Google Professional Data Engineer

The correct answer is B. Create a Cloud Monitoring dashboard based on the BigQuery metric slots/allocated_for_project Why B is correct:

BigQuery provides a native metric called slots/allocated_for_project (and slots/allocated_for_project_and_job_type) that directly reports slot allocations per project. This metric is available in Cloud Monitoring and can be filtered by project. Each analytics team can be granted access to a project-specific dashboard using this metric to monitor their own slot usage in real time — no custom logging or metrics required. It is efficient, accurate, and purpose-built for this use case.

Why the others are incorrect: A. query/scanned_bytes → Measures data scanned, not slot usage. Slot consumption depends on query complexity and concurrency, not just bytes scanned. Incorrect. C. Log export per project + custom metric from totalSlotMs → totalSlotMs is per job, not real-time slot allocation. → Requires complex log parsing, custom metrics, and aggregation delays. → Overly complicated and not real-time. Not recommended. D. Aggregated log export at org level → Same issues as C, but now centralized — harder to isolate per-project usage. → Still relies on delayed logs and custom metrics. Not suitable.