30.5 Visualization, Service and Security

Q46. What is the difference between drill down and drill through?

Drill down navigates levels of a hierarchy within the same visual (Year → Quarter → Month). Drill through takes the user to a different page filtered by the selected item (e.g. a Dark Store Details page).

Q47. What are bookmarks used for?

Capturing a report page state (filters, slicers, visibility, sort) – for reset buttons, toggling visuals, and storytelling/navigation.

Q48. What is the difference between a field parameter and a what-if parameter?

A field parameter lets users switch the measures or dimensions shown in a visual. A what-if parameter creates a numeric series and slider to test scenarios (e.g. price increase %) used in measures.

Q49. What is a gateway, and when is it needed?

The on-premises data gateway lets the cloud Service refresh or query data inside a private network (on-premises databases, local files). Not needed for most cloud sources. Standard mode is shared and supports DirectQuery; personal mode is for one user and Import refresh only.

Q50. What is Row-Level Security? Static vs dynamic?

RLS restricts the rows a user can see using roles with DAX filters. Static RLS hard-codes values in each role (e.g. [City] = "Pune"). Dynamic RLS uses the user's identity, e.g. [Manager Email] = USERPRINCIPALNAME(), often with a security mapping table, so one role serves all users.

Q51. Does RLS apply to workspace members?

No. RLS applies to users with Viewer permission (or users consuming via apps/shares). Workspace Admins, Members and Contributors can see all data.

Q52. What is an app in Power BI Service?

A packaged collection of reports and dashboards from a workspace, published to specific audiences. It separates development (workspace) from consumption and lets you update content in a controlled way.

Q53. How do you improve a slow Power BI report?

Identify slow visuals with Performance Analyzer; reduce model size (remove columns, reduce cardinality, split date-time); use a star schema with single-direction relationships; replace calculated columns with measures or Power Query columns; optimise DAX (variables, avoid FILTER on large tables); reduce the number of visuals per page; use query reduction options; for large data consider incremental refresh or aggregations.

Q54. What is incremental refresh?

A policy that partitions a table by date and refreshes only recent data (e.g. last 10 days) while keeping historical partitions. It needs RangeStart/RangeEnd Date/Time parameters, and works best with a source that supports query folding.

Q55. What is the difference between a semantic model and a report?

The semantic model (formerly dataset) contains data, relationships, measures and security. A report is the visual layer built on a semantic model. Many reports can share one semantic model.