GCP Professional Data Engineer Practice Question

Your company wants to adopt a data-mesh model on Google Cloud. Retail, Logistics, and Finance teams will own their own raw Cloud Storage buckets and transformed BigQuery tables. A central data office must still enforce uniform data-classification, masking, and access-audit policies across all domains with minimal custom scripting and without taking over schema or pipeline ownership. Which architecture best satisfies these requirements?

Create a shared VPC; let each domain run separate BigQuery projects and maintain Cloud Functions that replicate organization-wide IAM roles and DLP templates to every project.
Schedule Cloud Composer workflows that nightly copy each domain's data into a central warehouse where the governance team applies masking views before republishing to consumers.
Use Dataplex to create one lake with separate zones and assets per domain, delegate asset-level ownership to the domain projects, and apply centralized taxonomy-based policy tags that propagate to BigQuery tables.
Put all data for every domain into a single BigQuery dataset secured with row-level security filters and managed entirely by the central data office.

Report Issue

Answer Description

Dataplex lets you build a lake that can include Cloud Storage and BigQuery assets. Inside the lake you create individual zones (and their assets) that can be mapped to business domains; you then grant fine-grained IAM roles so each domain team controls the data structure and processing in its zone. At the same time, Dataplex integrates with Data Catalog policy tags and Cloud Logging to apply centralized governance, data-classification, masking, and auditing rules across every asset in every zone automatically. The other options either rely on extensive custom scripting, concentrate all data in a single dataset (eliminating domain autonomy), or require copying data into a central warehouse, which increases cost and operational burden without solving policy enforcement consistently.

Ask Bash

Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.