Question

A data engineering team runs an Amazon Redshift cluster storing three years of transactional data in a large fact table. Analysts almost always filter queries by an event_date column to analyze recent activity (for example, the last 30 days), and these queries currently scan far more data than needed, causing slow response times. The team wants to minimize the number of blocks Redshift reads for these date-filtered queries. Which schema design change should they implement?

Accepted Answer

Define event_date as the compound sort key on the fact table so Redshift can skip blocks outside the queried date range using zone maps.

Answer

Set the distribution style to KEY on event_date so all rows for a date range reside on a single node.

Answer

Increase the number of nodes in the cluster to spread the fact table across more slices.

Answer

Apply column-level RAW encoding to event_date to speed up range comparisons.