GCP Professional Data Engineer Practice Question

Your organization ingests terabytes of JSON telemetry every hour into several multi-regional Cloud Storage buckets that serve as the raw landing zone of a newly built data lake. Security teams require that:

All new objects be automatically discovered and registered in a central metadata catalog that exposes schema and lineage.
Sensitive attributes such as e-mail or credit-card numbers be detected and tagged so that only designated data stewards can view them.
Objects that have not been accessed for 180 days be moved to a lower-cost storage class without operator intervention. Which approach satisfies all requirements while minimizing operational overhead?

Expose each bucket through external tables in BigQuery, apply column-level security policies there, schedule Cloud Functions to scan objects with the Data Loss Prevention API, and build a weekly Dataflow job to rewrite aged objects into Coldline class.
Rely on Cloud Asset Inventory and Cloud Monitoring to catalog and secure bucket contents, trigger Pub/Sub notifications to a Cloud Function for PII detection and storage-class changes, and manage data access through custom Cloud IAM roles on each object.
Create a Dataplex lake with separate raw zones for each bucket, enable automatic discovery to publish metadata to Data Catalog, configure Dataplex data profiling with Cloud DLP to apply policy tags on detected PII, and add Object Lifecycle Management rules to move objects to Coldline after 180 days.
Deploy an Apache Atlas cluster on-premises to catalog Cloud Storage via connectors, manage access with per-bucket IAM policies, run the DLP API on demand for PII scans, and use cron-driven gsutil scripts to transition objects to Coldline.

Report Issue

Answer Description

Dataplex can group buckets into lakes and zones, continuously crawl Cloud Storage to infer schemas, capture lineage, and publish technical metadata to Data Catalog, providing a unified search experience. Within Dataplex you attach fine-grained IAM and Data Catalog policy tags that restrict access to sensitive fields. You can invoke Cloud DLP data profilers from Dataplex to detect and classify PII automatically, feeding those tags. Finally, Cloud Storage Object Lifecycle Management rules natively transition objects-for example, to Coldline-after 180 days of inactivity, giving hands-off cost control. The other options depend on custom scripts or external tooling (Apache Atlas, Cloud Functions, Cloud Asset Inventory) to provide capabilities that Dataplex, Data Catalog, and Cloud Storage already deliver as managed services, increasing operational burden and missing integrated governance features.

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.