Migrate your scaling plan - AWS Auto Scaling
Step 1: Review your existing setupStep 2: Create predictive scaling policiesStep 3: Review the forecasts that the predictive scaling policies generateStep 4: Prepare to delete the scaling planStep 5: Delete the scaling planStep 6: Reactivate dynamic scalingStep 7: Reactivate predictive scalingAmazon EC2 Auto Scaling reference for migrating target tracking scaling policiesApplication Auto Scaling reference for migrating target tracking scaling policiesAdditional information

Migrate your scaling plan

You can migrate from a scaling plan to Amazon EC2 Auto Scaling and Application Auto Scaling scaling policies.

Important

To migrate a scaling plan, you must complete multiple steps in exact order. While you migrate your scaling plan, don't update it, as that breaks the order of operations and could cause undesirable behavior.

Step 1: Review your existing setup

To determine which scaling settings you must move over, use the describe-scaling-plans command. 

aws autoscaling-plans describe-scaling-plans \
  --scaling-plan-names my-scaling-plan

Make a note of items that you want to preserve from the existing scaling plan, which can include the following:

  • MinCapacity – The minimum capacity of the scalable resource.

  • MaxCapacity – The maximum capacity of the scalable resource.

  • PredefinedLoadMetricType – A load metric for predictive scaling.

  • PredefinedScalingMetricType – A scaling metric for target tracking (dynamic) scaling and predictive scaling.

  • TargetValue – The target value for the scaling metric.

Differences between scaling plans and scaling policies

There are a few important differences between scaling plans and scaling policies:

  • A scaling policy can enable only one type of scaling: either target tracking scaling or predictive scaling. To use both scaling methods, you must create separate policies.

  • Likewise, you must define the scaling metric for predictive scaling and the scaling metric for target tracking scaling separately within their respective policies.

Step 2: Create predictive scaling policies

If you don't use predictive scaling, then skip ahead to Step 4: Prepare to delete the scaling plan.

To provide time to evaluate the forecast, we recommend that you create predictive scaling policies before other scaling policies.

For any Auto Scaling groups with an existing load metric specification, do the following to turn it into an Amazon EC2 Auto Scaling-based predictive scaling policy.

To create predictive scaling policies

  1. In a JSON file, define a MetricSpecifications structure as shown in the following example:

    {
  "MetricSpecifications":[
    {
      ... 
    }
  ]
}

  2. In the MetricSpecifications structure, for each load metric in your scaling plan, create a PredefinedLoadMetricSpecification or CustomizedLoadMetricSpecification using the equivalent settings from the scaling plan.

    The following are examples of the structure of the load metric section.

    With predefined metrics
    {
  "MetricSpecifications":[
    {
      "PredefinedLoadMetricSpecification":{
        "PredefinedMetricType":"ASGTotalCPUUtilization"
      },
      ...
    }
  ]
}

    For more information, see PredictiveScalingPredefinedLoadMetric in the Amazon EC2 Auto Scaling API Reference.

    With custom metrics
    {
  "MetricSpecifications":[
    {
      "CustomizedLoadMetricSpecification":{
        "MetricDataQueries":[
          {
            "Id":"load_metric",
            "MetricStat":{
              "Metric":{
                "MetricName":"MyLoadMetric",
                "Namespace":"MyNameSpace",
                "Dimensions":[
                  {
                    "Name":"MyOptionalMetricDimensionName",
                    "Value":"MyOptionalMetricDimensionValue"
                  }
                ]
              },
              "Stat":"Sum"
            }
          }
        ]
      },
      ...
    }
  ]
}

    For more information, see PredictiveScalingCustomizedLoadMetric in the Amazon EC2 Auto Scaling API Reference.

  3. Add the scaling metric specification to the MetricSpecifications and define a target value.

    The following are examples of the structure of the scaling metric and target value sections.

    With predefined metrics
    {
  "MetricSpecifications":[
    {
      "PredefinedLoadMetricSpecification":{
        "PredefinedMetricType":"ASGTotalCPUUtilization"
      },
      "PredefinedScalingMetricSpecification":{
        "PredefinedMetricType":"ASGCPUUtilization"
      },
      "TargetValue":50
    }
  ],
  ...
}

    For more information, see PredictiveScalingPredefinedScalingMetric in the Amazon EC2 Auto Scaling API Reference.

    With custom metrics
    {
  "MetricSpecifications":[
    {
      "CustomizedLoadMetricSpecification":{
        "MetricDataQueries":[
          {
            "Id":"load_metric",
            "MetricStat":{
              "Metric":{
                "MetricName":"MyLoadMetric",
                "Namespace":"MyNameSpace",
                "Dimensions":[
                  {
                    "Name":"MyOptionalMetricDimensionName",
                    "Value":"MyOptionalMetricDimensionValue"
                  }
                ]
              },
              "Stat":"Sum"
            }
          }
        ]
      },
      "CustomizedScalingMetricSpecification":{
        "MetricDataQueries":[
          {
            "Id":"scaling_metric",
            "MetricStat":{
              "Metric":{
                "MetricName":"MyUtilizationMetric",
                "Namespace":"MyNameSpace",
                "Dimensions":[
                  {
                    "Name":"MyOptionalMetricDimensionName",
                    "Value":"MyOptionalMetricDimensionValue"
                  }
                ]
              },
              "Stat":"Average"
            }
          }
        ]
      },
      "TargetValue":50
    }
  ],
  ...
}

    For more information, see PredictiveScalingCustomizedScalingMetric in the Amazon EC2 Auto Scaling API Reference.

