AWS-CDK ECS Service Step AutoScaling Error

[deromka]

AWS-CDK Version - 0.22.0

Trying to setup the ECS Service Step AutoScaling as defined here

When deploying getting the following error:

1/44 | 2:29:14 PM | CREATE_FAILED        | AWS::ApplicationAutoScaling::ScalingPolicy  | ECS/MyService/TaskCount/Target/MyServiceStepCpuScaling/UpperPolicy (ECSMyServiceTaskCountTargetMyServiceStepCpuScalingUpperPolicyD9DB0A83) Either ScalingTargetId or all three of [ResourceId, ScalableDimension, ServiceNamespace] must be specified
       new StepScalingAction (/Users/user/devel/project/node_modules/@aws-cdk/aws-applicationautoscaling/lib/step-scaling-action.js:18:26)
       \_ new StepScalingPolicy (/Users/user/devel/project/node_modules/@aws-cdk/aws-applicationautoscaling/lib/step-scaling-policy.js:52:32)
       \_ ScalableTarget.scaleOnMetric (/Users/user/devel/project/node_modules/@aws-cdk/aws-applicationautoscaling/lib/scalable-target.js:66:16)
       \_ ScalableTaskCount.doScaleOnMetric (/Users/user/devel/project/node_modules/@aws-cdk/aws-applicationautoscaling/lib/base-scalable-attribute.js:42:21)
       \_ ScalableTaskCount.scaleOnMetric (/Users/user/devel/project/node_modules/@aws-cdk/aws-ecs/lib/base/scalable-task-count.js:18:22)
       \_ EcsServicesAndTasksConstruct.setupServiceAutoScaling (/Users/user/devel/project/dist/lib/constructs/ecs-services-and-tasks-construct.js:83:24)
       \_ /Users/user/devel/project/dist/lib/constructs/ecs-services-and-tasks-construct.js:56:26
       \_ Array.map (<anonymous>)
       \_ new EcsServicesAndTasksConstruct (/Users/user/devel/project/dist/lib/constructs/ecs-services-and-tasks-construct.js:47:39)
       \_ SmsApplicationStack.deployDRMaaSOnECS (/Users/user/devel/project/dist/lib/stacks/app-stack.js:203:23)
       \_ new SmsApplicationStack (/Users/user/devel/project/dist/lib/stacks/app-stack.js:47:40)
       \_ Object.<anonymous> (/Users/user/devel/project/dist/lib/app/sms-app.js:14:27)
       \_ Module._compile (internal/modules/cjs/loader.js:689:30)
       \_ Object.Module._extensions..js (internal/modules/cjs/loader.js:700:10)
       \_ Module.load (internal/modules/cjs/loader.js:599:32)
       \_ tryModuleLoad (internal/modules/cjs/loader.js:538:12)
       \_ Function.Module._load (internal/modules/cjs/loader.js:530:3)
       \_ Function.Module.runMain (internal/modules/cjs/loader.js:742:12)
       \_ startup (internal/bootstrap/node.js:283:19)
       \_ bootstrapNodeJSCore (internal/bootstrap/node.js:743:3)

My code snippet:

setupServiceAutoScaling(name : string, service : Ec2Service, props : MyProps)
    {
        const serviceScaling = service.autoScaleTaskCount({
            minCapacity: props.ecsMinContainerCount,
            maxCapacity: props.ecsMaxContainerCount
        })

        serviceScaling.scaleOnMetric(name + 'StepCpuScaling', {
            metric: service.metricCpuUtilization(),
            scalingSteps:[
                { upper: 30, change: -1 },
                { lower: 60, change: +1 }
            ],
            adjustmentType: autoscaling.AdjustmentType.ChangeInCapacity,
        })
}

Generated CF template fragment:

