Amazon SAP-C02: Cost Optimization & Governance — Study Guide
Part of the AWS Solutions Architect Professional SAP-C02 — Study Guide. Practice with verified answers in the Amazon exam hub, or take timed practice tests on ExamRoll.io.
Cost visibility, allocation, and reporting
Achieving true cost visibility begins with consistent, enforced cost allocation and high-fidelity reporting. Deploy AWS Organizations with a designated payer account and enable consolidated billing to centralize invoices while maintaining per-account charge separation. Activate the AWS Cost and Usage Report (CUR) with hourly granularity, S3 delivery, and Amazon Athena integration so you can run ad hoc queries and join cost records with resource metadata. Implement a strict tagging strategy: define a mandatory set of cost-allocation tags (environment, project, owner, business-unit, and cost-center) and enforce them at provisioning using CloudFormation StackSets, Service Control Policies to limit untagged resource creation patterns, and AWS Config Rules that evaluate and auto-remediate missing tags. Pair Cost Explorer with reserved-entity and rightsizing reports to understand spend trends and idle capacity. Common traps include incomplete tag coverage that skews chargebacks, relying on the Billing Console alone without CUR analytics, and failing to capture cross-account or inter-region transfer costs. Decision trade-offs often come down to timeliness versus granularity: enabling hourly CURs and Athena costs more in processing but yields the precise attribution needed for business decisions, whereas coarse monthly summaries are cheaper operationally but mask transient spikes and inefficient resources.
Committed pricing, rightsizing, and purchase strategy
Optimizing committed spend requires choosing between Savings Plans, Reserved Instances, and transient pricing like Spot, while using rightsizing to match capacity to workload profiles. Begin with usage patterns from Cost Explorer and Compute Optimizer to identify steady-state CPU/memory baselines and opportunities for switching families or instance sizes. Prefer Compute Savings Plans when workload flexibility across instance types and Regions is needed, and use EC2 Instance Reservations when specific instance families and AZ placements justify deeper discounts. Leverage Spot for fault-tolerant batch and microservice workloads, but avoid Spot for single-AZ stateful services without robust checkpointing. Rightsizing should combine automated recommendations from Compute Optimizer and Trusted Advisor with manual review to avoid overaggressive downsizing that impacts performance. Watch for traps including overcommitting on 3-year RIs when business forecasts are uncertain, underutilizing Savings Plans because of untagged or account-isolated usage, and assuming instance family interchangeability without testing. Trade-offs are typically cost versus operational flexibility: deeper long-term discounts reduce unit cost but add business risk if demand drops or architecture changes; conversely, Spot and on-demand offer agility at higher per-unit cost.
Governance, policy enforcement, and automated remediation
Centralized governance establishes guardrails that prevent uncontrolled spend while enabling autonomous teams. Use AWS Control Tower or a well-architected Organizations baseline to provision accounts with preconfigured guardrails and centralized logging. Apply Service Control Policies to limit high-cost services or unapproved Regions, and deploy AWS Config Rules to detect noncompliant configurations such as public S3 buckets, oversized instance types, or missing encryption and tags. Integrate AWS Budgets with automated actions: set budget thresholds that trigger SNS notifications and automated remediation via Lambda or Systems Manager (for example, stop/terminate idle instances or reduce RDS instance class). Trusted Advisor complements governance by surfacing cost-optimization checks, but do not treat it as the only signal; Trusted Advisor’s free checks are limited and detailed insights need a Business or Enterprise Support plan. Common traps include using SCPs too restrictively and blocking legitimate operational changes, relying solely on notifications without automated enforcement, and granting excessive IAM privileges that permit bypassing policies. Evaluate governance decisions by weighing business autonomy against risk: stricter controls prevent runaway costs but can slow velocity and require a well-defined exception process.
Cost-aware architectural patterns and data transfer considerations
Architectural choices profoundly influence ongoing cost. Offload high-volume content and global distribution to Amazon CloudFront to reduce S3 origin requests and egress charges; use S3 Transfer Acceleration only when latency benefits justify higher transfer costs. For cross-AZ and cross-Region architectures, remember that inter-AZ data transfer can be charged; design for intra-AZ traffic locality where possible, or aggregate traffic through regional services. For large file distribution, consider Amazon S3 with multipart upload and lifecycle policies, S3 Intelligent-Tiering for unpredictable access, and EFS One Zone for single-AZ workloads where resilience trade-offs reduce cost. When migrating container workloads, evaluate Fargate versus EC2-backed ECS/EKS: Fargate increases operational simplicity and reduces cluster management overhead but typically costs more per vCPU/memory than well-packed EC2 Spot-backed node groups. Common traps include underestimating inter-region replication, misplacing high-churn logs in infrequent-access storage classes, and assuming VPC endpoints are free — they save on NAT egress but add per-hour and per-GB charges. Decision criteria should weigh data gravity, latency SLAs, and durability needs: choose cheaper storage or compute only where resilience and performance requirements permit.
Practical Problem: Use-Case Scenario
Scenario: Acme Global Enterprises operates a mature AWS environment with 18 member accounts under AWS Organizations, a centralized payer account, and workloads across three Regions. They have partial tagging adoption, run several long-lived EC2 fleets, and use S3 extensively for analytics data.
Challenge: They need to reduce monthly AWS spend by 20% within six months while preserving performance SLAs and enabling autonomous teams to deploy features.
Recommended Approach:
- Enable the Cost and Usage Report to S3 with hourly granularity and integrate with Amazon Athena; create Cost Explorer reserved and rightsizing reports for the last 6–12 months.
- Deploy Compute Optimizer and analyze steady-state instance use; buy a mix of Compute Savings Plans for broad coverage and 1-year Convertible RIs for predictable, family-specific workloads.
- Enforce tagging via CloudFormation StackSets and AWS Config Rules with auto-remediation for missing tags; funnel cost reports per tag to business-unit owners and create AWS Budgets with automated SNS and lambda remediation for threshold breaches.
- Shift static and globally distributed content to Amazon CloudFront, convert infrequently accessed data to S3 Intelligent-Tiering with lifecycle transitions, and identify batch jobs suitable for Spot with checkpointing to migrate off on-demand capacity.
Rationale: Centralized visibility (CUR + Athena) exposes concrete rightsizing and reserved-buy opportunities while automated guardrails (Config, StackSets, Budgets) enforce cost allocation and prevent regressions; combining Savings Plans with selective RIs balances discount depth and flexibility for reliable cost reduction without sacrificing performance.
← Resilience · All domains · Deployment →
Practice these questions → · Timed practice on ExamRoll.io →
Pass the whole exam — not just this question
You found this answer. Get every verified question and explanation in one place, and save hours of prep. Free to start.
Pass your exam →