The correct option is First maps to a Data Lake and second maps to a Data Warehouse and third maps to a Database.

The first need describes hundreds of terabytes of raw images, short videos, and HTTP logs, which are best placed in a Data Lake. This paradigm is built for large volumes of unstructured and semi structured data and offers cost effective and elastic storage that can feed downstream processing and analytics. On Google Cloud, Cloud Storage commonly implements this pattern with durable object storage and lifecycle management.

The second need calls for complex analysis over seven years of cleaned and structured order history to uncover trends, which is exactly what a Data Warehouse is designed for. A warehouse supports analytical SQL at scale with columnar storage, partitioning, clustering, and the ability to efficiently scan and aggregate across large datasets. In Google Cloud, BigQuery is the standard warehouse for this use case.

The third need requires sub second retrieval of a single customer’s current order status, which fits a Database that supports transactional reads and indexed point lookups. On Google Cloud this is typically served by Cloud SQL, Firestore, or Spanner depending on the required consistency model, relational features, and scale.

First maps to a Data Warehouse and second maps to a Data Lake and third maps to a Database is incorrect because it reverses the first two paradigms. Raw files do not belong in an analytical warehouse and curated historical analytics do not belong in a file oriented repository. Even though the third part uses a database, the overall mapping does not match the workloads.

First maps to a Data Lake and second maps to a Database and third maps to a Data Warehouse is incorrect because the second need is analytical at scale and not an operational lookup store, and the third need requires fast transactional access rather than an analytical engine.

First maps to a Database and second maps to a Data Warehouse and third maps to a Data Lake is incorrect because the first need involves massive unstructured assets that do not fit a transactional system, and the third need is an operational lookup which does not fit a file based repository.

Full AWS Practitioner Certification Question