Google Associate Data Practitioner 1-10

表示モード

画像位置

文字位置

理解度の自動記録

問題幅 % 60〜100%

Q1Google Associate Data Practitioner

Q1. You have prepared a curated dataset organizing market trends in BigQuery and want to share it with multiple external partners.
You want to control which rows and columns each partner can access.
You want to implement this in line with Google’s recommended practices.
How should you proceed?

A. Publish the dataset as a listing in Analytics Hub.
Use the subscription feature to grant each partner dataset-level access.
B. Provision a dedicated Cloud Storage bucket for each partner.
Export the dataset to each bucket and assign each partner to their respective bucket.
Set bucket-level access with IAM roles.
C. Grant each partner an IAM role that provides read access to the BigQuery dataset.
D. Create a dedicated project for each partner and copy the dataset into each project.
Publish the copied dataset in Analytics Hub and grant each partner dataset-level access via subscription.

Show answer

Correct answer: A. Publish the dataset as a listing in Analytics Hub. Use the subscription feature to grant each partner dataset-level access.

Analytics Hub is a service that lets the provider retain access control while sharing data with external partners.
Delivering data via subscription allows query execution only, without letting partners copy the data, and lets you finely manage the scope of sharing.
Option B: Cloud Storage is meant for file storage and is not suitable as a means of sharing a BigQuery dataset.
Option C: IAM roles alone cannot provide fine-grained row- and column-level control.
Option D: Creating and copying to a separate project per partner requires significant effort and lacks scalability.
Therefore, sharing through Analytics Hub is optimal.
Introduction to Analytics Hub | BigQuery

Q2Google Associate Data Practitioner

Q2. Using data you have on hand, you want to demonstrate BigQuery’s capabilities to your organization’s executives.
You need to load a file stored on your local machine into BigQuery just once, with as little effort as possible.
How should you proceed?

A. Write and run a Python script that uses the BigQuery Storage Write API library.
B. Create a Dataproc cluster, copy the file to Cloud Storage, and write an Apache Spark job that uses the spark-bigquery-connector.
C. Run the bq load command on your local machine.
D. Create a Dataflow job that uses Apache Beam’s FileIO and BigQueryIO connectors with the local runner.

Show answer

Correct answer: C. Run the bq load command on your local machine.

To perform a one-time load with minimal effort, a ready-to-use tool is the right choice.
The bq load command is a CLI tool included in the Google Cloud SDK that can upload a local CSV or JSON file directly to BigQuery with a single command.
It requires no additional coding and lets you leverage the existing SDK as-is.
Option A: A Python script using the Storage Write API requires authentication, library setup, and debugging, which is overkill.
Option B: Involving a Dataproc cluster and file transfer is far too complex.
Option D: Building a Dataflow job is also too much effort for a one-time load.
Using the bq command-line tool | BigQuery

Q3Google Associate Data Practitioner

Q3. Your team needs to analyze a large dataset in BigQuery to understand trends in user behavior.
The analysis involves advanced statistical computations, Python packages, and visualization.
You need to propose a managed, collaborative environment where the team can develop and share the analysis.
What should you recommend?

A. Create a Colab Enterprise notebook and connect it to BigQuery.
Share the notebook with the team and perform data analysis and visualization in Colab Enterprise.
B. Build a statistical model with BigQuery ML.
Share the query with the team, analyze the data, and create visualizations in Looker Studio.
C. Create a Looker Studio dashboard and connect it to BigQuery.
Share the dashboard with the team and perform data analysis and visualization in Looker Studio.
D. Connect Google Sheets to BigQuery with Connected Sheets.
Share the Google Sheet with the team and perform data analysis and visualization in Google Sheets.

Show answer

Correct answer: A. Create a Colab Enterprise notebook and connect it to BigQuery. Share the notebook with the team and perform data analysis and visualization in Colab Enterprise.

A Colab Enterprise notebook connected to BigQuery provides a managed, collaborative environment well suited to advanced statistical computations, Python packages, and visualization.
It supports Python libraries for advanced analysis and integrates seamlessly with BigQuery for querying large-scale data.
Collaborative development and sharing across the team are also easy, and it includes visualization capabilities.
Options B and C are visualization-centric tools and are not well suited to complex Python computations.
Option D: Connected Sheets is oriented toward spreadsheet analysis and is not suitable for advanced statistical processing or custom visualization.
Introduction to Colab Enterprise | Vertex AI

Q4Google Associate Data Practitioner

Q4. You are a database administrator who manages regional sales transaction data in a BigQuery table.
You need to ensure that each sales representative can view only the transactions for the region they are responsible for.
How should you proceed?

A. Grant appropriate IAM permissions on the dataset.
B. Create data masking rules.
C. Create row-level access policies.
D. Add policy tags in BigQuery.

Show answer

Correct answer: C. Create row-level access policies.

