Analysts need to occasionally join a small Redshift dimension table against petabytes of historical data sitting in S3 — without loading that data into Redshift. Which feature enables this?
A.Redshift Spectrum querying external tables over S3
Correct. Federated queries over the data lake.
B.COPY the entire S3 dataset into Redshift
Petabyte COPY is slow, expensive, and wastes cluster storage.
C.Redshift materialized views
Materialized views still require the data loaded first.
D.AWS DMS continuous replication into Redshift
DMS replicates databases, not analytical lake queries.
Why this answer
Spectrum lets Redshift query exabytes in S3 directly via external tables (Glue Data Catalog-backed), joining lake data with local tables — no loading required for infrequently accessed data. Loading petabytes to answer occasional queries is exactly what Spectrum exists to avoid.
Correct answer: A · Last reviewed 2026-11-02
Official AWS references
Practice this interactively
Answer without seeing the solution first in the Data Stores & Lakes practice session.
Practice Data Stores & Lakes