AWS Certified Solutions Architect Professional SAP-C02 Practice Question

The correct answer is to use a high-memory EC2 instance within an Auto Scaling group with a fixed size of one. This architecture directly addresses the core constraints of the workload. The single-threaded, high-memory nature of the application necessitates a scale-up approach, which is satisfied by using a memory-optimized EC2 instance type (such as the 'u', 'x', or 'r' families) that can provide more than 512 GiB of RAM. Placing this single instance in an Auto Scaling group configured with a minimum, maximum, and desired capacity of 1 provides high reliability. If the instance or its underlying Availability Zone fails, the Auto Scaling group will automatically terminate the unhealthy instance and launch a new one in a healthy Availability Zone, ensuring the job can restart and meet its SLA without manual intervention.

• Using a fleet of smaller EC2 instances with a load balancer is incorrect because the core application is monolithic and single-threaded, meaning it cannot be distributed across multiple instances to scale out. This approach would require a complete re-architecture of the application itself.
• Using AWS Batch with Fargate is incorrect due to a specific service limitation. AWS Fargate tasks have a maximum memory limit of 120 GiB, which is insufficient for the application's requirement of over 512 GiB.
• Launching a single instance without an automated recovery mechanism is incorrect because it introduces a single point of failure and does not meet the 'highly reliable' requirement. Relying on a manual process for failover is slow, error-prone, and not a resilient architecture.