Oracle Cloud Infrastructure 2021 Multicloud Architect Associate — Question 46

You have created an Autonomous Data Warehouse (ADW) service in your company's Oracle Cloud Infrastructure (OCI) tenancy and you now have to load historical data into it. You have already extracted this historical data from multiple data marts and data warehouses. This data is stored in multiple CSV text files and these files are ranging in size from 25 MB to 20 GB.
Which is the most efficient and error tolerant method for loading data into ADW? (Choose the best answer.)

Answer options

• A. Create Auth token, use it to create an object storage credential by executing DBMS_CLOUD.CREATE_CREDENTIAL, using the web console upload the CSV files to an OCI object storage bucket, create the tables in the ADW database and then execute DBMS_CLOUD.COPY_DATA for each CSV file to copy the contents into the corresponding ADW database table.
• B. Create the tables in the ADW database and then execute SQL*Loader for each CSV file to load the contents into the corresponding ADW database table.
• C. Create Auth token, use it to create an object storage credential by executing DBMS_CLOUD.CREATE_CREDENTIAL, using OCI CLI upload the CSV files to an OCI object storage bucket, create the tables in the ADW database and then execute Data Pump Import for each CSV file to copy the contents into the corresponding ADW database table.
• D. Create Auth token, use it to create an object storage credential by executing DBMS_CLOUD.CREATE_CREDENTIAL, using OCI CLI upload the CSV files to an OCI object storage bucket, create the tables in the ADW database and then execute DBMS_CLOUD.COPY_DATA for each CSV file to copy the contents into the corresponding ADW database table.

Correct answer: D

Explanation

Option D is the most efficient and error-tolerant method as it leverages the DBMS_CLOUD.COPY_DATA procedure, which is designed for bulk data transfer and handles errors effectively. Option A uses the web console for uploads, which is less efficient for large datasets. Option B does not utilize object storage and relies on SQL*Loader, which is not as optimized for ADW. Option C, while it uses OCI CLI, suggests using Data Pump Import, which isn't as suitable for CSV file handling in this context.