Explanation
Why the waste happens and who it affects.
However, when used for databases with constant load, the service's elasticity offers little advantage and adds cost overhead. Serverless instances run continuously in steady workloads, resulting in persistent ACU billing at a higher effective rate than a provisioned cluster of similar size. In addition, Serverless configurations cannot use Reserved Instances, which apply only to provisioned Aurora. Database Savings Plans can discount Serverless ACU usage, but provisioned instances can use Reserved Instances or Database Savings Plans, so compare both options with their commitment discounts applied.
Billing model
The pricing dimensions that drive this cost.
Aurora Provisioned clusters are billed by instance type and size, with options for Reserved Instance or Savings Plan discounts. Aurora Serverless clusters are billed per Aurora Capacity Unit (ACU) per second, with pricing optimized for variable workloads; Serverless usage is eligible for Database Savings Plans but not Reserved Instances. When workloads are steady, Serverless' dynamic pricing leads to consistently higher costs compared to equivalent provisioned instances.
How to detect
4 checks to find it in your estate.
- Review Aurora Serverless clusters where capacity usage remains constant over time
- Assess whether average and peak ACU consumption show minimal fluctuation across representative time windows
- Confirm whether workload patterns reflect consistent utilization rather than bursty or intermittent usage
- Evaluate whether provisioned Aurora instances could deliver similar performance at lower cost, especially with Reserved Instance coverage
How to fix
4 ways to remove the waste.
- Convert steady-state Aurora Serverless clusters to provisioned Aurora instances with appropriate instance sizing, after comparing Serverless with Database Savings Plans against provisioned instances with Reserved Instances
- Apply Reserved Instances or Savings Plans to reduce ongoing compute costs for predictable workloads
- Retain Aurora Serverless only for workloads with irregular, unpredictable, or low-utilization patterns
- Validate that performance requirements are met after migration; provisioned instances often provide lower latency and jitter
Documentation
Vendor references for pricing and configuration.