An analyst wants to run ad hoc ANSI SQL over Parquet files stored in an Amazon S3 bucket. She wants to pay only for the data each query scans and does not want to provision or manage a cluster. Which AWS service should she use?
Show answer and explanation
Answer: A
Amazon Athena lets you issue standard SQL against objects in S3 with nothing to provision. Charges depend on the volume of data each query scans, and columnar formats such as Parquet reduce that volume. Redshift is a data warehouse that typically loads data into its own managed storage, Aurora is a relational database engine for transactional workloads, and Keyspaces is a managed Cassandra-compatible NoSQL service.
Why the other options are wrong
- B. Redshift is a data warehouse that typically stores data in its own managed storage after loading it. Running ad hoc queries over raw S3 files there is not its serverless, pay-per-scan pattern.
- C. Aurora is a managed relational database for transactional applications. It is not designed for scanning large collections of files in S3 with ad hoc SQL.
- D. Keyspaces is a managed Apache Cassandra-compatible service for wide-column NoSQL workloads. It does not run SQL queries over Parquet objects in S3.