You have two projects running BigQuery jobs: a production project that typically uses >=300 slots and spikes an extra 500, and an ad-hoc project that uses <=200 slots and should be billed on data scanned. How do you allocate compute and billing?

The correct option effectively addresses both projects' requirements. The production project, needing a guaranteed baseline and spike capacity, benefits from an Enterprise Edition reservation with a 300-slot baseline and autoscaling up to 500 additional slots (total 800 slots, matching its peak need of 300+500). The ad-hoc project, with lower usage and a preference for data-scanned billing, is best served by on-demand pricing, which aligns with its "billed on data scanned" requirement. Incorrect options: A single Enterprise reservation for both projects doesn't allow for separate billing models or dedicated slot allocation for the production project's critical workload. Creating an Enterprise reservation with a 0-slot baseline and ignoring idle slots for the ad-hoc project doesn't align with the "billed on data scanned" requirement, as reservations are slot-based. An 800-slot baseline for the SLA project is inefficient if its typical usage is 300, and autoscaling for the ad-hoc project still implies slot-based billing, not data-scanned.