Learning Objectives
By the end of this chapter, you will be able to:
- Design cost-optimized storage solutions.
- Design cost-optimized compute solutions.
- Design cost-optimized database solutions.
- Design cost-optimized network architectures.
Executive Summary
- Every AWS service has multiple pricing dimensions; the single largest cost lever is matching pricing models (On-Demand, Reserved, Savings Plans, Spot) and tiers (storage classes, capacity modes) to actual usage patterns rather than peak assumptions.
- Storage cost optimization centers on S3 storage class selection and lifecycle policies, EBS volume type migration (gp2 to gp3), and eliminating orphaned resources such as unattached volumes and old snapshots.
- Compute cost optimization follows a three-layer strategy: right-size first, then commit to Savings Plans or Reserved Instances for steady-state baseline, and use Spot Instances for fault-tolerant or flexible workloads.
- Network cost optimization focuses on reducing data transfer charges through VPC endpoints (free Gateway Endpoints for S3 and DynamoDB), same-AZ placement, CloudFront caching, and eliminating unnecessary NAT Gateway traffic.
Assumptions
- All pricing references in this chapter are directional and region-dependent. Exam questions test your understanding of relative cost relationships (e.g., Glacier Deep Archive is cheaper per GB than Standard), not memorization of exact dollar amounts.
- This chapter assumes familiarity with the services introduced in Chapters 1-3 (VPC, S3, EC2, RDS, Aurora, DynamoDB, Lambda, CloudFront, NAT Gateway, etc.) and focuses on the cost dimension of those services.
- Unless stated otherwise, examples reference the US East (N. Virginia) region, which typically has the lowest pricing.