Question

A media company is migrating a 400 GB PostgreSQL database from an on-premises data center to Amazon Aurora PostgreSQL. The migration involves an initial full load followed by an ongoing change data capture (CDC) phase to keep the target in sync until cutover. The source database has highly variable write activity: near-zero changes overnight and heavy bursts during business hours. The team wants to minimize operational overhead and pay only for the replication capacity actually consumed, without manually sizing or scaling replication instances. Which approach best meets these requirements?

Accepted Answer

Use AWS DMS Serverless, which automatically provisions and scales replication capacity (in DCUs) based on workload and charges only for capacity used.

Answer

Provision a large AWS DMS replication instance sized for peak load and leave it running for the entire migration window to avoid CDC lag.

Answer

Use AWS DataSync to perform the full load and CDC replication, scheduling incremental tasks hourly to capture ongoing changes.

Answer

Deploy a fleet of small DMS replication instances behind an Auto Scaling group so capacity scales in and out with the source write volume.