Question

A retail company runs a production Amazon Aurora MySQL-compatible cluster supporting an e-commerce platform that processes thousands of transactions per minute. As part of a modernization effort, the database team must perform a major version upgrade of the Aurora engine. The business requires that the upgrade be tested against production-like data and traffic patterns before cutover, that the switchover to the new version take no more than a few minutes, and that a fast rollback path exists if issues are detected after cutover. Which approach best meets these requirements with the least operational effort?

Accepted Answer

Create an Amazon RDS Blue/Green Deployment for the Aurora cluster, upgrade the engine on the green environment, validate it while replication keeps it in sync with blue, and switch over during a maintenance window.

Answer

Take a manual snapshot of the cluster, restore it to a new cluster running the target engine version, redirect application traffic to the new cluster, and delete the old cluster after validation.

Answer

Use AWS DMS to replicate data from the current cluster to a newly provisioned cluster running the target engine version, then cut over the application endpoint once DMS reports full-load and CDC completion.

Answer

Modify the cluster in place to the target engine version during a scheduled maintenance window and rely on the automated pre-upgrade snapshot for rollback if the upgrade fails.