What is a data warehouse, data lake, and data lakehouse, and when would you use each?

• A. Data warehouse stores structured data for reporting with optimized queries; a data lake stores raw/varied data for exploration; a lakehouse combines lake flexibility with warehouse performance for analytics
• B. A data warehouse stores unstructured data; a data lake stores fixed-schema data
• C. A data lakehouse rejects governance and focuses on speed
• D. A data warehouse stores only metadata

Answer: (A)

The idea being tested is how these storage architectures differ in structure, governance, and analytics, and when each is appropriate. A data warehouse is designed around structured data and is optimized for fast, repeatable SQL queries used in reporting and dashboards; the data is cleansed, organized, and governed to support reliable business insights. A data lake stores raw, diverse data at scale and uses a schema-on-read approach, which makes it ideal for exploration, data science, and machine learning where flexibility and cost-effective storage matter. A data lakehouse combines the best of both by using a data lake as the storage layer while adding warehouse-like features—rigorous metadata management, governance, and ACID transactions—to deliver fast analytics and reliable BI/ML workloads on diverse data. Use the data warehouse for stable, governed reporting on structured data; use the data lake for large-scale discovery and experimentation with varied data; use the lakehouse when you need flexible data storage with strong analytics performance and governance in a single platform.