1Z0-184-25 Question 135
Single answerYou are designing a RAG (Red-Amber-Green) application in Oracle Cloud Infrastructure to track project tasks� urgency levels and store classification logic in PL/SQL. You want the solution to be maintainable, secure, and easily adjustable for changing thresholds. Which approach should you use to implement the classification logic in your Oracle Database?
- A
Place all classification logic in separate anonymous PL/SQL blocks that are invoked from the front-end application for each request.
- B
Implement the classification logic as stand-alone procedures, duplicating similar logic in each procedure to handle Red, Amber, and Green statuses.
- C
Create a packaged PL/SQL function with thresholds and classification routines centralized in a package, ensuring a single source of truth for future updates.
- D
Embed the classification logic directly in client-side code, passing only final classification results to the database.
Show answer and explanation
Correct answer: C
Explanation
A best practice when creating PL/SQL-based solutions�such as a RAG application�is to store related logic in a package. This approach allows for centralized definitions, easier version control, and secure access through package specifications. Refer to Oracle documentation on PL/SQL package best practices (e.g., Oracle Database PL/SQL Language Reference) for guidelines on structuring maintainable and secure application logic.
- A. Incorrect.
Option 1 is incorrect because relying on anonymous blocks for each request leads to scattered logic, making maintenance and version control difficult. Also, repeated blocks can introduce inconsistencies in classification thresholds.
- B. Incorrect.
Option 2 is incorrect because while stand-alone procedures can work, duplicating logic across multiple procedures increases the risk of inconsistent thresholds. Centralizing logic is typically a recommended best practice.
- C. Correct.
Option 3 is correct because packaging the logic in a PL/SQL package ensures a central location for thresholds and classification routines, simplifies maintenance, and provides a clean interface for other database objects and the application.
- D. Incorrect.
Option 4 is incorrect because embedding classification logic in client-side code makes it harder to maintain consistent thresholds across different clients and can introduce security risks by exposing sensitive logic outside the database.