Describe the feature
The current scaling policy defined in QueueProcessingServiceBase sets the targetUtilizationPercent at 50%, and is not configurable.
|
protected configureAutoscalingForService(service: BaseService) { |
This value should be configurable by users.
Use Case
We have an application that works on a high throughput queue, but the processing is not latency-sensitive and we can afford to run hotter on the CPU, allowing us to run fewer ECS tasks and save on cost.
Proposed Solution
This value should be configurable through an optional parameter added to QueueProcessingServiceBaseProps (and , by extension, QueueProcessingFargateServiceProps). If not specified by the user, this will default to 50% utilization, for backward compatibility.
Other Information
No response
Acknowledgements
CDK version used
1.141
Environment details (OS name and version, etc.)
NodeJS-14
Describe the feature
The current scaling policy defined in
QueueProcessingServiceBasesets thetargetUtilizationPercentat 50%, and is not configurable.aws-cdk/packages/@aws-cdk/aws-ecs-patterns/lib/base/queue-processing-service-base.ts
Line 344 in fd5808f
This value should be configurable by users.
Use Case
We have an application that works on a high throughput queue, but the processing is not latency-sensitive and we can afford to run hotter on the CPU, allowing us to run fewer ECS tasks and save on cost.
Proposed Solution
This value should be configurable through an optional parameter added to
QueueProcessingServiceBaseProps(and , by extension,QueueProcessingFargateServiceProps). If not specified by the user, this will default to 50% utilization, for backward compatibility.Other Information
No response
Acknowledgements
CDK version used
1.141
Environment details (OS name and version, etc.)
NodeJS-14