View Full Databricks Certified Data Engineer Associate Exam Dumps and Practice Test Dumps.

 

Question 301

A data engineer needs to store non-tabular files such as images, PDFs, and raw JSON files in a governed Databricks environment. Which Unity Catalog capability is designed for this purpose?

1. Volumes
2. SQL warehouses
3. Job clusters
4. Materialized views

Correct Answer: 1

Explanation

Unity Catalog volumes provide governed storage for files that are not necessarily organized as tables. They are useful for managing files such as images, documents, JSON files, and other unstructured or semi-structured content while applying access controls through Unity Catalog. SQL warehouses provide SQL compute, job clusters provide workflow compute, and materialized views contain query results. Therefore, volumes are the appropriate capability when a data engineering team needs governed access to non-tabular files within Databricks.

Question 302

A company needs to allow Databricks users to access a specific cloud storage path through centrally managed permissions. Which Unity Catalog objects should be configured to establish this governed access?

1. Views and temporary tables
2. Storage credential and external location
3. Git branches and repositories
4. SQL warehouse and cluster policy

Correct Answer: 2

Explanation

A storage credential provides the authentication configuration needed to access cloud storage, while an external location associates that authenticated access with a specific storage path. Together, these Unity Catalog objects provide a governed approach to cloud storage access. Views and temporary tables are query objects, Git branches manage source code, and SQL warehouses and cluster policies concern compute. Therefore, configuring a storage credential and external location is the appropriate approach for centrally governed access to a specific cloud storage path.

Question 303

A data engineer wants to allow a service account to read a production table while preventing it from modifying the data. Which permission should be granted for the table?

1. MODIFY
2. OWN
3. SELECT
4. CREATE

Correct Answer: 3

Explanation

The SELECT privilege allows an identity to read data from a table without granting permissions to modify the table contents. This follows the principle of least privilege when an automated process only needs to consume production data. MODIFY provides data modification capabilities, OWN provides ownership-level control, and CREATE relates to creating objects in applicable securable contexts. Therefore, SELECT is the appropriate table privilege when a service account only needs read access.

Question 304

A data engineer is investigating why a user cannot query a table even though the user belongs to the correct project team. Which access-control chain should be checked first?

1. Git branch → notebook → cluster
2. SQL warehouse → query → dashboard
3. Spark task → stage → executor
4. Catalog → schema → table privileges

Correct Answer: 4

Explanation

Unity Catalog permissions follow the hierarchy of governed objects, including catalogs, schemas, and tables. When a user cannot query a table, the engineer should verify that the user or group has the required privileges at the relevant levels and that the necessary permissions have been granted. Git branches, SQL warehouses, and Spark execution components do not determine table authorization. Therefore, checking the catalog, schema, and table privilege chain is appropriate when troubleshooting governed data access.

Question 305

A data engineering team wants to run a workload using compute that is created specifically for the job and can be terminated after the workload completes. Which option is most appropriate?

1. Job cluster
2. All-purpose compute
3. External location
4. Unity Catalog volume

Correct Answer: 1

Explanation

Job clusters are intended for automated workloads and can be created specifically for job execution. Depending on configuration, the compute can terminate after the job completes, helping isolate production workloads from interactive development activity. All-purpose compute is generally intended for interactive work and can remain available for repeated use. External locations and volumes are storage governance capabilities rather than compute resources. Therefore, a job cluster is appropriate when compute should be dedicated to an automated workload.

Question 306

A developer is experimenting interactively with a notebook and repeatedly changing code, inspecting results, and testing transformations. Which compute option is generally intended for this workflow?

1. Job cluster
2. All-purpose compute
3. External location
4. Storage credential

Correct Answer: 2

Explanation

All-purpose compute is designed for interactive development and exploratory workloads. Developers can attach notebooks, run commands repeatedly, inspect results, and iterate on transformations without creating a new isolated job environment for every execution. Job clusters are more closely associated with automated workload execution. External locations and storage credentials provide governed access to cloud storage rather than notebook compute. Therefore, all-purpose compute is generally suitable for interactive development where engineers need an environment for repeated experimentation.

Question 307

An administrator wants to ensure that users can create compute only with approved runtime versions and worker configurations. Which capability can enforce these restrictions?

1. Data Explorer
2. Cluster policy
3. Delta Sharing
4. Query history

Correct Answer: 2

Explanation

Cluster policies allow administrators to define restrictions on compute configuration. Policies can be used to control settings such as runtime versions, worker configurations, instance types, and other supported parameters. This helps organizations maintain standards and control resource usage. Data Explorer supports data discovery, Delta Sharing supports governed data sharing, and query history provides information about executed queries. Therefore, a cluster policy is the appropriate capability for enforcing approved compute configurations.

