Serverless Computing

Correct Answer: DynamoDB

DynamoDB is the correct choice when designing a serverless architecture for high scalability and low latency. This is because DynamoDB is a fully managed NoSQL database service provided by Amazon for applications that require consistency and low latency even at scale. It integrates well with serverless architectures where developers do not have to manage server infrastructure. On the contrary, RDS is a managed relational database service suitable for structured data and applications that require complex joins and transactions. Aurora is a part of the RDS service and is compatible with MySQL and PostgreSQL but does not have the scalability and low latency advantage of DynamoDB at high throughput. Lastly, Redshift is a data warehouse best designed for large scale data analytics and not suitable for low latency serverless applications.

Correct Answer: AWS Kinesis Data Firehose

AWS Kinesis Data Firehose is the most suitable for real-time data processing and analysis. It is designed for the real-time processing of streaming data - it can capture, transform, and load streaming data into AWS destinations such as Amazon Simple Storage Service (Amazon S3), Amazon Redshift, Amazon Elasticsearch Service (Amazon ES), and Splunk. This characteristic makes it the best fit for handling data from IoT devices. On the other hand, AWS EMR is for large-scale data processing, not real-time data, RDS is a relational database service which is not typically used for real-time analysis, and AWS IoT Core is more about device communication and security rather than real-time data processing and analysis.

Correct Answer: Implement a queue-based load leveling pattern to process the orders efficiently and according to their priority.

By implementing a queue-based load-leveling pattern, the company can manage the processing of orders efficiently, taking into account their size and priority. This allows the processing of orders to be more flexible and better utilize system resources. The other options may not effectively manage the varying processing times and priorities of the orders.