Cross-Region Aurora Replicas - AWS Prescriptive Guidance

Cross-Region Aurora Replicas

With Amazon Aurora MySQL-Compatible Edition clusters, you can use cross-Region Aurora Replicas to create a replica of your primary DB cluster in another AWS Region by using the database engine native replication mechanism. Aurora MySQL-Compatible uses binary log (binlog) replication. This kind of replication is based on the change data capture (CDC) process. CDC identifies and tracks changes to data in a database. Changes to the database on the primary are recorded in a change log file, which is transferred to the machine that runs the Aurora Replica in the second Region. A process on the secondary machine reads these change logs and generates SQL statements that are then applied to the secondary database. 

The following diagram shows the high-level architecture of a cross-Region Aurora Replica.

Description follows the diagram.

The diagram shows the following transaction flow:

  1. Transactions are committed on the primary Aurora clusters.

  2. A process records these changes on the source by using the binlog dump thread, which writes them to a binary log file.

  3. A replication agent sends the change logs through an I/O thread to relay logs in the secondary Region.

  4. A process reads the change logs, generates SQL statements, and applies the statements to the secondary Aurora cluster.

Aurora MySQL-Compatible clusters can have up to five read replicas created this way, each in a different Region. Aurora PostgreSQL-Compatible doesn't support cross-Region Aurora Replicas. However, for Aurora PostgreSQL DB clusters, you can use an Aurora global database.