Data-Engineer-Associate - Data Ingestion and Loading (21% of the exam) - Section 2.4

Configure Lakeflow Connect to reliably ingest data from diverse enterprise sources into Unity Catalog governed tables.

Set up Lakeflow Connect ingestion from enterprise sources such as SaaS applications and databases, including connections, ingestion pipelines and destination tables in Unity Catalog. Recognise what the managed connector handles for you (incremental change capture, retries) and what remains the engineer's job.

Lakeflow Connectmanaged connectorsUnity Catalog connectionsingestion pipelinechange data capture

Practice question for this objective

Free sampleData Ingestion and Loadingmedium

A data engineer must bring several tables from a self-managed SQL Server database into Unity Catalog and keep them current from the database's change data, using Lakeflow Connect rather than bespoke extraction code. The engineer asks which components a database connector puts in place. Which description is correct?

AA Unity Catalog connection holds the source credentials, an ingestion gateway extracts the initial snapshot and ongoing change data from the database into a staging location, and an ingestion pipeline applies them to Unity Catalog tables. Correct
BA single ingestion pipeline connects to SQL Server on every scheduled run and overwrites each destination table in full, with no separate component involved in reading change data.
CAn Auto Loader stream reads transaction log files that the database team exports to cloud storage, and a Lakeflow Connect pipeline then registers the resulting Delta tables in Unity Catalog.
DA Unity Catalog external location points at the database host, and a scheduled Lakeflow Job runs a MERGE statement against the source tables to pull the rows that have changed since the previous run.

A Lakeflow Connect database connector combines a Unity Catalog connection, an ingestion gateway that stages change data, and an ingestion pipeline writing governed tables. Database sources are read by a gateway that authenticates through the Unity Catalog connection, takes an initial snapshot and then follows the source's change data, staging both. A separate ingestion pipeline consumes that staged output and applies inserts, updates and deletes into Unity Catalog tables, which keeps credential handling, source extraction and governed writing as distinct, separately managed pieces.

Why A is correct: Correct. A database connector separates authentication, extraction and application: the connection object governs access to the source, the gateway reaches the database and stages snapshot plus change data, and the ingestion pipeline writes the governed destination tables.

Why B is wrong: This is tempting because a full reload is how a hand-written JDBC extraction is usually built, and it does produce a correct table. It is wrong because a Lakeflow Connect database connector is designed to read change data incrementally, and it uses a gateway rather than one component doing everything.

Why C is wrong: This is plausible because Auto Loader is the standard mechanism for incremental file ingestion and the description sounds like change data capture. It is wrong because Auto Loader only reads files, and Lakeflow Connect is not a registration step layered on top of another ingestion product.

Why D is wrong: This is tempting because external locations and MERGE are both genuine Unity Catalog and Delta features used in ingestion work. It is wrong because an external location describes a cloud object storage path, not a database host, and a MERGE runs against Delta tables rather than a remote source.

See more Data-Engineer-Associate practice questions, answers explained.

Exam traps in Data Ingestion and Loading

Answers that look right on this material and are not. Each one is a distractor from a different question in the Data-Engineer-Associate bank for this domain.

Write a Structured Streaming query that calls the Salesforce REST API inside foreachBatch and appends the responses to the target Unity Catalog tables each hour.
Why it is wrong: This is technically achievable and foreachBatch is a real sink API, but it puts the authentication, pagination and change tracking that the team cannot maintain back into hand-written code.
A folder of Parquet files that a partner writes into an external location every night, which must be topped up into a bronze table each morning.
Why it is wrong: Tempting because it is still ingestion into Unity Catalog, but files already sitting in cloud storage are the case Auto Loader and COPY INTO cover, with no source connector involved.
It registers a cloud object storage path together with a storage credential, so that the connector can list and read the raw files that the source system exports.
Why it is wrong: This describes an external location and its storage credential, which is how Unity Catalog governs paths in cloud object storage. A managed connector talks to the application over its own interface, so no storage path is being registered here.

Examworthy is not affiliated with or endorsed by Databricks. Original, blueprint-aligned practice material only.