Distinguish between a data lake and a data warehouse; give examples in cloud.

Distinguish between a data lake and a data warehouse; give examples in cloud.

• A. Data lake stores raw, diverse data; data warehouse stores structured, analytics-ready data; examples: AWS S3 + Lake Formation/Glue vs Redshift; Google Cloud: Cloud Storage with BigQuery; Azure: Data Lake Storage + Synapse.
• B. Data lake stores only structured data; data warehouse stores unstructured data; examples: S3 and Databricks; BigQuery.
• C. Data lake stores only files; data warehouse stores only relational tables; examples: S3 and Snowflake; BigQuery and Synapse.
• D. Data lake and data warehouse are the same concept; both store the same data.

Answer: (A)

A data lake stores raw, diverse data in its native formats—structured, semi-structured, and unstructured—so you can ingest a wide range of sources and decide how to use them later. It usually follows a schema-on-read approach, meaning the structure is defined when you read the data, not when you store it. A data warehouse, on the other hand, is built for analytics on cleaned, structured data and uses schema-on-write, so data is transformed and organized before storage to support fast, repeatable SQL queries. In cloud environments, a lake is typically backed by object storage and complemented with cataloging and governance tools (for example, AWS S3 with Lake Formation or Glue, or Google Cloud Storage with BigQuery data surfaces, among others). A warehouse is provided as a purpose-built query service (such as AWS Redshift, Google BigQuery, or Azure Synapse) that’s optimized for analytical workloads. That combination matches the statement well: it accurately distinguishes the data types and processing approaches, and pairs them with valid cloud examples. The other options mix up what a lake or warehouse stores (one says only structured data or only files) or claim they’re the same thing, which doesn’t reflect how these architectures are designed to handle data and analytics.

A data lake stores raw, diverse data in its native formats—structured, semi-structured, and unstructured—so you can ingest a wide range of sources and decide how to use them later. It usually follows a schema-on-read approach, meaning the structure is defined when you read the data, not when you store it. A data warehouse, on the other hand, is built for analytics on cleaned, structured data and uses schema-on-write, so data is transformed and organized before storage to support fast, repeatable SQL queries.

In cloud environments, a lake is typically backed by object storage and complemented with cataloging and governance tools (for example, AWS S3 with Lake Formation or Glue, or Google Cloud Storage with BigQuery data surfaces, among others). A warehouse is provided as a purpose-built query service (such as AWS Redshift, Google BigQuery, or Azure Synapse) that’s optimized for analytical workloads.

That combination matches the statement well: it accurately distinguishes the data types and processing approaches, and pairs them with valid cloud examples. The other options mix up what a lake or warehouse stores (one says only structured data or only files) or claim they’re the same thing, which doesn’t reflect how these architectures are designed to handle data and analytics.