DBX-DAA Study Notes in Questions: Understanding of the Databricks Data Intelligence Platform

5 questions, one domain: Understanding of the Databricks Data Intelligence Platform from the Databricks Data Analyst Associate (DBX-DAA) exam. It makes up about 11% of the exam. Pick your answer, then check it against the explanation.

Databricks Data Analyst Associate (DBX-DAA) · Understanding of the Databricks Data Intelligence Platform · 5 questions with answers and explanations

🧭 Reading Understanding of the Databricks Data Intelligence Platform questions

Platform questions ask what a part of Databricks is and where it runs, often from the point of view of an analyst who is new to it. The split between the control plane, compute and storage usually decides. Strong options name the component that does the job; weaker ones mix up where data lives with where queries run.

Three things to look for

• The web app and services run in the control plane; data stays in your cloud storage, and serverless compute runs in Databricks' account.
• Delta Lake adds ACID transactions on files in object storage, so SQL and other tools share one copy.
• Unity Catalog, Catalog Explorer and Marketplace each answer a different question: who may see it, what exists, and what others offer.

Question 1 of 5

A company runs a workspace for finance analysts and a second workspace for marketing analysts in the same cloud region, both attached to the same Unity Catalog metastore. A marketing analyst who was granted SELECT on finance.reports.q3_summary opens the marketing workspace. What does the analyst see?

AThe same table, because the metastore holds the metadata and permissions that every attached workspace shares

BA read-only snapshot taken when the analyst was granted access, which is not refreshed afterwards

CThe table only after a workspace admin recreates the grant separately inside the marketing workspace

DNothing, because each workspace keeps its own catalogs and the table must be copied into the marketing workspace

Show answer

✅ Answer: A

Difficulty: applied · Topic: Unity Catalog

One metastore, many workspaces
The Unity Catalog metastore is the top-level container for metadata and permissions. A region has one metastore that all workspaces in that region can share, so catalogs, schemas, tables and grants are defined once instead of per workspace.
• Objects are addressed as catalog.schema.object in every attached workspace.
• A privilege granted on an object applies wherever the user connects from, unless the catalog is bound to specific workspaces (workspace-catalog binding).
• Compute still runs per workspace, but the data view and the governance are the same.

Why each option works or fails
• ✓ A. Shared metastore (correct): the table and the analyst's SELECT grant live in the metastore, so the marketing workspace shows the same table without copying data.
• ✗ B. Snapshot: the analyst queries the live table, and Unity Catalog does not create a copy at grant time.
• ✗ C. Recreated grant: privileges are stored in the metastore and apply in every attached workspace; nothing is granted again per workspace.
• ✗ D. Per-workspace catalogs: catalogs belong to the metastore, not to one workspace, so no copy is needed.

Question 2 of 5

A table was created in a notebook as CREATE TEMPORARY VIEW recent_orders AS SELECT .... A colleague cannot find recent_orders in Catalog Explorer and gets "table or view not found" in her own notebook. Why?

AA temporary view is scoped to the session or query that created it and is not registered in any schema or catalog

BTemporary views are registered in the metastore but are visible only after the workspace admin certifies them

CThe view was dropped automatically because Unity Catalog allows views only inside the default catalog

DViews cannot be shared between users, so every analyst must recreate the view in a separate workspace

Show answer

✅ Answer: A

Difficulty: applied · Topic: Views

Scope of views
A view created with CREATE VIEW is registered in a schema (catalog.schema.view) and can be queried from anywhere you have permission. A temporary view is not registered in any schema or catalog:
• in notebooks and jobs it is scoped to the notebook and disappears when the notebook detaches from compute;
• in Databricks SQL it is scoped to the query and cannot be referenced from other queries, even in the same dashboard.
To share logic, create a persistent view or a materialized view in a schema and grant access.

Why each option works or fails
• ✓ A. Session scope (correct): the temporary view exists only for the notebook or query that defined it, so Catalog Explorer and other users never see it.
• ✗ B. Certification: certification is a tag for trust signals and does not control the visibility of views.
• ✗ C. Default catalog only: views can be created in any catalog and schema where the user has privileges.
• ✗ D. Cannot be shared: persistent views are designed for sharing and can be granted to other users and groups.

Question 3 of 5

A sales analyst wants to find a ready-made third-party dataset that can be queried next to company tables in Unity Catalog. Which Databricks capability is built for this?

ADatabricks Marketplace, an open exchange where providers publish datasets, notebooks and AI models for consumers

BLakeflow Jobs, a scheduler that publishes the company's tables to external buyers on a cron schedule

CSQL alerts, a notification feature that emails analysts whenever a new public dataset appears

DGenie Code, an assistant that finds and downloads open datasets from the internet when asked in natural language, then registers them as tables

Show answer

✅ Answer: A

Difficulty: applied · Topic: Databricks Marketplace

Marketplace
Databricks Marketplace is an open exchange where data providers, software vendors and partners publish offerings: datasets (typically catalogs of tabular data, also volumes), notebooks, AI models, apps and MCP servers. Consumers browse listings, request access, and use shared data through OpenSharing (the sharing protocol formerly called Delta Sharing in many docs). In a Unity Catalog workspace, accessed data appears as a read-only catalog that analysts query like any other.

