Master AWS DevOps Pro 2026 – Conquer the Cloud with Confidence!

The recommended strategy for upgrading an RDS cluster with minimal downtime is to create a new read replica, update its engine version, and then promote it. This approach minimizes the impact on the existing production workload as it allows for testing and validation of the new instance before making it the primary source of data.

By creating a read replica, you can run the updated engine version in a separate instance while the original cluster continues to operate normally. This means that existing applications can continue to access the current version of the database without any interruptions. Once the new read replica is fully synchronized and you are confident in its performance and compatibility, you can promote it to be the primary instance. This promotion is generally a quick operation and can be performed with minimal disruption to users.

This strategy contrasts with the other options. Performing upgrades during off-peak hours may reduce the impact on users but does not eliminate downtime risks, nor does it provide an opportunity for testing the upgrade before it is applied live. Upgrading the existing cluster directly can lead to significant downtime, especially if issues arise during the upgrade process. Creating a backup and restoring it to a new instance is also a more disruptive approach, as it involves data migration and could lead to substantial downtime, making it less desirable for environments