Question 308

A SQL workload has significantly higher demand during business hours and much lower demand overnight. Which capability can help adjust SQL compute resources according to workload requirements?

1. Generated columns
2. Git integration
3. Warehouse scaling
4. Table constraints

Correct Answer: 3

Explanation

Warehouse scaling can adjust SQL compute resources in response to workload requirements, depending on the configured warehouse behavior. This is useful for workloads where query demand changes significantly throughout the day. During periods of higher demand, additional resources can help accommodate increased concurrency, while lower-demand periods can avoid unnecessary resource consumption. Generated columns derive values, Git integration manages source code, and table constraints enforce data rules. Therefore, warehouse scaling is the capability most relevant to changing SQL workload demand.

Question 309

A data engineer wants to accelerate supported SQL and DataFrame workloads using Databricks’ optimized execution engine. Which capability should be considered?

1. Photon
2. Git folders
3. Unity Catalog volumes
4. Service principals

Correct Answer: 1

Explanation

Photon is Databricks’ optimized execution engine designed to accelerate supported SQL and DataFrame workloads. It uses an optimized processing approach to improve query and data-processing performance while allowing users to continue working with familiar Databricks and Spark interfaces. Git folders provide source-control integration, Unity Catalog volumes govern file access, and service principals provide application identities. Therefore, Photon is the capability specifically associated with improving execution performance for supported workloads.

Question 310

A data engineering team needs to authenticate an automated application to Databricks without tying the workload to a specific employee. Which identity is designed for this purpose?

1. Temporary view
2. Service principal
3. SQL warehouse
4. DataFrame

Correct Answer: 2

Explanation

A service principal is an identity intended for applications and automated processes. It allows an organization to assign permissions to a workload independently of an employee’s personal account. This is useful for production pipelines, deployment automation, and other machine-to-machine operations. Temporary views and DataFrames are data-processing constructs, while a SQL warehouse provides SQL compute. Therefore, a service principal is the appropriate identity for authenticating automated applications without relying on an individual user’s account.

Question 311

A team wants to grant the same Unity Catalog permissions to analysts who regularly join and leave the project. Which approach reduces the need to modify individual table permissions every time membership changes?

1. Grant ownership to every analyst
2. Use group-based permissions
3. Store permissions in notebook variables
4. Give every analyst administrator access

Correct Answer: 2

Explanation

Group-based permissions allow administrators to grant privileges to a group rather than repeatedly assigning the same permissions to individual users. When analysts join or leave the project, administrators can update group membership while keeping the underlying data grants unchanged. Granting ownership or administrator access provides broader privileges than necessary and violates least-privilege principles. Notebook variables are not a governance mechanism. Therefore, group-based access is an efficient way to manage shared Unity Catalog permissions.

Question 312

A data engineer is reviewing a Delta table and wants to know the exact operations associated with earlier table versions. Which command should be used?

1. DESCRIBE TABLE
2. DESCRIBE DETAIL
3. DESCRIBE HISTORY
4. SHOW COLUMNS

Correct Answer: 3

Explanation

DESCRIBE HISTORY displays the transaction history of a Delta table, including versions and operations associated with changes. This makes it useful for investigating when modifications occurred and understanding the sequence of operations that produced historical table states. DESCRIBE TABLE focuses primarily on schema information, while DESCRIBE DETAIL provides broader metadata and storage-related information. SHOW COLUMNS lists column information. Therefore, DESCRIBE HISTORY is the appropriate command for reviewing historical Delta operations.

Question 313

A data engineer wants to inspect detailed metadata about a Delta table, including its storage location and other table-level characteristics. Which command is most appropriate?

1. DESCRIBE DETAIL
2. DROP TABLE
3. DELETE FROM
4. CREATE VIEW

Correct Answer: 1

Explanation

DESCRIBE DETAIL provides detailed information about a Delta table, including metadata such as its location, format, and other table-level characteristics. Engineers can use it when investigating how a table is stored or when they need more information than a basic schema description provides. DROP TABLE removes a table, DELETE FROM modifies table data, and CREATE VIEW creates a query object. Therefore, DESCRIBE DETAIL is the appropriate command for examining detailed table metadata.

Question 314

A development team wants to create an isolated copy of a Delta table for testing while keeping the source table unchanged. Which capability provides a direct way to create such a table copy?