Row-level access policies filter rows based on conditions such as the region a user is responsible for, restricting which rows each user can view.
This achieves fine-grained access control without creating separate tables or views, and aligns with the principle of least privilege.
Option A: IAM permissions control access at the dataset level and do not support row-level control.
Option B: Data masking is a feature that hides column values and cannot be used to filter rows.
Option D: Policy tags are used for column-level access control.
Therefore, row-level access policies are optimal.
Introduction to row-level security | BigQuery

Q5Google Associate Data Practitioner

Q5. Your organization needs to introduce near-real-time analytics for events arriving at Pub/Sub at thousands per second.
The incoming messages require transformation.
You need to build a pipeline that processes and transforms the data and loads it into BigQuery while minimizing development time.
How should you proceed?

A. Use a Google-provided Dataflow template to process the Pub/Sub messages, apply the transformation, and write to BigQuery.
B. Create a Cloud Data Fusion instance and configure Pub/Sub as the source.
Process and transform the messages in Data Fusion and write to BigQuery.
C. Use a Cloud Storage subscription to ingest the data from Pub/Sub into Cloud Storage.
Create a Dataproc cluster and use PySpark to transform the data in Cloud Storage and write it to BigQuery.
D. Use a Cloud Run function to process the Pub/Sub messages, apply the transformation, and write to BigQuery.

Show answer

Correct answer: A. Use a Google-provided Dataflow template to process the Pub/Sub messages, apply the transformation, and write to BigQuery.

Google-provided Dataflow templates are pre-built and optimized for streaming data processing, so you can deploy the pipeline with minimal development.
They handle ingestion from Pub/Sub, transformation, and writing to BigQuery all at once, ensuring scalability and low latency.
Option B: Building a Cloud Data Fusion setup takes significant effort, which conflicts with the requirement to minimize development time.
Option C: Routing through a Dataproc cluster is complex and increases latency.
Option D: Cloud Run functions are not well suited to continuous stream processing.
Google-provided templates | Dataflow

Q6Google Associate Data Practitioner

Q6. You operate and manage a large volume of data on Cloud Storage, including raw data, processed data, and backups.
Your organization is subject to strict compliance regulations that mandate data immutability for certain data types.
You want to reduce storage costs in an efficient manner while meeting the retention requirements.
How should you proceed?

A. Configure lifecycle management rules that move objects to the appropriate storage class based on access patterns.
Enable versioning on all objects to meet the immutability requirement.
B. Move objects to different storage classes based on age and access patterns.
To meet the immutability requirement, encrypt specific objects using customer-managed encryption keys (CMEK) with Cloud KMS.
C. Create a Cloud Run function that periodically checks object metadata and moves objects to the appropriate storage class based on age and access patterns.
To enforce immutability, apply object holds to specific objects.
D. Apply object holds to specific objects to enforce immutability, and configure lifecycle management rules that move objects to the appropriate storage class based on age and access patterns.

Show answer

Correct answer: D. Apply object holds to specific objects to enforce immutability, and configure lifecycle management rules that move objects to the appropriate storage class based on age and access patterns.

Object holds prevent objects from being deleted or overwritten, ensuring immutability, while lifecycle management rules automatically move objects to lower-cost storage classes based on age.
This combination achieves both compliance and cost reduction.
Option A: Versioning does not guarantee immutability.
Option B: CMEK is encryption, not a means of ensuring immutability.
Option C: Manual management via a Cloud Run function creates unnecessary effort.
Object holds | Cloud Storage

Q7Google Associate Data Practitioner

Q7. You are building a data pipeline that handles highly sensitive customer data stored in a Cloud Storage bucket.
You need to keep this data accessible even if a single zone fails.
How should you proceed?

A. Configure Cloud CDN in front of the bucket.
B. Enable object versioning on the bucket.
C. Store the data in a multi-region bucket.
D. Store the data in Nearline storage.

Show answer

Correct answer: C. Store the data in a multi-region bucket.

A multi-region bucket replicates data across multiple locations, so access to the data is maintained even if a single zone fails.
It is well suited to sensitive data that requires high availability and durability.
Option A: Cloud CDN is intended to improve delivery performance and does not protect the availability of the source data.
Option B: Object versioning is a feature for guarding against overwrites and deletions and does not ensure availability during a zone failure.
Option D: Nearline is a storage class choice and is unrelated to redundancy against zone failures.
Bucket locations | Cloud Storage

Q8Google Associate Data Practitioner

Q8. Your organization sends IoT event data to a Pub/Sub topic.
A subscriber application reads the messages, transforms them, and stores them in a data warehouse.
You have found that during peak hours, when writes to the topic are especially heavy, the subscriber application fails to acknowledge (ack) messages within the deadline.
You need to rework the pipeline so that message processing can continue while handling these load spikes.
How should you proceed?

A. Keep retrying until the messages are acknowledged.
B. Implement flow control on the subscriber side.
C. Forward unacknowledged messages to a dead-letter topic.
D. Seek to the last acknowledged message.

Show answer

Correct answer: B. Implement flow control on the subscriber side.

