AWS Certified Solutions Architect Professional SAP-C02 Practice Question

The correct solution is to use an Amazon SQS FIFO queue with the client's unique ID as the MessageGroupId. This approach effectively decouples the producers (client apps) from the consumers (EC2 instances) and provides a buffer to smooth out traffic spikes, preventing the backend from being overwhelmed.

Using a FIFO (First-In-First-Out) queue is essential to meet the strict ordering requirement. By assigning the unique client ID to the MessageGroupId, all messages from that specific client are processed in the exact order they are received, while messages from different clients (with different MessageGroupIds) can be processed in parallel. This provides both the required ordering and high throughput.

• An Amazon SQS Standard queue is incorrect because it only provides best-effort ordering and does not guarantee that messages will be processed in the order they were sent, which violates a critical requirement.
• While Amazon Kinesis Data Streams can maintain order using a partition key, it is primarily designed for real-time streaming analytics and is more complex to manage than SQS for this simple work-offloading use case. SQS FIFO is the more purpose-built and efficient solution for decoupling and buffering with ordered processing for backend workers.
• An Application Load Balancer does not provide a buffering mechanism. It forwards requests directly to the backend targets. During a traffic spike, the ALB would still overwhelm the EC2 instances, leading to the same failures.