Architecture
- Leader node parses queries and distributes work; compute nodes store data and execute queries.
- Distribution styles: AUTO, KEY, EVEN, ALL — choose based on join patterns and table size.
- Sort keys enable zone maps for efficient range-based filtering.
- Redshift Serverless provides on-demand capacity without cluster management.
Loading and Querying
- The COPY command is the fastest way to load data from S3, DynamoDB, or EMR.
- UNLOAD exports query results to S3 in parallel.
- Redshift Spectrum queries data directly in S3 without loading it into Redshift tables.
- Materialized views store precomputed results for frequently executed queries.
Exam Cues
- Need query S3 data without loading: Redshift Spectrum.
- Need fastest S3-to-Redshift ingestion: COPY command.
- Need prioritize critical queries: workload management (WLM) queues.
- Need reduce storage cost for infrequent data: keep in S3, query via Spectrum.
Practice Redshift Questions
Put your knowledge to the test with practice questions.