Microsoft DP-700 Practice Test 8
Question 70
A report is shared to a stakeholder but its Direct Lake data cannot be queried directly from OneLake. What should be explained?
- Grant access to the relevant supported item and dataset without a general workspace membership.
- Grant the vendor the necessary item-level access plus the supported query-data permission.
- Grant the appropriate underlying OneLake permission only when direct data access is required.
- Use the pipeline's supported item execution permission rather than broad workspace edit rights.
Explanation
Correct Answer
C:Read access to a report's metadata or rendered visuals does not automatically authorize direct reads of its underlying files. Attempt a direct OneLake read using the stakeholder identity.
Incorrect Answers
A:Item-scoped permission avoids visibility and editing of unrelated jobs and assets.
B:Item sharing and data query rights are separable; workspace-wide roles are unnecessary for a single item.
D:Execute and Write are distinct item-level capabilities where supported; editing permission is unnecessary merely to run a job.
Question 71
A developer needs to execute one pipeline but must not alter its definition. Which control should be evaluated?
- Configure the supported SQL analytics endpoint data permission in addition to basic item visibility.
- Use the pipeline's supported item execution permission rather than broad workspace edit rights.
- Provide report-consumption permissions without unnecessary direct OneLake file grants.
- Check the warehouse's data-plane SQL permissions rather than only workspace visibility.
Explanation
Correct Answer
B:Execute and Write are distinct item-level capabilities where supported; editing permission is unnecessary merely to run a job. Test execution and confirm pipeline editing is denied.
Incorrect Answers
A:Item Read permission alone can expose metadata without automatically conferring SQL SELECT on the lakehouse endpoint.
C:Artifact consumption and underlying file access are independent; broad data grants can leak information beyond the approved report.
D:Object listing is not the same as data access; SQL permissions determine allowed table operations for the identity.
Question 72
A shared lakehouse item is visible but SQL queries report no permission. What should be granted if the use case requires SQL data access?
- Review and revoke the specific item's direct share or permission assignment.
- Configure the supported SQL analytics endpoint data permission in addition to basic item visibility.
- Grant only Read all with Apache Spark and general item visibility while leaving SQL analytics endpoint query permissions unchanged.
- Provide report-consumption permissions without unnecessary direct OneLake file grants.
Explanation
Correct Answer
B:Item visibility alone does not grant the SQL access needed for T-SQL queries. Verify the recipient has the appropriate Read all with SQL analytics endpoint permission or granular endpoint SELECT access, and confirm the SQL query succeeds only for authorized objects.
Incorrect Answers
A:Direct item permission may survive changes to workspace membership; remediation must cover both grant paths. Item Read permission alone can expose metadata without automatically conferring SQL SELECT on the lakehouse endpoint. Run a scoped SELECT under the recipient account.
C:Read all with Apache Spark authorizes the Spark/OneLake access path, not the separate SQL analytics endpoint data query permission. A user may still be unable to run the required T-SQL query until the appropriate SQL-specific permission is granted.
D:Artifact consumption and underlying file access are independent; broad data grants can leak information beyond the approved report. Item Read permission alone can expose metadata without automatically conferring SQL SELECT on the lakehouse endpoint. Run a scoped SELECT under the recipient account.
Question 73
A report consumer must not receive raw file read access even though they can view a published report. What design should be preserved?
- Provide report-consumption permissions without unnecessary direct OneLake file grants.
- Grant the vendor the necessary item-level access plus the supported query-data permission.
- Configure the supported SQL analytics endpoint data permission in addition to basic item visibility.
- Use the pipeline's supported item execution permission rather than broad workspace edit rights.
Explanation
Correct Answer
A:Artifact consumption and underlying file access are independent; broad data grants can leak information beyond the approved report. Confirm report rendering works while direct file access fails.
Incorrect Answers
B:Item sharing and data query rights are separable; workspace-wide roles are unnecessary for a single item.
C:Item Read permission alone can expose metadata without automatically conferring SQL SELECT on the lakehouse endpoint.
D:Execute and Write are distinct item-level capabilities where supported; editing permission is unnecessary merely to run a job.
Question 74
An operator must share one eventhouse query result with an external reviewer but not the entire workspace. Which boundary fits?
- Grant access to the relevant supported item and dataset without a general workspace membership.
- Use the pipeline's supported item execution permission rather than broad workspace edit rights.
- Grant the appropriate underlying OneLake permission only when direct data access is required.
- Provide report-consumption permissions without unnecessary direct OneLake file grants.
Explanation
Correct Answer
A:Item-scoped permission avoids visibility and editing of unrelated jobs and assets. Validate effective permissions for the external reviewer.
Incorrect Answers
B:Execute and Write are distinct item-level capabilities where supported; editing permission is unnecessary merely to run a job.
C:Read access to a report's metadata or rendered visuals does not automatically authorize direct reads of its underlying files.
D:Artifact consumption and underlying file access are independent; broad data grants can leak information beyond the approved report.
Question 75
A user was removed from a workspace but can still access one separately shared item. Where should access be revoked?
