS usage as well, plus offers flexibility if you do intend to change around the instance family you're using.