SAA-C03 Design High-Performing Architectures Practice Question
A retail company is deploying a read-heavy product catalog on Amazon Aurora MySQL. The primary instance is heavily loaded with read traffic, and the team wants to offload reads to Aurora Replicas while keeping the application resilient to replica failures. The application connects using a single endpoint. Which two actions should the team take to meet these requirements? (Choose two.)
⚠ Common exam trap
The trap here is assuming that connecting directly to each replica provides better control, when it actually bypasses the built-in load balancing and failover of the reader endpoint.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Enable Aurora Auto Scaling for the cluster so that Aurora Replicas are added or removed based on read workload.
Aurora separates the writer endpoint from the reader endpoint, and the reader endpoint automatically distributes connections across all Aurora Replicas. Pairing that endpoint with Aurora Auto Scaling lets the cluster add or remove replicas based on read demand, so the application gains both read offloading and resilience without custom failover logic. Direct instance endpoints, caching, and multi-master do not fulfill the specific goal of routing reads through replicas.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Configure the application to connect directly to each Aurora Replica by its individual instance endpoint and round-robin between them in code.
Why it's wrong here
Connecting directly to individual replica endpoints requires custom failover logic in the application and does not automatically adapt when replicas are added or removed. It increases operational complexity and can cause errors when a replica is unhealthy. The Aurora reader endpoint already provides this distribution and resilience, so manual round-robin is unnecessary and less reliable.
- ✓
Enable Aurora Auto Scaling for the cluster so that Aurora Replicas are added or removed based on read workload.
Why this is correct
Aurora Auto Scaling monitors read workload and automatically adds or removes Aurora Replicas within a defined range. Combined with the reader endpoint, this keeps read capacity aligned with demand and improves resilience because the endpoint always includes healthy replicas. It reduces manual intervention and helps maintain performance during traffic spikes.
- ✗
Create an Amazon ElastiCache for Memcached cluster and cache all product catalog queries there.
Why it's wrong here
Caching can reduce read load, but it does not offload reads to Aurora Replicas as required, and it introduces cache invalidation complexity. Memcached also lacks the high availability and failover features of Redis, so it does not directly improve resilience. The scenario specifically asks to use Aurora Replicas, making this approach a different architectural choice rather than the requested action.
- ✗
Convert the Aurora cluster to a multi-master cluster so that all instances accept writes and reads.
Why it's wrong here
Aurora multi-master is intended for write scaling and high availability with multiple writer nodes, not for read offloading. It does not use the reader endpoint model and adds complexity around conflict handling. The requirement is to offload reads to replicas, which is better served by standard Aurora Replicas and the reader endpoint.
- ✓
Create one or more Aurora Replicas and connect the application to the reader endpoint of the Aurora cluster.
Why this is correct
The Aurora reader endpoint load-balances connections across all Aurora Replicas in the cluster. By pointing read traffic at this endpoint, the application automatically distributes reads and continues working if a replica is added or removed. This directly offloads the primary instance and supports resilience without hardcoding individual replica hostnames.
Go deeper
Related to this question
About these practice questions
This SAA-C03 question is part of Courseiva's 935-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Amazon Web Services exam blueprint
This SAA-C03 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the SAA-C03 exam.