Amazon Aurora Serverless

Aurora Serveless provides a relatively simple, cost-effective option for infrequent, intermittent, or unpredictable workloads.

• An on-demand, auto-scaling configuration for MySQL and Postgres compatible edition

• Aurora Serverless DB cluster automatically stands up, shout down, and scales capacity up or down based on your application's needs.

Aurora Serverless is:

• On Demand and Auto Scaling service

• Automation of monitoring workloads and adjusting capacity for databases.

• Based on Demand: capacity adjusted based on application demands.

• Billing: Charged only for resources consumed by DB clusters; per-second billing.

• Budged Friendly: helps customers to stay well within budgets with the Auto Scaling and per-second billing features.

Aurora Serveless Concepts:

1. Aurora Capacity Units (ACUs): Measuraments on how your clusters scale

2. Set a minimum and maximum of ACUs for scaling requirements - can be zero.

3. Allocated (quickly) by AWS-managed warm pools

4. combination of about 2GiB of memory, matching CPU, and networking capability

5. Same data resiliancy as Aurora provisioned: six copies of data across three AZs

6. Multi-AZ deployments for establishing highly available clusters

Use Cases

• Variable Workloads

• Multi-Tenants Apps: Let the service mange database capacity for each individual app.

• New Apps: unsure what database istance needs are required.

• Dev and Test new features.

• Mixed-Use Apps. Application that serve more then one purpose with different traffic spikes.

• Capacity Planning. Easily swap from provisioned to serverless or vice versa.