PL-300 sample exam questions

Importing the dataflow data and scheduling a daily refresh gives fast, in-memory visuals and data up to the previous day with minimal online processing. DirectQuery against the dataflow would push queries to the compute engine every time someone uses a visual, and hourly refreshes through Power Automate add unnecessary processing.

Select the discount column and set its type to Decimal Number, because Whole Number would round values such as 0.08 to 0. Then use Replace Errors to put 0.05 in place of the errors. This keeps every row, while Remove Errors would delete rows.

INTERSECT(ProductsGroupA, ProductsGroupB) returns the rows of the first table (A) that also appear in the second table (B), keeping duplicates from A. EXCEPT would return the rows of A with no match in B, and UNION would return all rows from both.

The relationship is one-to-many from Table1 to Table2 with single cross-filter direction, so filters flow only from Table1 to Table2. Table1 code SM has no matching row in Table2, so Table2's Code shows a blank for it.

Sync slicers carries the first page's selection to every page. Hiding the slicer on pages two and three, while keeping it synced, means those pages still apply the filter but users there see only the results and cannot change the selection.

In Power BI Desktop's Canvas settings, choose a custom page size with a width and height that match the portrait display, so the report fills the screen. Background image fit only stretches the image, Personalize visuals is for readers, and the Pages pane takes up space.

RLS roles are defined in Power BI Desktop, but membership is assigned after publishing. In the Power BI service, open the semantic model's Security settings and add the HR team (preferably a security group) to the HR role. Sharing the dataset alone does not map users to the role.

Build permission lets User1 create reports from the HR dataset without being able to modify it, whereas a WorkspaceA Contributor could edit it. Contributor on WorkspaceB allows publishing reports there but not adding users, which Member and Admin can do.

The case requires near-real-time data from the on-premises SQL Server, so DirectQuery through the on-premises data gateway is needed. The gateway data source uses a stored database credential. Import would show stale data, and passing end-user credentials (SSO) is not required by the scenario.

RLS roles and their DAX filters are defined in Power BI Desktop, then published with the model. Members are assigned to roles in the service. RLS applies only to Viewers, so giving users Contributor or Member would bypass it because those roles have Write permission.

The community chose app distribution, but an app grants only read access. The sales department must create, edit and delete content, manage permissions, publish and unpublish the app, and assign groups to workspace roles. Microsoft lists exactly these as Member role capabilities, so the least-privilege fit is Member, not Admin.

Balance sheet amounts are semi-additive snapshots, so the ending balance is the balance on the last date of the selected period. CALCULATE(SUM(BalanceSheet[BalanceAmount]), LASTDATE('Date'[Date])) returns it. DATESQTD would add up balances across the quarter, MAX returns the largest single row, and FIRSTNONBLANK returns a date rather than the ending balance.

After merging the manager tables into Region, creating a hierarchy (region manager name, then sales manager name) lets executives drill from region managers down to sales managers, as their reporting requirements ask. RLS, bidirectional relationships and a lookup column do not provide the drill-down.

The percentage of late orders has to be compared across many countries and shipping regions, which a bar chart shows clearly. A pie chart only works for a few parts of a whole and cannot show the percentage per country well, and a scatterplot compares two numeric measures.

The scenario's reports draw on data with different security and refresh requirements that cannot share one model. Two imported semantic models (datasets) is the minimum that supports all the reports. DirectQuery is not needed for the required data freshness and would slow the reports.