Set up cascading read replicas for a Cloud SQL destination instance

Overview

Database Migration Service lets you migrate data to a Cloud SQL destination instance, and then set up cascading read replicas for that instance.

Cascading read replicas let you create a read replica under a Cloud SQL read replica in the same or different region before promoting your instance to primary. You can add up to four levels of cascading read replicas, including the primary instance.

Cascading read replicas in Cloud SQL for MySQL are supported only for global transaction identifier (GTID)-based replication . If you want to use cascading read replicas in your migration, set GTID_MODE to ON on your source instance.

Set up cascading read replicas

To migrate your data to a Cloud SQL destination that is using cascading read replicas:

1. Set up and start your migration job. For more information, see Create a migration job .

2. Wait until the full dump phase of the job completes and the migration job enters the change data capture (CDC) phase. Your migration job is in the CDC phase when the Statuscolumn in the Migration jobspage indicates a CDC substatus.

3. Once in the CDC phase, you can set up your cascading read replica. For information about how to create cascading read replicas for a Cloud SQL destination instance, see Set up a cascading replica .

4. You can promote your Cloud SQL replica to primary instance using Database Migration Service. For information about promoting a destination instance, see Promote a migration .

When you promote the Cloud SQL replica to a primary instance, the replica's cascading replica hierarchy is retained.