SOA-C02 Cost and Performance Optimization Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✓

  Purchase Reserved Instances for the baseline capacity.

  Why this is correct

  Reserved Instances provide a significant hourly cost reduction over On-Demand pricing in exchange for a one- or three-year commitment, directly addressing the stem’s requirement to optimise costs. By purchasing Reserved Instances for the baseline capacity—the minimum number of instances always running—the company locks in lower rates for that steady-state workload while retaining On-Demand or Spot instances to handle any variable traffic spikes, thus preserving performance.

• ✗

  Use Dedicated Hosts for all instances to control placement.

  Why it's wrong here

  Dedicated Hosts are physical servers reserved for your exclusive use, enabling license-bound software or strict compliance requirements. They are billed per host regardless of how many instances run, so they are substantially more expensive than On-Demand or Reserved Instances. For a typical web application without such constraints, this would inflate costs rather than optimize them, and it does nothing to address varying demand.

• ✓

  Implement Auto Scaling to match capacity with demand.

  Why this is correct

  Auto Scaling is an AWS service that automatically adjusts the number of EC2 instances based on demand using CloudWatch alarms and scaling policies. It prevents over-provisioning during low-traffic periods and under-provisioning during spikes, ensuring you only pay for the compute capacity actually needed. Combining Auto Scaling with a right-sized launch template and multiple purchase options is a central cost-optimization best practice.

• ✗

  Enable T2/T3 unlimited to handle spikes without throttling.

  Why it's wrong here

  T2/T3 Unlimited lets instances burst beyond their baseline CPU credits for a short time, but once the credit balance is exhausted, you are charged extra per vCPU-hour. While it avoids throttling, it can lead to unpredictable and higher costs, especially for workloads with sustained high CPU. For cost-sensitive production workloads, choosing a larger instance type or using Auto Scaling is more effective than relying on unlimited mode.

• ✗

  Use multiple instance types in each Availability Zone.

  Why it's wrong here

  Launching multiple instance types across Availability Zones is a high-availability strategy that protects against instance-type or AZ failures, but it does not directly reduce costs. In fact, without proper governance, this can increase waste by keeping unused capacity running in each AZ. Cost optimization relies on scaling capacity with demand and using the most appropriate pricing model, not on spreading instances across types and zones.

Go deeper

Related to this question