Question

A media company is building a new customer-facing analytics portal. Business users need sub-second interactive dashboards over a 40 TB dataset stored in Amazon S3 in Parquet format. The workload is highly variable: usage spikes to thousands of concurrent dashboard viewers during weekday mornings and drops to near-zero overnight. The architecture team wants a fully managed solution that scales concurrency automatically, minimizes cost during idle periods, and avoids provisioning fixed compute capacity for query serving. Which design BEST meets these requirements?

Accepted Answer

Use Amazon Athena to query the S3 data directly and embed Amazon QuickSight dashboards backed by SPICE, refreshing SPICE datasets on a schedule.

Answer

Load the data into an Amazon Redshift provisioned cluster sized for peak concurrency and embed QuickSight dashboards querying Redshift directly.

Answer

Deploy a fleet of Auto Scaling EC2 instances running a Presto cluster behind an Application Load Balancer and connect QuickSight in direct query mode.

Answer

Store the data in Amazon RDS for PostgreSQL with read replicas and enable QuickSight direct query with a read replica scaling policy.