‹ All SAP-C02 questions

SAP-C02 Practice Question 7: Heterogeneous Oracle to Aurora Migration

AWS Certified Solutions Architect – Professional (SAP-C02) · Test 1 · question 7 of 10

Migration and Modernizationeasy

An insurance company must retire an 8 TB on-premises Oracle database that backs a claims application and move it to Amazon Aurora PostgreSQL-Compatible Edition to escape Oracle licensing costs. The schema contains hundreds of tables, secondary indexes, views, materialized views, and several thousand lines of PL/SQL in packages and triggers.

The company has a 10 Gbps AWS Direct Connect connection to AWS. The claims application must stay online during the bulk data transfer, the final cutover outage must be under one hour, and the team needs to be able to compare source and target rows before cutting over.

Which migration approach meets these requirements?

Select one answer
Show answer and explanation

Answer:

  • A. Convert the schema and PL/SQL with the AWS Schema Conversion Tool, apply the converted objects to Aurora PostgreSQL, then run an AWS DMS task in full load plus change data capture mode with data validation enabled, and cut over when replication lag reaches zero

A heterogeneous migration, a different source and target engine, is a two-part job. The AWS Schema Conversion Tool (or the equivalent DMS Schema Conversion feature) converts the schema and the procedural code: tables, indexes, views, packages, and triggers become PostgreSQL equivalents, and the assessment report lists the objects that need manual work. AWS DMS then moves the data. A task configured for full load and ongoing replication copies the existing 8 TB while the source stays online, then applies changes captured from the Oracle redo logs until the target is caught up, so the cutover is only as long as it takes to stop writers and repoint the application. DMS data validation compares source and target rows and reports mismatches, satisfying the verification requirement.

Option B skips schema conversion. AWS DMS creates only the target tables and their primary keys if they do not exist; it does not create secondary indexes, views, materialized views, sequences, default values, or any PL/SQL. The claims application would arrive on a target missing most of its schema and all of its stored logic.

Option C is a homogeneous-migration technique used on the wrong pair of engines. Oracle Data Pump produces an Oracle-proprietary dump file that PostgreSQL cannot read, and even where a dump-and-load is possible it is a point-in-time export with no change capture, so the outage would last as long as the export, transfer, and load of 8 TB.

Option D solves a bandwidth problem the company does not have. With a 10 Gbps Direct Connect connection, 8 TB transfers online in hours; shipping a Snowball Edge adds days of transit and still provides no ongoing replication or schema conversion, so the cutover outage would be far more than one hour.

Reference: What is AWS Database Migration Service?

Practice Test 1 ›

Study guides for SAP-C02