GCP-ADP Practice Test 4 — 25 Questions

Question 1

In Cloud Composer, what is the default behavior if a DAG execution takes longer than the 'dagrun_timeout' setting?

Accepted Answer

The DAG run is marked as failed.. In Cloud Composer (Apache Airflow), if a DAG run exceeds the dagrun_timeout setting, the scheduler marks the DAG run as failed. This prevents stuck runs from consuming resources indefinitely and allows alerting or retry logic to kick in.

Answer

The DAG enters a 'running' state indefinitely.

Answer

The DAG run is paused.

Answer

The scheduler automatically restarts the DAG.

Question 2

You want to share a BigQuery dataset with a partner who uses a different Google Cloud project. What is the most secure way to grant them access?

Accepted Answer

Grant the partner's service account the 'BigQuery Data Viewer' role on the dataset.. Granting the partner's service account the BigQuery Data Viewer role on the dataset is the most secure approach because it follows the principle of least privilege: the partner can read data in that specific dataset only, using their own identity, with no credential sharing. IAM roles can be granted at the dataset level, scoping access precisely and allowing easy revocation.

Answer

Make the dataset public.

Answer

Provide the partner with your project's service account key.

Answer

Export the data to a public Cloud Storage bucket.

Question 3

A data team is building a BigQuery ML model to predict customer churn directly from a BigQuery table. They want to evaluate the model's quality and understand which input features influenced its predictions. Which two BigQuery ML capabilities should they use to accomplish these goals? (Choose two.)

Accepted Answer

Use ML.EXPLAIN_PREDICT to return per-row feature attributions alongside each prediction.. ML.EVALUATE measures the churn model's predictive quality with classification metrics on held-out data, while ML.EXPLAIN_PREDICT returns per-row feature attributions that reveal which inputs drove each prediction. Together they cover both evaluation and interpretability. The other functions describe training data, training progress, or generate embeddings, none of which answers how good the model is or why it predicts churn.

Answer

Use ML.TRAINING_INFO to inspect the iteration history of the model's training run.

Answer

Use ML.GENERATE_EMBEDDING to convert customer records into vector embeddings for similarity search.

Answer

Use ML.FEATURE_INFO to list the summary statistics of each feature column used during training.

Question 4

You are identifying trends in a large dataset using BigQuery. Which SQL function is best suited to identify the top 5 products sold per region?

Accepted Answer

RANK() OVER (PARTITION BY region ORDER BY sales DESC).. RANK() OVER (PARTITION BY region ORDER BY sales DESC) assigns a rank to each product within each region based on sales, so filtering for rank <= 5 yields the top 5 products per region. This is the standard SQL window-function pattern for per-group top-N queries.

Answer

GROUP BY region, product.

Answer

SELECT TOP 5 * FROM sales.

Answer

SUM(sales) OVER (ORDER BY region).

Question 5

You manage a BigQuery dataset that contains a table with 500 TB of historical order data. Analysts frequently run queries that filter on the `order_date` column. Query costs are rising because every query scans the entire table. You want to reduce the bytes scanned and cost without changing the table name or existing SQL. What should you do?

Accepted Answer

Partition the table by the `order_date` column so queries filtering on that column scan only relevant partitions.. Partitioning a table by a column that queries commonly filter on lets BigQuery prune partitions and read only the data needed, which reduces bytes scanned and therefore cost. Because partitioning does not change the table name or require SQL changes, it satisfies all the constraints. The other choices either require SQL rewrites, do not affect scan volume, or block queries instead of optimizing them.

Answer

Set a custom query quota on the project so analysts cannot run queries that scan more than 1 TB.

Answer

Enable the BigQuery BI Engine reservation on the dataset to cache query results and lower cost.

Answer

Create a materialized view that pre-aggregates orders by date and instruct analysts to query the view instead.

Question 6

Your team runs an Apache Airflow environment on Google Cloud Managed Service for Apache Airflow. A daily ingestion DAG must wait for a partner team's separate DAG to finish loading a landing table before it can begin its own transformation. The partner DAG is in the same Airflow environment but runs on an unpredictable schedule. You want the dependency handled natively by Airflow rather than with a fixed sleep or time sensor. What should you configure?

Accepted Answer

An ExternalTaskSensor that points at the partner DAG and task. The requirement is a true task-level dependency on another DAG's completion, not a time or file heuristic. An ExternalTaskSensor reads the upstream task state from Airflow's metadata database, so the transformation proceeds only after the partner load succeeds. Time sensors, storage sensors, and API triggers either guess at completion or invert the dependency, none of which guarantees the landing table is fully populated.

Answer

A PythonOperator that calls the partner DAG's REST API trigger endpoint

Answer

A Cloud Storage sensor watching the landing bucket for new objects

Answer

A TimeSensor that waits until a predetermined clock time each day

Question 7

A data analyst has a BigQuery table named `sales` with columns `order_date` (DATE), `product_id` (STRING), and `revenue` (NUMERIC). They need to create a report that shows total revenue per product for the last 30 days, but only for products that generated more than $10,000 in that period. They want the result sorted by revenue descending. Which SQL query should they use?