ECSMyServiceTaskCountTargetMyServiceStepCpuScalingLowerPolicy6D568037:
    Type: AWS::ApplicationAutoScaling::ScalingPolicy
    Properties:
      PolicyName: drmaasromanappdev550026ECSMyServiceTaskCountTargetMyServiceStepCpuScalingLowerPolicyB5BEEEF9
      PolicyType: StepScaling
      StepScalingPolicyConfiguration:
        AdjustmentType: ChangeInCapacity
        MetricAggregationType: Average
        StepAdjustments:
          - MetricIntervalUpperBound: 0
            ScalingAdjustment: -1
    Metadata:
      aws:cdk:path: APP/ECS/MyService/TaskCount/Target/MyServiceStepCpuScaling/LowerPolicy/Resource
  ECSMyServiceTaskCountTargetMyServiceStepCpuScalingLowerAlarmD5D408D8:
    Type: AWS::CloudWatch::Alarm
    Properties:
      ComparisonOperator: LessThanOrEqualToThreshold
      EvaluationPeriods: 1
      Threshold: 30
      AlarmActions:
        - Ref: ECSMyServiceTaskCountTargetMyServiceStepCpuScalingLowerPolicy6D568037
      AlarmDescription: Lower threshold scaling alarm
      Dimensions:
        - Name: ClusterName
          Value:
            Fn::ImportValue: StackName:ecsEcsClusterClusterName3A20C485
        - Name: ServiceName
          Value:
            Fn::GetAtt:
              - ECSMyServiceService9B5D7DD3
              - Name
      MetricName: CPUUtilization
      Namespace: AWS/ECS
      Period: 60
      Statistic: Average
    Metadata:
      aws:cdk:path: APP/ECS/MyService/TaskCount/Target/MyServiceStepCpuScaling/LowerAlarm/Resource
  ECSMyServiceTaskCountTargetMyServiceStepCpuScalingUpperPolicyD9DB0A83:
    Type: AWS::ApplicationAutoScaling::ScalingPolicy
    Properties:
      PolicyName: drmaasromanappdev550026ECSMyServiceTaskCountTargetMyServiceStepCpuScalingUpperPolicy13DD87BD
      PolicyType: StepScaling
      StepScalingPolicyConfiguration:
        AdjustmentType: ChangeInCapacity
        MetricAggregationType: Average
        StepAdjustments:
          - MetricIntervalLowerBound: 0
            ScalingAdjustment: 1
    Metadata:
      aws:cdk:path: APP/ECS/MyService/TaskCount/Target/MyServiceStepCpuScaling/UpperPolicy/Resource
  ECSMyServiceTaskCountTargetMyServiceStepCpuScalingUpperAlarmC120A764:
    Type: AWS::CloudWatch::Alarm
    Properties:
      ComparisonOperator: GreaterThanOrEqualToThreshold
      EvaluationPeriods: 1
      Threshold: 60
      AlarmActions:
        - Ref: ECSMyServiceTaskCountTargetMyServiceStepCpuScalingUpperPolicyD9DB0A83
      AlarmDescription: Upper threshold scaling alarm
      Dimensions:
        - Name: ClusterName
          Value:
            Fn::ImportValue: StackName:ecsEcsClusterClusterName3A20C485
        - Name: ServiceName
          Value:
            Fn::GetAtt:
              - ECSMyServiceService9B5D7DD3
              - Name
      MetricName: CPUUtilization
      Namespace: AWS/ECS
      Period: 60
      Statistic: Average
    Metadata:
      aws:cdk:path: APP/ECS/MyService/TaskCount/Target/MyServiceStepCpuScaling/UpperAlarm/Resource

It seems that the ScalingPolicy missing some essential parameters:
Either ScalingTargetId or all three of [ResourceId, ScalableDimension, ServiceNamespace] must be specified

ECSMyServiceTaskCountTargetMyServiceStepCpuScalingUpperPolicyD9DB0A83:
    Type: AWS::ApplicationAutoScaling::ScalingPolicy
    Properties:
      PolicyName: drmaasromanappdev550026ECSMyServiceTaskCountTargetMyServiceStepCpuScalingUpperPolicy13DD87BD
      PolicyType: StepScaling
      StepScalingPolicyConfiguration:
        AdjustmentType: ChangeInCapacity
        MetricAggregationType: Average
        StepAdjustments:
          - MetricIntervalLowerBound: 0
            ScalingAdjustment: 1
    Metadata:
      aws:cdk:path: APP/ECS/MyService/TaskCount/Target/MyServiceStepCpuScaling/UpperPolicy/Resource

[RomainMuller]

Are you able to try to upgrade your project to the latest CDK (0.24.1 as of writing) and see if you still have the same problem?

[piradeepk]

This is still an issue, using the latest version of the CDK (0.30.0)

[piradeepk]

Pending PR review: 2522

[piradeepk]

Fix has been merged and issue has been resolved.