Question

A data engineering team supports a BI dashboard in Amazon QuickSight that queries a large fact table in Amazon Redshift. Business analysts complain that during peak reporting hours the Redshift cluster is heavily loaded by hundreds of concurrent dashboard viewers, slowing down the nightly ETL jobs. The dashboard only needs to reflect data as of the previous night's ETL completion, not real-time changes. Which change should the team make to reduce load on Redshift while meeting the freshness requirement?

Accepted Answer

Switch the QuickSight dataset from direct query mode to SPICE and schedule a SPICE refresh to run after the nightly ETL completes.

Answer

Enable Redshift concurrency scaling and keep the dashboard in direct query mode so each viewer runs a live query.

Answer

Migrate the fact table to Amazon Athena and point QuickSight at it using direct query mode.

Answer

Create a materialized view in Redshift and have QuickSight direct query it on every dashboard load.