Lab 3 - Failover to read Replica

• Migration Methodologies
• Preparation
• Lab 1 - Change Database Instance Type
• Lab 2 - Promote Read Replica
• Lab 3 - Failover to Read Replica
• Lab 4 - Restore from snapshot
• Cleanup

Overview

In this lab you will take the Amazon Aurora database created in the Preparation step, add a read replica, and then failover to the replica.

This method works for Amazon Aurora only (Postgres and MySQL).

This is a fast way to migrate, and applications can continue to use the existing database endpoints.

Add a read replica and failover

1. In the RDS Databases console, click on the Aurora cluster name, aurora-lab3 to go to the detail screen for the database
2. From the Actions menu select Add reader
3. Enter a DB Instance Identifier (e.g. aurora-graviton)
4. Select a graviton instance type e.g. db.t4g.medium from burstable classes
5. Leave all other settings as default, scroll to the bottom of the page and click Add reader
6. Once the graviton instance is in the Available state (this may take a few minutes), there should be an x86-64 writer and a Graviton reader

Failover to the Graviton Instance

1. Select the radio button next to the newly created Graviton instance named aurora-graviton, choose actions and then Failover

2. Note that the graviton instance now becomes the writer and the x86-64 instance is demoted to a reader node
3. Notice how quickly the failover happens compared to changing the instance type.
4. The lab is now complete and you can move on to the next one.