1. VACUUM
2. CLONE
3. OPTIMIZE
4. CHECK constraint

Correct Answer: 2

Explanation

Delta Lake cloning allows engineers to create a separate table based on an existing Delta table. This is useful for development, testing, experimentation, or other situations where a team needs a separate table representation without rebuilding the table manually. VACUUM removes eligible obsolete files, OPTIMIZE improves data layout, and CHECK constraints enforce data-quality rules. Therefore, CLONE is the capability most directly associated with creating a separate copy of a Delta table for testing purposes.

Question 315

A team needs an independent physical copy of a Delta table because the copy must remain self-contained even if the original table’s underlying files are later removed. Which clone type is appropriate?

1. Temporary clone
2. Shallow clone
3. Deep clone
4. View clone

Correct Answer: 3

Explanation

A deep clone copies the underlying data and metadata needed to create an independent Delta table. Because the cloned table has its own physical data files, it can remain independent from the source table’s underlying storage. A shallow clone instead references the source table’s data files, so it does not provide the same level of physical independence. Temporary and view clone options are not the relevant Delta cloning types. Therefore, deep clone is appropriate when an independent physical copy is required.

Question 316

A data engineer needs to create a lightweight Delta table copy for short-term testing and wants to avoid immediately copying all source data files. Which approach is appropriate?

1. Deep clone
2. Shallow clone
3. Full table export to Python
4. Materialized view

Correct Answer: 2

Explanation

A shallow clone creates a new Delta table while initially referencing the source table’s underlying data files. This can make the cloning operation faster and reduce immediate additional storage requirements compared with a deep clone. A deep clone creates an independent physical copy of the data. Exporting a full table to Python is unsuitable for large datasets, and a materialized view is designed for query results rather than table cloning. Therefore, shallow clone is appropriate for a lightweight testing copy.

Question 317

A streaming pipeline receives events from a source that may retry delivery after a temporary failure. The same event can therefore appear multiple times. Which design principle should the target pipeline follow?

1. Idempotent processing
2. Random row deletion
3. Driver-side collection
4. Manual file editing

Correct Answer: 1

Explanation

Idempotent processing ensures that repeated processing of the same event does not produce an incorrect or duplicated final result. Streaming systems can encounter retries or repeated deliveries, so pipelines should use stable identifiers, deduplication, merge logic, or other mechanisms appropriate to the workload. Random deletion could remove legitimate data, collecting records on the driver does not solve duplicate delivery, and manual file editing is unsuitable for automated pipelines. Therefore, idempotent processing is the appropriate design principle for handling repeated event delivery.

Question 318

A streaming transaction dataset needs to be enriched with customer information stored in a static Delta table. Which join pattern is suitable?

1. Stream-stream join only
2. Stream-static join
3. Cross join with every table
4. Git merge

Correct Answer: 2

Explanation

A stream-static join combines a continuously arriving streaming dataset with a static dataset. In this scenario, transaction events can be enriched with customer attributes from a static Delta table as the streaming records are processed. A stream-stream join is designed for two streaming inputs and has different state-management considerations. A cross join can generate unnecessary combinations, while Git merge relates to source-code management. Therefore, a stream-static join is the appropriate pattern for enriching streaming transactions with static customer data.

Question 319

A streaming query maintains aggregate results that can change when new events arrive, and the downstream sink should receive only the rows whose aggregate values changed. Which output mode should be selected?

1. Append
2. Complete
3. Update
4. Overwrite

Correct Answer: 3

Explanation

Update mode outputs rows whose results have changed since the previous trigger. This is appropriate for streaming aggregations where incoming records can modify existing aggregate values. Append mode is intended for newly finalized rows that will not change, while complete mode outputs the entire current result table. Overwrite is not a standard Structured Streaming output mode for this purpose. Therefore, update mode is suitable when the downstream sink should receive only aggregate rows that have changed.

Question 320

A data engineer wants to prevent invalid records with a negative quantity from being inserted into a Delta table. Which table-level rule can enforce this requirement?

1. Git branch
2. SQL warehouse
3. CHECK constraint
4. Cluster policy

Correct Answer: 3

Explanation

A CHECK constraint can enforce a Boolean condition on data written to a Delta table. For example, a constraint requiring quantity >= 0 can prevent records containing negative quantities from being accepted. This provides table-level enforcement of a defined data-quality rule. Git branches manage source-code changes, SQL warehouses provide SQL compute, and cluster policies control compute configurations. Therefore, a CHECK constraint is the appropriate mechanism for preventing invalid negative quantity values from entering the table.