The correct answer is the Control Plane. Databricks divides its services between a control plane and a data plane. The control plane, managed within Databricks' own cloud account, hosts backend services including jobs, pipelines, and queries. It's important to note that serverless features such as SQL Endpoint and DLT compute utilize shared compute resources within the Control Plane. The control plane also securely stores notebook commands and workspace configurations, encrypted at rest. Conversely, the data plane, managed in the customer's cloud account, is where data resides and is processed, with the capability to connect to external data sources via Azure Databricks connectors for data ingestion or storage.