Several teams share Athena. Operations needs separate query settings, query metrics, and a mechanism to stop individual queries that scan beyond an approved data limit. Which feature should it configure?

1. Send an AWS Budgets notification and treat notification delivery as an immediate hard stop on each oversized query.
2. Create separate Glue databases for each team and assume they automatically impose per-query scanned-byte limits.
3. A Redshift workload management queue for queries that execute only in Athena.
4. Athena workgroups with the required enforced settings, metrics, and per-query data-usage controls.