Which solution meets these requirements?
Amazon DynamoDB with on-demand capacity mode
Amazon Aurora with one writer node and an Aurora Replica with the parallel query feature enabled
Amazon DynamoDB with provisioned capacity mode with 5,000 write capacity units (WCUs) and 10,000 read capacity units (RCUs)
Amazon Aurora with one writer node and two cross-Region Aurora Replicas
Explanations:
Amazon DynamoDB with on-demand capacity mode is ideal for unpredictable, bursty traffic, automatically scaling to handle up to 5,000 orders per second while supporting low traffic volumes outside of the sale. This option also meets the high availability and low latency requirements cost-effectively without needing fixed WCUs or RCUs.
Amazon Aurora is not as optimized for handling highly variable workloads or high write throughput as efficiently as DynamoDB. It does not scale as seamlessly to 5,000 write requests per second without manual configuration and scaling adjustments, which would impact cost-effectiveness.
Provisioned capacity mode in DynamoDB would require predefining WCUs and RCUs, which may lead to underutilization and unnecessary cost during low-traffic periods outside of the sale. On-demand capacity mode better aligns with the need for automatic scaling for bursty and unpredictable workloads.
Amazon Aurora with multiple replicas is more suitable for read-heavy applications rather than write-heavy workloads with bursty and unpredictable traffic. It does not scale automatically to support 5,000 writes per second, nor does it meet the cost-effectiveness requirement for this use case.