DP-700 — frequently asked questions

No, it only syncs new tables created after enabling. You must explicitly select 'Apply to existing tables' to backfill old data.

The feature was likely enabled without the backfill option. Only new data added to the eventhouse is automatically copied to OneLake.

A lakehouse stores semi-structured Spark-written data and offers T-SQL via its SQL analytics endpoint, but KQL is not native there; you need an eventhouse or a shortcut to query it with KQL.

Yes. Eventhouse KQL databases support Spark ingestion via the Kusto Spark connector and still serve KQL and T-SQL reads.

Dataflow Gen1 is deprecated in Fabric, so exam answers favor a data pipeline with Copy data for a direct gateway-backed load from SQL Server to a warehouse.

A KQL queryset queries KQL databases in Eventhouse; it is not a general data-movement item and cannot copy SQL Server tables into Warehouse1.

No. Existing tables are copied only if you select the Apply to existing tables option for historic backfill; otherwise only new data is copied.

Because current docs mention an Apply to existing tables backfill option, but the stem never says that option was selected, so default behavior remains new data only.

Workspace identity is a Fabric-managed identity for workspace outbound access to resources; it is not a credential an external Azure DevOps pipeline uses to call Fabric APIs.

Yes for supported item types and operations, but check the current Fabric REST identity-support table because some deployment-pipeline operations limit service principals to Power BI items.

Because appropriate object-level permissions on DW1 are already assigned; Contributor would also allow edits to Pipeline1 and other workspace items, violating least privilege.

Yes, if DW1 has explicit object-level permissions for User3; the workspace Viewer role controls item visibility, while the warehouse permission grants the update capability.

No. Domain contributor status is separate from workspace roles; User3 needs an explicit Workspace3 role for item-level access.

Yes. New workspaces created by users in a default domain are automatically assigned to that domain, and Group1 members generally become domain contributors.

In KQL, order by and sort by default to descending order, so omitting asc sorts No_Bikes from highest to lowest. Ascending must be requested explicitly with the asc keyword.

No — order by and sort by are synonyms and behave identically, both defaulting to descending. Neither fixes the missing asc keyword in the exhibit code.

Caching stores data in Fabric-managed cache, not in the bronze, silver, or gold folders of Lakehouse1, so the lakehouse itself holds no raw copy while egress drops sharply.

A workspace identity handles authentication to services such as Key Vault; it neither avoids copying S3 files nor reduces cross-cloud data transfer, so it satisfies neither requirement.

Stream Analytics does detect anomalies, but it runs as a standalone Azure resource outside the Fabric workspace, requiring extra provisioning, query writing, and output configuration — which violates the minimize-development-effort requirement.

No. Dataflow Gen2 is the batch ingestion tool already used for the medallion loads and cannot stream IoT sensor data for near-real-time dashboards.

Integer keys improve join speed, but they do not lower the 500 million row count, so the table still exceeds the F32 Direct Lake limit and keeps falling back to DirectQuery.

Yes. V-Order is enabled by default for Fabric warehouses and can only be disabled, not newly enabled, so option C changes nothing about the current performance.

SalesAmount is an additive measure and Date identifies the sale event, so both are fact-table columns. Dimension tables hold only attributes that describe the product itself.

SCD tracking uses surrogate keys and generated valid-from/valid-to dates created during transformation, not the transaction Date from the original flat table.

Lakehouse sharing in Fabric grants read access to the lakehouse's contents, so the engineer would still be able to read the Employee table and its PII. Sharing cannot hide individual tables.

After the migration, Workspace2 holds the Lakehouse2 that contains the PII-bearing Employee table. Granting the engineer any role there would expose exactly the data you are trying to protect.

A schema only groups database objects such as tables and policies; it has no logic that can test USER_NAME() and allow or deny individual rows of the sales table.

The function implements the filter predicate logic, and CREATE SECURITY POLICY binds that function to the table so every query is filtered automatically for the connected user.

Yes. Only the model definition is deployed, so the models arrive empty in the target stage and you must refresh them (and recreate the refresh schedule) to load data.

Yes, Dataflow Gen1 is a supported item type in Fabric deployment pipelines, so the flows deploy as definitions — but verify connections and credentials per stage after deployment.

Deployment pipelines copy or synchronize items between workspaces at the item level; they do not parse warehouse T-SQL code, so broken object references deploy silently.

No — both would require custom logic to query catalog views and compare dependencies, which violates the 'minimize development effort' requirement.

The retention period for cached files in OneLake shortcuts can be configured from 1 to 28 days.

Because the cache retention is configurable up to 28 days, older files can still be served from cache if accessed within that window.

Products targets ProductFile on ADLS Gen2, and shortcut caching is only supported for GCS, S3 and S3-compatible shortcuts, so workspace enablement does not apply to it.

TripsFile is about 2 GB, and individual files greater than 1 GB are never cached, so only smaller S3 files such as StoreFile are stored in the cache.

A domain admin manages a domain that already exists; the user must first create the sales domain, which only a Fabric admin can do. Least privilege still applies because no lower role can complete all four tasks.

A domain contributor can add workspaces to an existing domain, but it cannot create subdomains or domains, so it fails two of the required tasks.

Because the items with matching names are already paired, deployment overwrites them in place. Deleting or renaming would break that pairing and force a clean deploy instead.

No. Items unique to the target stage stay untouched, and new source items are created as copies. The question only requires overwriting the corresponding items.

Yes. Fabric Warehouse supports T-SQL dynamic data masking, so you can use ALTER TABLE Table1 ALTER COLUMN Phone ADD MASKED WITH to apply a partial mask.

Table-level UNMASK gives Group2 unmasked values without also letting Group1 see full Phone values; a database-level grant would be broader than the requirement allows.

DimCustomer has a foreign key to DimGeography, so loading it first could create customer rows that reference missing geography keys.

No. Both are fact tables that depend on dimension keys; loading them first risks orphaned facts and referential-integrity failures.