A new web application will run on Amazon EC2 and use Amazon DynamoDB for backend storage. Traffic is unpredictable and expected read/write throughput is moderate to high. The application must scale with traffic and be cost-effective. Which DynamoDB table configuration best meets these requirements?

The correct option is configuring DynamoDB in on-demand mode using the DynamoDB Standard table class. On-demand capacity mode is ideal for unpredictable workloads because it automatically scales read and write capacity to accommodate traffic spikes without requiring manual provisioning or auto scaling configuration. You pay only for the reads and writes your application performs, making it cost-effective for fluctuating usage. The Standard table class is suitable for most workloads with moderate to high throughput. Provisioned capacity with auto scaling (first and third options) requires defining initial capacity and scaling policies, which is less optimal for unpredictable traffic compared to on-demand. While auto scaling helps, on-demand offers greater simplicity and cost efficiency for highly variable workloads. The Standard-IA table class (third and fourth options) is designed for data that is accessed less frequently and has higher read/write costs than the Standard class, making it less suitable for an application with moderate to high expected throughput.