Question

A company runs an internal reporting application on an Aurora MySQL provisioned cluster with two db.r6g.2xlarge instances (writer plus reader). The application is heavily used for about four hours each business day and is nearly idle the rest of the time, including nights and weekends. Finance flags the cluster as one of the highest recurring costs, yet the team must keep the database instantly available whenever a user logs in, and they cannot tolerate the multi-minute delays of stopping and starting the cluster on a schedule. Which change best reduces cost while preserving on-demand responsiveness?

Accepted Answer

Migrate the cluster to Aurora Serverless v2, configuring a low minimum ACU capacity and a higher maximum ACU so capacity scales up during the daily peak and scales down to the minimum during idle periods.

Answer

Keep the provisioned cluster but purchase a 3-year All Upfront Reserved Instance for both nodes to lower the hourly rate.

Answer

Create an EventBridge Scheduler rule to stop the Aurora cluster each evening and start it each morning, accepting the startup latency.

Answer

Replace the reader with an Aurora Auto Scaling policy that adds read replicas based on CPU, and downsize the writer to db.r6g.large.