Aurora Serverless

• Aurora Serverless automatically provisions, scales, and pauses database capacity based on actual workload.

• Designed for workloads that are infrequent, intermittent, or highly unpredictable.

• Eliminates the need for manual capacity planning.

• Billing is per second of active database usage, making it potentially more cost-effective for variable workloads.

How It Works

• The client connects to a Proxy Fleet managed by Aurora.

• The proxy layer routes requests to Aurora instances that are dynamically started or stopped as needed.

• Aurora automatically scales in and out the compute layer based on metrics such as connections and CPU usage.

• All compute instances share the same distributed storage volume, which is always available and decoupled from compute.

Benefits

• Seamless scaling – no downtime when scaling up or down.

• Cost efficiency – pay only for actual compute usage.

• High availability – storage remains online even when all compute is paused.

• No maintenance – Aurora manages patching, backups, and failover automatically.