Implementing flow control on the subscriber side lets you adjust the rate at which messages are fetched and processed, so that even during load spikes the subscriber does not become overloaded and can acknowledge within the deadline.
This keeps the pipeline stable even during peak hours and avoids unnecessarily dropping or delaying messages.
Option A: Retrying alone does not resolve the overload.
Option C: A dead-letter topic is a destination for processing failures and is not a countermeasure for spikes.
Option D: Seeking is a means of reprocessing and does not address surges.
Handle backlog with flow control | Pub/Sub

Q9Google Associate Data Practitioner

Q9. You operate a web application that stores data in a Cloud SQL database.
You need to offload read traffic from the primary database instance and improve the application’s read performance.
You want to implement this while minimizing effort and cost.
How should you proceed?

A. Use Cloud CDN to cache frequently accessed data.
B. Store frequently accessed data in a Memorystore instance.
C. Migrate the database to a larger Cloud SQL instance.
D. Enable automated backups and create a read replica of the Cloud SQL instance.

Show answer

Correct answer: D. Enable automated backups and create a read replica of the Cloud SQL instance.

Creating a read replica lets you offload read traffic from the primary instance, distributing the load and improving read performance.
It is easy to implement within Cloud SQL and is cost-efficient.
The primary can focus on writes while the replica handles read queries, improving performance with minimal effort.
Option A: Cloud CDN is intended for static web content and is not appropriate for databases.
Option B: Memorystore is an additional caching layer that increases implementation effort.
Option C: Scaling up is costly and does not distribute the read load.
About replication | Cloud SQL

Q10Google Associate Data Practitioner

Q10. Your organization has decided to migrate its existing enterprise data warehouse to BigQuery.
The data pipeline tool currently in use already supports connectors for BigQuery.
You need to determine the approach that optimizes migration speed.
How should you proceed?

A. Create a temporary file system that facilitates data transfer from the existing environment to Cloud Storage.
Use Storage Transfer Service to migrate the data to BigQuery.
B. Build the data pipeline using the Cloud Data Fusion web interface.
Create a directed acyclic graph (DAG) to facilitate orchestration of the pipeline.
C. Reconfigure the data mappings using the existing data pipeline tool’s BigQuery connector.
D. Rebuild the data pipeline using BigQuery Data Transfer Service and migrate the data to BigQuery.

Show answer

Correct answer: C. Reconfigure the data mappings using the existing data pipeline tool’s BigQuery connector.

Since the existing tool already supports a BigQuery connector, reconfiguring the data mappings with that connector is the most efficient approach.
Leveraging the current tool reduces migration complexity and setup time, optimizing migration speed.
You can feed data directly into BigQuery without going through additional services or intermediate steps.
Options A and D require new services or rebuilding the pipeline, which increases effort.
Option B: Rebuilding in Cloud Data Fusion introduces unnecessary complexity.
Introduction to data warehouse migration | BigQuery

Google Associate Data Practitioner 1-10

Choose confidence levels to display

Q1. You have prepared a curated dataset organizing market trends in BigQuery and want to share it with multiple external partners.You want to control which rows and columns each partner can access.You want to implement this in line with Google’s recommended practices.How should you proceed?

Q2. Using data you have on hand, you want to demonstrate BigQuery’s capabilities to your organization’s executives.You need to load a file stored on your local machine into BigQuery just once, with as little effort as possible.How should you proceed?

Q4. You are a database administrator who manages regional sales transaction data in a BigQuery table.You need to ensure that each sales representative can view only the transactions for the region they are responsible for.How should you proceed?

Q7. You are building a data pipeline that handles highly sensitive customer data stored in a Cloud Storage bucket.You need to keep this data accessible even if a single zone fails.How should you proceed?

Q9. You operate a web application that stores data in a Cloud SQL database.You need to offload read traffic from the primary database instance and improve the application’s read performance.You want to implement this while minimizing effort and cost.How should you proceed?

Q10. Your organization has decided to migrate its existing enterprise data warehouse to BigQuery.The data pipeline tool currently in use already supports connectors for BigQuery.You need to determine the approach that optimizes migration speed.How should you proceed?

Q1. You have prepared a curated dataset organizing market trends in BigQuery and want to share it with multiple external partners.
You want to control which rows and columns each partner can access.
You want to implement this in line with Google’s recommended practices.
How should you proceed?

Q2. Using data you have on hand, you want to demonstrate BigQuery’s capabilities to your organization’s executives.
You need to load a file stored on your local machine into BigQuery just once, with as little effort as possible.
How should you proceed?

Q4. You are a database administrator who manages regional sales transaction data in a BigQuery table.
You need to ensure that each sales representative can view only the transactions for the region they are responsible for.
How should you proceed?

Q7. You are building a data pipeline that handles highly sensitive customer data stored in a Cloud Storage bucket.
You need to keep this data accessible even if a single zone fails.
How should you proceed?

Q9. You operate a web application that stores data in a Cloud SQL database.
You need to offload read traffic from the primary database instance and improve the application’s read performance.
You want to implement this while minimizing effort and cost.
How should you proceed?

Q10. Your organization has decided to migrate its existing enterprise data warehouse to BigQuery.
The data pipeline tool currently in use already supports connectors for BigQuery.
You need to determine the approach that optimizes migration speed.
How should you proceed?