Examkin

Question

Difficulty: MediumDatabase Migration and Schema Conversion using DMS and SCT

A company is migrating a high-throughput, on-premises Oracle database to an Amazon Aurora PostgreSQL-Compatible Edition DB cluster. The migration strategy requires minimal downtime, and a solutions architect has configured AWS DMS with a Change Data Capture (CDC) replication task. During testing, the source Oracle database experiences extremely high CPU utilization, and the DMS task cannot keep up with the volume of changes. Which of the following solutions should the solutions architect implement to reduce CPU utilization on the source database and improve CDC performance?

  1. Configure AWS DMS to use the Binary Reader method instead of the default Oracle LogMiner to read the redo and archived log files directly.Answer
  2. B
    Enable supplemental logging on all source Oracle tables and configure the AWS DMS task to use Oracle LogMiner with the useLogminerReader task setting set to true.
  3. C
    Modify the AWS DMS replication task settings to disable transaction-based replication and switch to statement-level logging on the source Oracle database.
  4. D
    Perform a replatforming migration using AWS SCT to continuously extract transaction logs directly to Amazon S3, and load them into Aurora PostgreSQL using pg_restore.

Answer

Configure AWS DMS to use the Binary Reader method instead of the default Oracle LogMiner to read the redo and archived log files directly.
The correct solution is to configure the AWS DMS task to use the Binary Reader method. When migrating from a high-throughput Oracle database, the default LogMiner method consumes significant CPU resources because it operates within the database engine to reconstruct changes. The Binary Reader method reads the redo and archived logs directly from the database file system or Automatic Storage Management (ASM), which minimizes the resource footprint on the source Oracle database and enhances replication performance.

Step-by-Step Solution

1
Analyze the impact of the default Oracle LogMiner mechanism on the source database.
Confirm that Oracle LogMiner reconstructs database changes using SQL queries, which causes high CPU usage under heavy transactional workloads.
To determine the root cause of the performance bottleneck on the source Oracle database.
2
Select a highly performant alternative CDC reading method suitable for high-volume Oracle databases.
Identify that AWS DMS Binary Reader reads the redo and archived logs directly from the OS file system or ASM disks, avoiding the SQL engine query overhead.
To choose a replication architecture that offloads processing from the source database instance.
3
Implement configuration changes on the source Oracle database and AWS DMS endpoint settings.
Grant AWS DMS the necessary permissions to read the Oracle directories containing the redo and archived logs, and enable the Binary Reader settings in the DMS source endpoint configuration.
To apply the Binary Reader method configuration to successfully complete CDC with minimal CPU utilization on the source database.

Key Concept

AWS DMS replication methods (LogMiner vs. Binary Reader) for Oracle CDC
Rate this question

Topics

Database Migration and Schema Conversion using DMS and SCT

More questions from this topic