Autonomous Database Fundamentals
Domain 1: Autonomous Database Fundamentals
This domain tests whether you can explain the Autonomous Database family and select the right deployment, workload, and scaling model. Questions often hinge on Serverless versus Dedicated, workload type, compute units, and which operations Oracle automates.
Core rules and concepts
- Serverless is fully managed by Oracle and billed by ECPU or OCPU; Dedicated runs on Exadata infrastructure you provision, with Autonomous Container Databases and Autonomous Databases layered on it.
- Workload types include Transaction Processing, Data Warehouse, JSON, and APEX. The choice sets defaults and tuning behavior; it is not just a label.
- Compute is independent of storage. 1 OCPU = 2 ECPUs in Autonomous Database. Auto scaling can raise compute above the base count to absorb peaks, but it is bounded, not infinite.
- Storage grows automatically; it is not tied to the compute shape. Backups, patching, statistics gathering, indexing, and tuning are automated.
- Autonomous features include automatic indexing, automatic statistics, automatic backups, and automatic maintenance. These reduce manual DBA work but do not remove the need to monitor.
Worked example
A team needs a small APEX application with occasional JSON documents, public internet access limited to corporate IPs, and the lowest operational overhead. Choose Serverless, an APEX workload type, a small ECPU base with auto scaling, and a public endpoint restricted by an access control list. Reasoning: Serverless avoids Exadata infrastructure management; APEX supplies the low-code stack; auto scaling handles occasional peaks; the ACL narrows exposure without requiring a private endpoint.
Traps that cost marks
- Treating OCPU and ECPU as interchangeable; a 2-ECPU shape is not 2 OCPUs.
- Assuming auto scaling removes the need to size a base or that it scales storage.
- Choosing Dedicated for a small workload where isolation does not justify cost.
- Confusing automated maintenance with the absence of monitoring.