Aurora Serverless Scaling execution block
Category: Database scaling
During a Region switch, your destination Aurora Serverless cluster may have ACU (Aurora Capacity Unit) settings that are far below what's needed to absorb production traffic. The Aurora Serverless Scaling execution block automatically calculates and applies the correct min and max ACU capacity to your destination cluster based on the source cluster's actual usage, ensuring your serverless database can handle the incoming workload without throttling or connection failures.
Key benefits
Usage-based capacity calculation: Rather than relying on static configuration, Region switch derives target capacity from the source cluster's actual peak utilization over the last 24 hours, giving you right-sized capacity based on real traffic patterns.
Cross-engine-type intelligence: Whether your source is Serverless, Provisioned, or a hybrid configuration, Region switch knows how to translate the source capacity into appropriate ACU settings for the destination Serverless cluster.
Percentage-based scaling for active-active: Configure a target percentage above 100% (for example, 200%) for active-active architectures where the destination must absorb combined traffic from both Regions.
When to use
Active-passive with Serverless standby: Your destination Region runs a Serverless cluster at minimal ACUs and needs to scale up before receiving production traffic.
Active-active failover: Both Regions serve traffic, and during a switch the remaining Region must handle the combined load – use a target percentage above 100%.
Mixed-engine Global Databases: Your source Region uses provisioned instances but your destination uses Serverless – Region switch handles the capacity translation automatically.
How Aurora Serverless Scaling compares to alternatives
Without this execution block, customers must manually calculate ACU requirements and modify cluster settings before switching traffic – a complex, error-prone process especially when source and destination use different engine types.
| Approach | Pros | Cons | |
|---|---|---|---|
| 1 | Aurora Serverless Scaling block | Automated calculation from real usage, handles cross-engine translation, percentage-based control, integrated with plan orchestration | Only scales up; modifies ACU settings which may drift from IaC |
| 2 | Manual ACU adjustment | Full control | Requires calculating ACU equivalents under pressure; slow; error-prone |
| 3 | Scripted automation | Customizable | Must replicate cross-engine translation logic; no plan evaluation; maintenance burden |
| 4 | Pre-provisioning (max ACU always high) | No failover delay | Expensive; defeats the cost benefit of Serverless; wasteful in standby Region |
The Aurora Serverless Scaling block is the right choice when you need automated, usage-aware capacity scaling that handles the complexity of cross-engine ACU translation.
How it works
After you configure an Aurora Serverless Scaling execution block, Region switch confirms that there is one source cluster and one destination cluster in the specified global database. The target capacity is determined based on the source cluster type:
Source is Serverless:
Min ACU = source cluster's peak observed ACU utilization (the
ServerlessDatabaseCapacityCloudWatch metric) over the last 24 hoursMax ACU = peak of the source cluster's max ACU over the last 24 hours
Source is Provisioned:
Maps the source cluster's EC2 instance memory to equivalent ACUs (instance memory in GiB ÷ 2)
Sets max ACU to 256
Source is Hybrid (Provisioned + Serverless):
Min ACU = maximum of the provisioned instance ACU equivalent and the observed serverless ACU utilization over 24 hours
Max ACU = 256
Region switch then applies the target percentage to calculate final values:
destination min ACU = round_to_nearest_0.5(targetPercent × source min ACU) destination max ACU = round_to_nearest_0.5(targetPercent × source max ACU)
If the destination cluster's current capacity is already at or above the calculated target, Region switch completes the step without making changes. Region switch does not scale down cluster capacity. When the destination cluster is not Serverless, the block completes successfully as a no-op.
For active-active plans, Region switch uses the other configured Region as the source. If a Region is being deactivated, Region switch uses the other active Region as the source to calculate the percentage to scale.
Note
Executing this block modifies the min and max ACU capacity settings of your Aurora Serverless clusters, which may cause configuration drift if you manage these values through infrastructure-as-code tools or other automation. Ensure your configuration management processes account for these changes to prevent unintended rollbacks.
Configuration
When you configure the Aurora Serverless Scaling execution block, you enter the global cluster identifier for your Aurora Global Database and the database cluster ARNs for each Region that you want to be scaled up during plan execution.
Important
Before you configure the execution block, make sure that the plan's execution role has the correct IAM policy in place. For more information, see Aurora serverless scaling execution block sample policy.
To configure an Aurora Serverless Scaling execution block, enter the following values:
Step name: Enter a name.
Step description (optional): Enter a description of the step.
Aurora Global Database cluster name: Enter the global cluster identifier.
Cluster ARN for Region: Enter the database cluster ARN to use in each Region for your plan.
Target percent (optional): Enter the percentage of derived source capacity to scale the destination cluster to. Defaults to 100. For active-active plans, consider a higher value (for example, 200%) to account for combined traffic.
Timeout: Enter a timeout value.
Then, choose Save step.
What is evaluated as part of plan evaluation
When Region switch evaluates your plan, Region switch performs several critical checks on your Aurora Serverless Scaling execution block configuration and permissions. Region switch evaluation verifies that Aurora Serverless clusters are present in both Regions, ensures that they are properly configured and accessible, and notes the current capacity in each Region. It also confirms that the maximum capacity in the target Region's cluster is sufficient to handle the specified percentage match of scale for the required capacity.
Region switch also validates that the plan's IAM role has the correct permissions for Aurora Serverless scaling. For more information about the required permissions for Region switch execution blocks, see Aurora serverless scaling execution block sample policy. If any of the checks fail, Region switch returns warning messages, which you can view in the console. Or, you can receive the validation warnings through or by using API operations.
Related resources
Managing Aurora Serverless v2 capacity in the Amazon Aurora User Guide
