AWS Database Migration Service: Preparing for Full Load

Answer: D.

Option A is incorrect because data migration can be performed from the primary or secondary node of the MongoDB replica set.

Using the secondary node of the source replica set for the read operations is recommended as it can result in reducing the overall load on the primary instance.

Option B is incorrect because CDC is required in online data migration.

AWS DMS uses CDC to replicate changes to target DocumentDB.Option C is incorrect because AWS DMS uses the source MongoDB oplog to perform data migration.

During online data migration, the oplog on each replica set should be large enough to contain all changes made during the entire duration of the data migration process.

Option D is CORRECT because AWS DMS does not migrate indexes.

For large data migrations, it is most efficient to pre-create indexes in the target Amazon DocumentDB cluster before migrating the data.

Reference:

https://aws.amazon.com/blogs/database/migrating-to-amazon-documentdb-with-the-online-method/

When migrating data from an on-premises MongoDB cluster to Amazon DocumentDB using AWS Database Migration Service (DMS), there are some best practices that a database specialist should follow to minimize the source database downtime.

Before beginning a full load of migration data, the specialist should turn off CDC (Change Data Capture) on the source database. CDC captures all the changes that occur on the source database during the migration process, which can cause issues such as replication lag, network congestion, and other performance problems.

In addition, the specialist should set the oplog to 0 on the source database. The oplog is the MongoDB operation log that stores all the changes that occur on the source database. By setting the oplog to 0, the specialist can ensure that DMS can perform a full load of the data without being interrupted by new changes that occur on the source database during the migration process.

Creating indexes on the target database is also a best practice, as it can help to optimize the performance of the target database. Indexes can speed up queries and improve the overall performance of the database, which can be especially important when dealing with large volumes of data.

Lastly, it is essential to ensure that DMS is using the primary node of the source cluster. The primary node is the node that is responsible for all write operations on the cluster. By using the primary node, DMS can ensure that all the data is migrated correctly and that there are no data inconsistencies or errors during the migration process.

In summary, before beginning a full load of migration data from an on-premises MongoDB cluster to Amazon DocumentDB using AWS DMS, a database specialist should turn off CDC on the source database, set the oplog to 0 on the source database, create indexes on the target database, and ensure that DMS is using the primary node of the source cluster.