Data Analyst Associate · Domain 2: Managing Data · 8% of the exam
Task 2.1: Use Unity Catalog to discover, query, and manage certified datasets; Use the Catalog Explorer to tag a data asset with governed tags and view its lineage, including external lineage
Finding trusted data through search, comments, insights and certification, applying governed tags and their allowed values, and reading lineage that includes systems outside Databricks.
Study it
Discovering certified data, governed tags and lineage
Lesson coming
Sample questions
Try each one before opening the answer. Every option is explained, with the Databricks documentation page that proves it.
Question 1 · choose 1
Ravi runs a query on the certified table main.finance.revenue in the SQL editor and gets an INSUFFICIENT_PERMISSIONS error. He holds no privileges on the catalog main or the schema main.finance. He uses Request access from the error message to ask for SELECT on the table. What happens next?
ADatabricks adds USE CATALOG and USE SCHEMA requests for the parent objects
BThe request is refused because Ravi lacks BROWSE on the catalog main
CThe request stays open in the reviewer's queue until a reviewer acts on it, however long that takes
DA reviewer can approve it only by granting SELECT directly to Ravi
Show the answer and why
ADatabricks adds USE CATALOG and USE SCHEMA requests for the parent objects
Correct
When a user requests a privilege such as SELECT, Databricks checks the prerequisite privileges and generates additional requests for the missing ones, routed to the parent objects' approvers.
BThe request is refused because Ravi lacks BROWSE on the catalog main
Incorrect
The SQL editor error offers Request access pre-filled with the referenced tables; BROWSE is what lets users navigate the catalog tree in Catalog Explorer.
CThe request stays open in the reviewer's queue until a reviewer acts on it, however long that takes
Incorrect
Unanswered requests expire after 7 days, with email reminders after 3 days.
DA reviewer can approve it only by granting SELECT directly to Ravi
Incorrect
Reviewers can also provision access by adding the principal to a group.
Request for access lets users ask for Unity Catalog privileges from Catalog Explorer, the SQL editor and notebooks, dashboards, Genie One and Genie Agents. For a SELECT request Databricks checks USE CATALOG and USE SCHEMA and generates the missing requests. Requests are routed to the configured destinations; reviewers provision by direct grant or group membership or decline with a reason; open requests expire after 7 days.
Maya holds SELECT on main.gold.sales_summary but not MODIFY. A Tableau dashboard outside Databricks reads that table, and the platform team has created an external metadata object for the dashboard in Catalog Explorer and granted Maya MODIFY on it. Maya wants the dashboard to appear downstream of the table in the lineage graph. What else does she need?
AMODIFY on main.gold.sales_summary in addition to SELECT
BCREATE EXTERNAL METADATA on the metastore
COwnership of main.gold.sales_summary, transferred to her
DNothing more; she already holds what is needed
Show the answer and why
AMODIFY on main.gold.sales_summary in addition to SELECT
Incorrect
Write privileges on the Unity Catalog object are required for an upstream relationship to it, not a downstream one.
BCREATE EXTERNAL METADATA on the metastore
Incorrect
That privilege is needed to create the external metadata object, which already exists.
COwnership of main.gold.sales_summary, transferred to her
Incorrect
Ownership is not required; the documented requirement for a downstream relationship is a read privilege on the object.
DNothing more; she already holds what is needed
Correct
She needs MODIFY on the external metadata object and, for a downstream relationship, a read privilege such as SELECT on the table: she has both.
External lineage adds systems outside Databricks to Unity Catalog lineage. Creating an external metadata object needs CREATE EXTERNAL METADATA on the metastore; defining a relationship needs MODIFY on that object, plus read privileges on the Unity Catalog object for a downstream relationship or write privileges for an upstream one.
Analysts keep asking what the columns of main.sales.orders mean. The table owner wants the answer visible where people browse data in Catalog Explorer. What should she do?
ARename the columns to longer names
BAdd a README notebook next to the table
CAdd comments to the table and its columns
DGrant the analysts MODIFY on the table
Show the answer and why
ARename the columns to longer names
Incorrect
Renaming changes the schema for every query; descriptions belong in comments.
BAdd a README notebook next to the table
Incorrect
A notebook is a separate workspace object; comments are shown on the table and its columns.
CAdd comments to the table and its columns
Correct
Comments on catalogs, schemas, tables and columns help users discover and understand data.
DGrant the analysts MODIFY on the table
Incorrect
MODIFY lets users change data; it does not explain the columns.
Discovering data in Databricks relies on metadata such as comments on catalogs, schemas, tables and columns, which Catalog Explorer shows alongside each object.
An analyst submitted several access requests for Unity Catalog tables and wants to see where each one stands. Where does she look?
AOn the Requests page in Catalog Explorer, under My requests
BIn the query history of the SQL warehouse
CIn the table's Lineage tab
DIn the workspace admin settings, under the Notifications page
Show the answer and why
AOn the Requests page in Catalog Explorer, under My requests
Correct
Requesters track their requests on the Requests page in Catalog Explorer; My requests lists the ones she submitted.
BIn the query history of the SQL warehouse
Incorrect
Query history lists queries, not access requests.
CIn the table's Lineage tab
Incorrect
Lineage shows data flow, not the status of access requests.
DIn the workspace admin settings, under the Notifications page
Incorrect
Access requests are tracked in Catalog Explorer.
Requesters, reviewers and metastore admins track access requests from the Requests page in Catalog Explorer: My requests for requests you submitted, Review queue for requests routed to you, and All requests for metastore admins.
In Catalog Explorer, an analyst creates an external metadata object for a report in an in-house BI tool that is not in the System type list. What should she select as the System type?
ALeave the field empty, since it is optional
BThe closest listed BI tool
CCustom, as the system is unlisted
DUnity Catalog
Show the answer and why
ALeave the field empty, since it is optional
Incorrect
System type is a required field of an external metadata object.
BThe closest listed BI tool
Incorrect
Picking another system mislabels the asset; the form offers a choice for unlisted systems.
CCustom, as the system is unlisted
Correct
If your system is not in the list of common external systems, select Custom.
DUnity Catalog
Incorrect
The object represents an entity outside Unity Catalog, so its system is the external one.
An external metadata object needs a name, a system type and an entity type, with optional URL, description, columns and properties. The system type comes from a list of common external data and BI systems, or Custom if yours is not listed.