Why each option works or fails
• ✓ A. Marketplace (correct): discover and connect to third-party data products directly from the workspace.
• ✗ B. Lakeflow Jobs: it orchestrates workloads and does not publish listings to buyers.
• ✗ C. SQL alerts: alerts notify on query results meeting a condition and do not monitor available datasets.
• ✗ D. Genie Code: it helps write and debug code and does not acquire external datasets.

Question 4 of 5

An analyst gets a table with salary, home address and several business columns. Sales managers need only the business columns, and the team wants them to query a governed object without being granted SELECT on the table itself. What fits?

AA table comment that tells the managers which columns they should not query, since comments restrict columns

BA view that selects only the business columns, with SELECT granted on the view and not on the base table

CA copy of the table in a new schema that keeps every column, with the salary column renamed

DA temporary view created in the analyst's notebook, which the managers can open from their own sessions

Show answer

✅ Answer: B

Difficulty: applied · Topic: Views, Unity Catalog

Views as a restricted window
A view stores a query. Because it is a securable object in a schema, you can grant SELECT on the view alone, and users read only the columns and rows the query exposes. Unity Catalog lets view readers query it without privileges on the base table when the view owner has access to it (together with USE CATALOG and USE SCHEMA on the parents).
• Column subset: list only permitted columns in the SELECT.
• Row or column masking by group: a dynamic view.
• Temporary views are session-scoped and are not governed objects.

Why each option works or fails
• ✗ A. Comment: comments document data and enforce nothing, so the managers could still query salary if they could read the table.
• ✓ B. View with its own grant (correct): managers get a governed object that exposes only the permitted columns, with no access to the base table.
• ✗ C. Full copy: it keeps the sensitive columns and creates a duplicate that drifts; renaming a column does not hide its values.
• ✗ D. Temporary view: it exists only in the creating session and is not registered in a schema, so others cannot open or be granted it.

Question 5 of 5

A team lead asks which Databricks component an analyst uses to run SQL queries and build dashboards against lakehouse data without managing clusters, as opposed to the component data engineers use to build ingestion and transformation pipelines. Which pairing is correct?

AAnalysts use Mosaic AI to run SQL queries; data engineers use Unity Catalog to transform and load the data into tables

BAnalysts use Lakeflow Jobs to run ad hoc queries; data engineers use Databricks SQL to schedule and orchestrate their pipelines

CAnalysts use Lakebase to build dashboards; data engineers use Delta Sharing to run scheduled transformations in notebooks

DAnalysts use Databricks SQL with SQL warehouses; data engineers use Lakeflow for ingestion, pipelines and jobs

Show answer

✅ Answer: D

Difficulty: applied · Topic: Databricks SQL, Lakeflow

Who uses what
• Databricks SQL is the data warehousing experience: SQL editor, SQL warehouses (the compute that executes queries), AI/BI dashboards, alerts and Genie. Analysts work here.
• Lakeflow is the data engineering solution: Lakeflow Connect (ingestion connectors), Lakeflow Declarative Pipelines (transformations) and Lakeflow Jobs (orchestration).
• Unity Catalog governs data and AI assets for everyone; Mosaic AI covers building and serving AI models and agents; Lakebase is a managed Postgres database for applications.
Older docs call Lakeflow Jobs "Workflows" and Declarative Pipelines "Delta Live Tables".

Why each option works or fails
• ✗ A. Mosaic AI and Unity Catalog: Mosaic AI is about building and serving AI, and Unity Catalog is a governance layer; neither is the SQL query interface or the loading tool.
• ✗ B. Swapped roles: Lakeflow Jobs orchestrates tasks and is not the analyst's ad hoc query tool, and Databricks SQL is not the pipeline orchestrator.
• ✗ C. Lakebase and Delta Sharing: Lakebase is a transactional Postgres database for apps, and Delta Sharing (OpenSharing) shares data across organizations rather than running transformations.
• ✓ D. Databricks SQL plus Lakeflow (correct): analysts query through SQL warehouses and dashboards, while engineers build ingestion, pipelines and orchestration with Lakeflow.

🎯 Takeaways

• One metastore per region is shared by its workspaces, so tables and grants are defined once.
• Temporary views are session-scoped and unregistered; persistent views live in a schema.
• Marketplace = exchange for third-party data, AI models and notebooks.
• Grant SELECT on a view that exposes only the needed columns to restrict access without copying data.
• Analysts = Databricks SQL + SQL warehouses; engineers = Lakeflow; governance = Unity Catalog.

📊 This set at a glance

Exam: DBX-DAA · 45 questions · 90 minutes
This set: 5 applied · 2 scenario, 1 comparison, 1 concept check, 1 troubleshooting
Topics: Unity Catalog, Views, Databricks Marketplace, Databricks SQL, Lakeflow

🔗 Related practice sets

• All DBX-DAA practice sets
• Every exam on this blog

📚 Keep studying DBX-DAA

More DBX-DAA questions: https://hyunhp.github.io/cert-bank/#exam=DBX-DAA
Study guide: https://hyunhp.github.io/cert-bank/study.html#exam=DBX-DAA

These are original practice questions, not real exam questions, and cert-bank is not affiliated with any certification vendor.

Original practice questions written for this blog, not exam dumps.