Explain how AWS DMS works.
AWS DMS operates by setting up a temporary replication server that reads data from the source and writes it to the destination.
1. Replication Instance
When you start a migration, AWS provisions a managed EC2 instance in the background called the Replication Instance. This instance performs the heavy lifting—connecting to your source, reading the data, formatting it, and loading it into the target.
2. Endpoints (Source and Target)
DMS needs to know exactly where to connect. You define two endpoints:
- Source Endpoint: The database you are moving away from (can be on-premises, on an EC2 instance, or in another cloud).
- Target Endpoint: The destination database inside AWS (e.g., Amazon RDS, Aurora, Amazon S3, or DynamoDB).
3. Replication Tasks
A DMS Task defines exactly how the data moves. You can choose from three types of migration styles:
- Full Load: Migrates all existing data from the source to the target at that specific point in time.
- Change Data Capture (CDC): Captures ongoing changes (Inserts, Updates, Deletes) in real-time by reading the source database's transaction logs.
- Full Load + CDC: The most common method. It copies the entire bulk database first, and then seamlessly transitions into capturing live changes. This keeps both databases in sync until you are ready to switch your application to the new AWS database.
