Scaling OCPUs in Autonomous Database is designed to be seamless. The two true statements are:

Scaling can happen while there are active transactions in the database (B):ADB supports online scaling, meaning you can increase or decrease OCPUs (e.g., from 2 to 4) via the OCI console or CLI (e.g., oci db autonomous-database update --cpu-core-count 4) without stopping the database. Active transactions (e.g., INSERT INTO orders VALUES (...)) continue running during this process. Oracle’s architecture ensures the database remains available, adjusting resources in the background. For example, a web app processing orders won’t notice the scaling operation starting at 10:00 AM.

Active transactions continue running unaffected (C):During scaling, existing transactions are not interrupted, terminated, or paused. They complete normally, with Oracle managing resource allocation transparently (e.g., shifting CPU usage without killing sessions). For instance, a long-running UPDATE statement started before scaling finishes successfully, leveraging the database’s high-availability design. The status shows “SCALING IN PROGRESS,” but users experience no downtime.

The incorrect options are:

Active transactions are terminated and rolled back (A):False. Scaling is non-disruptive; transactions aren’t killed or rolled back, preserving data integrity and user experience. Termination only occurs during explicit stops or failures, not scaling.

Active transactions are paused (D):False. There’s no pausing mechanism during scaling; transactions run continuously, as pausing would disrupt OLTP or analytical workloads, countering ADB’s autonomous promise.

This online scaling capability is a key benefit, ensuring uninterrupted service.