Google Professional Data Engineer Question 200
Select 3Google Cloud PlatformYour organization uses BigQuery for analytics and has a dataset containing sensitive customer information. You need to share aggregated insights from this dataset with a third-party partner, ensuring that sensitive information is not exposed. Which steps should you take to define rules for securely sharing the data?
- A
Use BigQuery authorized views to share only the aggregated data.
- B
Grant the third-party partner the 'roles/bigquery.dataEditor' role on the dataset.
- C
Apply column-level security to restrict access to sensitive columns.
- D
Use row-level security to filter data based on the third-party partner's access requirements.
- E
Export the dataset as a CSV file, manually redact sensitive data, and share the file via Cloud Storage.
Show answer and explanation
Correct answers: A, C, D
Explanation
To securely share data in BigQuery, you can use authorized views to expose only aggregated or filtered results, apply column-level security to restrict access to sensitive fields, and use row-level security to filter data based on access policies. These methods ensure compliance and prevent unauthorized access while sharing insights. Granting direct dataset-level roles or manually redacting data is not recommended, as it risks exposing sensitive information and lacks scalability.
- A. Correct.
Correct: Authorized views allow you to share only specific query results (e.g., aggregated data) without exposing the underlying raw data. This is a secure and scalable way to share data in BigQuery.
- B. Incorrect.
Incorrect: Granting the 'roles/bigquery.dataEditor' role gives the partner full access to the dataset, including the ability to update or delete data. This is not secure for sharing sensitive data.
- C. Correct.
Correct: Column-level security enables you to restrict access to specific sensitive columns in the dataset, ensuring the partner only sees permitted data.
- D. Correct.
Correct: Row-level security allows you to apply access policies to filter rows based on specific conditions, which helps in defining granular data-sharing rules.
- E. Incorrect.
Incorrect: Manually redacting and sharing data as a CSV file is error-prone, not scalable, and lacks the security and compliance features provided by BigQuery's native tools.