DP-700 - Ingest and Transform Data (33% of the exam) - Section 2.6

Create and manage OneLake shortcuts, and implement mirroring of external databases.

Create OneLake shortcuts to ADLS Gen2, Amazon S3, and internal Fabric items without copying data, and configure shortcut security and delegation. Contrast shortcuts with database mirroring for Azure SQL, Snowflake, and Cosmos DB, which provides near-real-time replication into OneLake.

OneLake shortcuts to ADLS Gen2, S3, and internal itemsshortcuts versus copying datadatabase mirroring (Azure SQL, Snowflake, Cosmos DB)near-real-time replicationshortcut security and delegation

Practice question for this objective

Free sampleIngest and Transform Datamedium

A team uses Snowflake as its operational analytics store and wants the same data available in Microsoft Fabric as Delta tables in OneLake that Spark and the SQL analytics endpoint can query. The replica should follow source changes closely, and the team wants a managed capability rather than a pipeline they have to build and operate. Which Fabric capability best meets these requirements?

AA OneLake shortcut from the Lakehouse to the Snowflake warehouse, so Fabric queries read the Snowflake tables in place without any replication.
BA Spark structured streaming job that continuously reads from Snowflake and writes Delta tables into the Lakehouse for the analytics endpoint.
CFabric mirroring for Snowflake, which replicates the Snowflake database into OneLake as Delta tables that Spark and the SQL analytics endpoint can query. Correct
DA scheduled Dataflows Gen2 refresh that reads Snowflake and writes the tables into the Lakehouse once an hour for downstream queries.

Use Fabric mirroring for Snowflake to land a managed, near-real-time Delta replica in OneLake instead of building a replication pipeline. Mirroring supports Snowflake as a source and continuously replicates its tables into OneLake in Delta format, so Spark and the SQL analytics endpoint can query a replica that closely follows the source; because the service manages the replication, the team avoids building and operating a streaming or refresh pipeline of its own.

Why A is wrong: Shortcuts point at storage locations and supported item types, not a Snowflake compute warehouse, so this does not surface Snowflake tables as queryable OneLake Delta tables.

Why B is wrong: A structured streaming job could replicate the data, but it is a pipeline the team must build, tune, and operate, which contradicts the requirement for a managed capability with low overhead.

Why C is correct: Fabric mirroring supports Snowflake as a source and continuously lands the data in OneLake as Delta tables, giving the managed, near-real-time replica the requirement asks for without a hand-built pipeline.

Why D is wrong: An hourly Dataflows Gen2 refresh lags the source by up to an hour and is a schedule the team manages, so it neither follows changes closely nor removes the operational burden mirroring eliminates.

See more DP-700 practice questions, answers explained.

Exam traps in Ingest and Transform Data

Answers that look right on this material and are not. Each one is a distractor from a different question in the DP-700 bank for this domain.

The consumer must hold their own role assignment on the Azure Data Lake Storage Gen2 account, and access is denied if they lack one even when reading through the shortcut.
Why it is wrong: This describes passthrough or delegated identity, but an Azure Data Lake Storage Gen2 shortcut created with a stored connection uses that connection's credential, so the consumer is not required to hold their own storage role.
Create a OneLake shortcut in a Lakehouse that points at the Azure Cosmos DB container so the documents appear in OneLake without being copied or replicated anywhere.
Why it is wrong: Shortcuts virtualise file and table storage such as ADLS Gen2, S3, or internal items, but a transactional Cosmos DB container is not a shortcut target, so this cannot expose the documents as Delta.
Build a Data Factory pipeline that reads the source change data capture tables on a short schedule and merges the changes into a Lakehouse Delta table.
Why it is wrong: A hand-built change-data-capture pipeline can approach the latency target, but it is exactly the operational overhead the requirement avoids, since the team must build, schedule, and maintain the merge logic itself.

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