Accepted Answer

SELECT product_id, SUM(revenue) AS total_revenue FROM sales WHERE order_date >= DATE_SUB(CURRENT_DATE(), INTERVAL 30 DAY) GROUP BY product_id HAVING total_revenue > 10000 ORDER BY total_revenue DESC. The correct query filters the last 30 days, groups by product, sums revenue, and then applies a HAVING clause to keep only products with total revenue above $10,000, finally sorting by that total descending. Using WHERE to filter individual revenues would incorrectly exclude products with many smaller sales, and placing HAVING after ORDER BY or ordering by a non-aggregated column would cause errors.

Answer

SELECT product_id, SUM(revenue) AS total_revenue FROM sales WHERE order_date >= DATE_SUB(CURRENT_DATE(), INTERVAL 30 DAY) AND revenue > 10000 GROUP BY product_id ORDER BY total_revenue DESC

Answer

SELECT product_id, SUM(revenue) AS total_revenue FROM sales WHERE order_date >= DATE_SUB(CURRENT_DATE(), INTERVAL 30 DAY) GROUP BY product_id ORDER BY total_revenue DESC HAVING total_revenue > 10000

Answer

SELECT product_id, SUM(revenue) AS total_revenue FROM sales WHERE order_date >= DATE_SUB(CURRENT_DATE(), INTERVAL 30 DAY) GROUP BY product_id HAVING SUM(revenue) > 10000 ORDER BY revenue DESC

Question 8

What is the benefit of using 'Scheduled Reports' in Looker?

Accepted Answer

To automatically email reports to stakeholders.. Scheduled Reports in Looker allow users to automate the delivery of dashboards, Looks, or visualizations via email or other destinations on a recurring schedule. This ensures stakeholders receive timely updates without manual intervention.

Answer

To increase the BigQuery slot limit.

Answer

To optimize the database performance.

Answer

To change the dashboard color.

Question 9

You need to ensure that deleted files in GCS can be restored for 30 days. What should you configure?

Accepted Answer

Object Versioning. Object Versioning preserves prior versions of objects when they are overwritten or deleted, allowing restoration for as long as the noncurrent versions are retained. To meet the 30-day restore requirement, you enable Object Versioning on the bucket and optionally add a lifecycle rule to delete noncurrent versions after 30 days. This directly satisfies the recoverability requirement without locking the bucket.

Answer

Cloud Storage bucket lock

Answer

Bucket Lifecycle policy

Answer

IAM

Question 10

You have data scattered across multiple projects and need to provide a unified data governance view. What is the recommended hierarchy?

Accepted Answer

One Lake, with assets added from multiple projects. Dataplex is designed so that a single Lake can contain Zones and Assets that reference resources from multiple projects, giving one unified governance and discovery view across the organization. This is the recommended hierarchy because it centralizes policy, metadata, and lineage while still allowing data to physically reside in different projects.

Answer

A single GCS bucket for all data

Answer

Multiple Lakes, one per project, and manage them independently

Answer

One Lake per project, then use a project-linking tool

Question 11

Your analytics team stores clickstream events in Cloud Storage as newline-delimited JSON files. They want to query this data with BigQuery using SQL without loading it into native BigQuery storage, and they want the table definition to refresh automatically when new files land in the bucket. What should you configure?

Accepted Answer

Create a BigQuery external table over the Cloud Storage bucket and enable automatic metadata refresh.. The scenario calls for querying Cloud Storage files with BigQuery SQL and having the table definition track new files automatically. An external table over Cloud Storage satisfies in-place querying, and automatic metadata refresh keeps the table aware of newly added objects without manual intervention. The other approaches either copy data into native storage, rely on a non-existent SQL endpoint, or move the workload to a different engine and dialect.

Answer

Create a Dataplex asset over the bucket and query it with the Dataplex SQL endpoint.

Answer

Mount the bucket with Cloud Storage FUSE on a Dataproc cluster and query it with Spark SQL.

Answer

Load the JSON files into a native BigQuery table using a recurring Cloud Composer job.

Question 12

Which TWO of the following tools allow you to interact with Dataplex? (Choose two)

Accepted Answer

Google Cloud CLI (gcloud). Option A, the Google Cloud CLI (gcloud), is correct because Dataplex provides gcloud commands (for example, gcloud dataplex lakes, zones, assets, and tasks) that let you create, manage, and query Dataplex resources from a terminal. Option D, the Google Cloud Console, is correct because Dataplex offers a graphical web UI in the Console for browsing lakes, zones, assets, and managing metadata, discovery, and data quality tasks. Option B, a local database client like DBeaver, is not correct because DBeaver connects to databases via JDBC/ODBC and does not natively manage Dataplex lakes, zones, or assets. Option C, a text editor for configuration files, is not correct because editing files alone does not provide an interface to Dataplex services; you would still need gcloud, the Console, or the API to apply them. Option E, a standard SSH client, is not correct because SSH provides remote shell access to hosts, not an interface to the Dataplex service.

Answer

A local database client like DBeaver

Answer

Text editor for configuration files

Answer

A standard SSH client

Question 13

You have a large table in BigQuery that is updated constantly. You need to perform a query that is consistent. What should you use?

