A company runs an ecommerce application on Amazon EC2 instances that store purchase details in an Amazon Aurora PostgreSQL DB cluster. Customers experience application timeouts during peak usage. A solutions architect must rearchitect so the application can scale to peak demand MOST cost-effectively. Which combination of actions should the architect take? (Choose two.)

The correct options address both application and database scaling for peak demand. Sending purchase requests to an SQS queue decouples the application from immediate processing, allowing it to handle sudden spikes in requests without timeouts. An Auto Scaling group of EC2 instances can then process these requests from the queue, scaling out as needed. This makes the processing asynchronous and resilient. Using Amazon RDS Proxy with Aurora PostgreSQL improves connection management and resilience, especially during high connection rates, by pooling and reusing database connections. This prevents the database from being overwhelmed by too many connections from scaling application instances. Using ElastiCache would help with read-heavy workloads but not directly with write-heavy purchase transactions. AWS Lambda for retries is less suitable for continuous processing of purchase requests compared to an Auto Scaling group. API Gateway with a usage plan is for API management and throttling, not for backend processing or database connection management.