- Use the pipeline's supported item execution permission rather than broad workspace edit rights.
- Provide report-consumption permissions without unnecessary direct OneLake file grants.
- Configure the supported SQL analytics endpoint data permission in addition to basic item visibility.
- Review and revoke the specific item's direct share or permission assignment.
Explanation
Correct Answer
D:Direct item permission may survive changes to workspace membership; remediation must cover both grant paths. Test access again after removing the item-level grant.
Incorrect Answers
A:Execute and Write are distinct item-level capabilities where supported; editing permission is unnecessary merely to run a job.
B:Artifact consumption and underlying file access are independent; broad data grants can leak information beyond the approved report.
C:Item Read permission alone can expose metadata without automatically conferring SQL SELECT on the lakehouse endpoint.
Question 76
An analyst can open a warehouse item but cannot run its intended queries. What should the owner inspect?
- Use the pipeline's supported item execution permission rather than broad workspace edit rights.
- Grant the vendor the necessary item-level access plus the supported query-data permission.
- Provide report-consumption permissions without unnecessary direct OneLake file grants.
- Check the warehouse's data-plane SQL permissions rather than only workspace visibility.
Explanation
Correct Answer
D:Object listing is not the same as data access; SQL permissions determine allowed table operations for the identity. Reproduce the query with an impersonated least-privileged identity.
Incorrect Answers
A:Execute and Write are distinct item-level capabilities where supported; editing permission is unnecessary merely to run a job.
B:Item sharing and data query rights are separable; workspace-wide roles are unnecessary for a single item.
C:Artifact consumption and underlying file access are independent; broad data grants can leak information beyond the approved report.
Question 77
Regional analysts must query the same table but see only rows for their assigned region. Which control addresses that requirement?
- Evaluate the effective union of OneLake security grants across the user's roles.
- Store the data as a supported Delta table in the proper Tables hierarchy before applying table security.
- Add a dynamic data masking policy to the warehouse email column using an appropriate mask function.
- Define and test row-level security based on the analyst's allowed region.
Explanation
Correct Answer
D:RLS evaluates which records a principal may access, instead of hiding an entire table or just masking a displayed value. Run identical queries as analysts from two regions.
Incorrect Answers
A:A second grant can expand visibility despite a restrictive first role; effective permissions must be tested together.
B:OneLake table security requires a valid table representation; an arbitrary file folder is not automatically eligible for RLS or CLS.
C:Dynamic data masking alters query-result presentation for nonprivileged users while leaving stored column values unchanged.
Question 78
A payroll table can be read, but salary columns must not be exposed to support staff. Which control is most direct?
- Apply supported column-level security restricting the salary columns.
- Add a dynamic data masking policy to the warehouse email column using an appropriate mask function.
- Evaluate the effective union of OneLake security grants across the user's roles.
- Store the data as a supported Delta table in the proper Tables hierarchy before applying table security.
Explanation
Correct Answer
A:CLS prevents unauthorized column access; merely masking the UI or filtering rows does not enforce the same column boundary. Query protected columns under the support role.
Incorrect Answers
B:Dynamic data masking alters query-result presentation for nonprivileged users while leaving stored column values unchanged.
C:A second grant can expand visibility despite a restrictive first role; effective permissions must be tested together.
D:OneLake table security requires a valid table representation; an arbitrary file folder is not automatically eligible for RLS or CLS.
Question 79
An entire sensitive table must be inaccessible to a group while other tables remain readable. Which data security scope matches?
- Apply object-level access control to the sensitive table.
- Separate Read-based RLS or CLS policies from any ReadWrite grant model.
- Add a dynamic data masking policy to the warehouse email column using an appropriate mask function.
- Evaluate the effective union of OneLake security grants across the user's roles.
Explanation
Correct Answer
A:Object-level permission removes the unauthorized table from the user's accessible data rather than filtering its cells. Test access to both allowed and denied tables.
Incorrect Answers
B:OneLake RLS and CLS are not configured on ReadWrite roles, so a single unrestricted writer role cannot enforce those reader filters.
C:Dynamic data masking alters query-result presentation for nonprivileged users while leaving stored column values unchanged.
D:A second grant can expand visibility despite a restrictive first role; effective permissions must be tested together.
Question 80
Two OneLake roles separately grant different rows, potentially broadening what a user sees. What should the engineer review?
- Evaluate the effective union of OneLake security grants across the user's roles.
- Add a dynamic data masking policy to the warehouse email column using an appropriate mask function.
- Define and test row-level security based on the analyst's allowed region.
- Apply object-level access control to the sensitive table.
Explanation
Correct Answer
A:A second grant can expand visibility despite a restrictive first role; effective permissions must be tested together. Inspect the user's effective row access with both memberships.
Incorrect Answers
B:Dynamic data masking alters query-result presentation for nonprivileged users while leaving stored column values unchanged.
C:RLS evaluates which records a principal may access, instead of hiding an entire table or just masking a displayed value.
D:Object-level permission removes the unauthorized table from the user's accessible data rather than filtering its cells.