Maintain a data analytics solution
Free Microsoft Certified: Fabric Analytics Engineer Associate practice — 6 questions on Maintain a data analytics solution, with explanations. No sign-up. Full 12-question mixed test →
A Fabric workspace contains a lakehouse used by multiple departments. The Finance team must be able to read only the files under Files/finance-reports/ in the lakehouse, but must not see any other folders, tables, or workspace items. The workspace already has other users with Viewer and Contributor roles. Which approach BEST meets this requirement?
OneLake data access roles let you define folder- or file-level scoped permissions directly on OneLake data, independent of workspace or item-level permissions, which is exactly the granularity needed here.
Two developers work on the same Fabric workspace connected to a Git branch. Developer A modifies a report in the workspace and commits it to Git. Meanwhile, Developer B pulls the branch, edits the same report definition file directly in Git, and pushes a change. When the workspace attempts to sync 'Update from Git,' a conflict is detected on that report item. Which statement correctly describes how Fabric resolves this conflict?
Fabric Git integration detects conflicts at the item level when the same item changed in both the workspace and Git; resolution requires selecting whether to keep the workspace version or the Git version for that item, since Fabric does not merge internal item contents automatically.
A Fabric Warehouse table stores customer credit card numbers in a column named CardNumber, formatted as 16 digits. Compliance requires that most users see only the last four digits (e.g., XXXX-XXXX-XXXX-1234) when querying the table, without altering the underlying stored data. Which masking function should you apply in the dynamic data masking rule?
The partial() masking function lets you specify a prefix length to hide, a custom padding string, and a suffix length to expose — exactly matching the requirement to show only the last four digits with a custom prefix pattern.
An organization uses a Fabric deployment pipeline with Dev, Test, and Production stages. A semantic model in each stage must connect to a warehouse that is unique to that stage (DevWarehouse, TestWarehouse, ProdWarehouse). Currently, after every deployment to Production, an engineer manually edits the semantic model's data source connection to point to ProdWarehouse. Which change eliminates this manual step going forward?
Fabric deployment pipelines support deployment rules, which let you define stage-specific override values (such as a different connection string or parameter) that are automatically applied every time content is deployed into that stage, removing the need for manual post-deployment edits.
A capacity administrator reviews the Fabric Capacity Metrics app and observes that interactive Power BI report queries are being rejected due to capacity overload, while a scheduled dataflow refresh on the same capacity continues running despite sustained high utilization. Which explains this difference in behavior?
Fabric capacity uses different smoothing windows depending on operation type: interactive operations are smoothed over 5 minutes, making overages trigger rejection quickly, while background operations are smoothed over 24 hours, allowing them to absorb bursts of usage for much longer before throttling or rejection occurs.
A team is building a semantic model in Direct Lake storage mode over a Fabric lakehouse. They define row-level security (RLS) roles with DAX filter expressions directly in the semantic model. A colleague asks whether they must also configure matching OneLake data access roles on the lakehouse tables for the semantic model's RLS to be enforced. What is the correct answer?
RLS defined on a semantic model is enforced by the Fabric/Analysis Services engine at query time for any client, regardless of whether the model uses Import, DirectQuery, or Direct Lake storage mode. Lakehouse-level security (like OneLake data access roles) governs direct access to lakehouse data separately and doesn't need to duplicate model-level RLS.
Ready for the real thing?
The full course: two full-length practice tests, video lessons for every exam domain, hands-on labs and detailed explanations.
$109.99 $34.99 with code FREETEST33 — valid through September 16.
Get my $34.99 deal →