Accepted Answer

Time travel / Table snapshots.. BigQuery time travel lets you query a table as it existed at any point within the last seven days, and table snapshots preserve a consistent view of the data at a specific moment. Because the table is constantly updated, querying the live table can return rows from different micro-batches; time travel or a snapshot gives you a point-in-time consistent read.

Answer

A stored procedure.

Answer

An external table.

Answer

A materialized view.

Question 14

You need to ingest data from an external API that requires an OAuth token. Which tool is best for executing this periodically?

Accepted Answer

Cloud Functions. Cloud Functions is the best fit because it can run scheduled (via Cloud Scheduler), execute custom code to perform the OAuth token exchange, call the external API, and write results to a destination like BigQuery or Cloud Storage. It supports secrets via Secret Manager and scales automatically. This makes it ideal for periodic, code-driven API ingestion requiring authentication.

Answer

Pub/Sub

Answer

BigQuery Data Transfer Service

Answer

Storage Transfer Service

Question 15

Which command-line tool do you primarily use to deploy and manage Workflows?

Accepted Answer

gcloud. The gcloud CLI is the primary tool for deploying and managing Workflows, using commands like gcloud workflows deploy, gcloud workflows execute, and gcloud workflows describe. It authenticates via gcloud auth and integrates with all Workflows management APIs. This is the documented and supported interface for lifecycle operations.

Answer

gsutil

Answer

kubectl

Answer

bq

Question 16

You are using Dataprep to prepare data. How can you share your work with a team?

Accepted Answer

Share the Flow. In Google Cloud Dataprep, a Flow is the primary container that holds all your recipes, datasets, and transformation logic. Sharing the Flow directly with team members grants them access to the entire pipeline, allowing them to view, edit, and run the same data preparation steps. This is the built-in collaboration mechanism designed for team workflows, and it respects IAM permissions on the underlying project and datasets.

Answer

Copy the JSON definition

Answer

Share the underlying GCS file

Answer

Export to a CSV

Question 17

A Cloud Composer environment is experiencing high latency in the Airflow UI. What is a common cause and mitigation for this?

Accepted Answer

Excessive DAG complexity; optimize/simplify DAG code.. Cloud Composer's Airflow UI (webserver) performance degrades when the scheduler and webserver must parse and render an excessive number of complex DAGs, tasks, and dependencies. Overly complex DAG code — deep task graphs, heavy top-level Python, dynamic task generation — increases parsing time and metadata DB load, which manifests as UI latency. Simplifying and optimizing DAG code (reducing task count, avoiding top-level DB calls, using static DAGs) is the standard mitigation.

Answer

GKE node power failure; restart cluster.

Answer

Too many users logged in; force logout.

Answer

Browser cache; clear your local cache.

Question 18

Which TWO of the following are true regarding Dataplex security? (Choose two)

Accepted Answer

IAM roles can be applied at the Lake, Zone, or Asset level. Option A is correct because Dataplex enforces fine-grained access control through IAM policies that can be attached at the Lake, Zone, or Asset level, allowing administrators to grant roles such as roles/dataplex.viewer or roles/dataplex.editor scoped to specific resources. Option E is correct because Dataplex supports VPC Service Controls perimeter integration, letting organizations restrict access to Dataplex resources and mitigate data exfiltration by adding dataplex.googleapis.com to service perimeters. Option B is not correct because Dataplex relies on Google Cloud IAM and federated identity rather than per-asset passwords. Option C is not correct because Dataplex respects and layers on top of existing Cloud Storage IAM permissions rather than ignoring them. Option D is not correct because Dataplex uses Google Cloud's existing authentication (Cloud Identity/Google Workspace, OAuth 2.0) instead of providing a separate authentication system.

Answer

Users must provide a password for each asset

Answer

Dataplex ignores existing storage permissions

Answer

Dataplex provides its own authentication system

Question 19

You are developing a dashboard in Looker. You notice that the 'Explore' menu is missing a specific set of fields. What should you check in the LookML model?

Accepted Answer

Verify the 'fields' parameter in the Explore definition.. If the Explore menu is missing a specific set of fields in Looker, you should verify the 'fields' parameter in the Explore definition within the LookML model. This parameter explicitly lists which fields are included in the Explore, and if a field is omitted, it will not appear. Other options like BigQuery location, table indexing, or browser cache are unrelated to LookML field visibility.

Answer

Check the project's 'BigQuery' location.

Answer

Check if the table is indexed.

Answer

Clear the browser cache.

Question 20

When using Cloud Scheduler to trigger a Workflows execution, what is the best practice for authenticating the scheduler to the Workflows API?

Accepted Answer

Use an OIDC token with a service account.. The best practice for Cloud Scheduler to authenticate to the Workflows API is to configure the job with an OIDC token tied to a service account that has permission to invoke the workflow. Cloud Scheduler supports OIDC authentication for HTTP targets, and the service account's identity is verified by Workflows via IAM. This avoids static credentials and follows least-privilege.

Answer

Use an API Key generated in the console.

Answer

Hardcode credentials in the scheduler job.

Answer

Use 'Public' access for the workflow.