  4. To forecast only, add the property Mode with a value of ForecastOnly. After you finish migrating predictive scaling and making sure that the forecast is accurate and reliable, you can change the mode to allow scaling. For more information, see Step 7: Reactivate predictive scaling. 

    {
  "MetricSpecifications":[
  ...
  ],
  "Mode":"ForecastOnly",
  ...
}

    For more information, see PredictiveScalingConfiguration in the Amazon EC2 Auto Scaling API Reference.

  5. If the ScheduledActionBufferTime property is present in your scaling plan, then copy its value to the SchedulingBufferTime property in your predictive scaling policy.

    {
  "MetricSpecifications":[
  ...
  ],
  "Mode":"ForecastOnly",
  "SchedulingBufferTime":300,
  ...
}

    For more information, see PredictiveScalingConfiguration in the Amazon EC2 Auto Scaling API Reference.

  6. If the PredictiveScalingMaxCapacityBehavior and PredictiveScalingMaxCapacityBuffer properties are present in your scaling plan, then you can configure the MaxCapacityBreachBehavior and MaxCapacityBuffer properties in your predictive scaling policy. These properties define what should happen if the forecast capacity approaches or exceeds the maximum capacity specified for the Auto Scaling group.

    Warning

    If you set the MaxCapacityBreachBehavior property to IncreaseMaxCapacity, then more instances could launch than intended unless you monitor and manage the increased maximum capacity. The increased maximum capacity becomes the new normal maximum capacity for the Auto Scaling group until you manually update it. The maximum capacity doesn't automatically decrease back to the original maximum.

    {
  "MetricSpecifications":[
  ...
  ],
  "Mode":"ForecastOnly",
  "SchedulingBufferTime":300,
  "MaxCapacityBreachBehavior": "IncreaseMaxCapacity",
  "MaxCapacityBuffer": 10
}

    For more information, see PredictiveScalingConfiguration in the Amazon EC2 Auto Scaling API Reference.

  7. Save the JSON file with a unique name. Make a note of the file name. You need it in the next step and again at the end of the migration procedure when you reactivate your predictive scaling policies. For more information, see Step 7: Reactivate predictive scaling.

  8. After you save your JSON file, run the put-scaling-policy command. In the following example, replace each user input placeholder with your own information.

    aws autoscaling put-scaling-policy --policy-name my-predictive-scaling-policy \
  --auto-scaling-group-name my-asg --policy-type PredictiveScaling \
  --predictive-scaling-configuration file://my-predictive-scaling-config.json

    If successful, this command returns the policy's Amazon Resource Name (ARN).

    {
  "PolicyARN": "arn:aws:autoscaling:region:account-id:scalingPolicy:2f4f5048-d8a8-4d14-b13a-d1905620f345:autoScalingGroupName/my-asg:policyName/my-predictive-scaling-policy",
  "Alarms": []
}

  9. Repeat these steps for each load metric specification that you're migrating to an Amazon EC2 Auto Scaling-based predictive scaling policy.

Step 3: Review the forecasts that the predictive scaling policies generate

If you don't use predictive scaling, then skip the following procedure.

A forecast is available shortly after you create a predictive scaling policy. After Amazon EC2 Auto Scaling generates the forecast, you can review the forecast for the policy through the Amazon EC2 Auto Scaling console and adjust as necessary.

To review the forecast for a predictive scaling policy

  1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

  2. In the navigation pane, choose Auto Scaling Groups, and then choose the name of your Auto Scaling group from the list.

  3. On the Automatic scaling tab, in Predictive scaling policies, choose your policy.

  4. In the Monitoring section, you can view your policy's past and future forecasts for load and capacity against actual values.

    For more information, see Review predictive scaling monitoring graphs in the Amazon EC2 Auto Scaling User Guide.

  5. Repeat these steps for each predictive scaling policy that you created.

Step 4: Prepare to delete the scaling plan

For any resources with an existing target tracking scaling configuration, do the following to collect any additional information that you need from the scaling plan before deleting it.

To describe the scaling policy information from the scaling plan, use the describe-scaling-plan-resources command. In the following example command, replace my-scaling-plan with your own information.

aws autoscaling-plans describe-scaling-plan-resources \
  --scaling-plan-name my-scaling-plan \
  --scaling-plan-version 1

Review the output and confirm you want to migrate the scaling policies described. Use this information to create new Amazon EC2 Auto Scaling and Application Auto Scaling-based target tracking scaling polices in Step 6: Reactivate dynamic scaling.

Step 5: Delete the scaling plan

Before creating new target tracking scaling policies, you must delete the scaling plan to delete the scaling policies that it created.

To delete your scaling plan, use the delete-scaling-plan command. In the following example command, replace my-scaling-plan with your own information.

aws autoscaling-plans delete-scaling-plan \
  --scaling-plan-name my-scaling-plan \
  --scaling-plan-version 1

After you delete the scaling plan, dynamic scaling is deactivated. So if there are sudden surges in traffic or workload, the capacity available for each scalable resource won't increase on its own. As a precaution, you might want to manually increase the capacity of your scalable resources in the short term.

To increase the capacity of an Auto Scaling group

  1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

  2. In the navigation pane, choose Auto Scaling Groups, and then choose the name of your Auto Scaling group from the list.

  3. On the Details tab, choose Group details, Edit.

  4. For Desired capacity, increase the desired capacity.

  5. When you're finished, choose Update.

To add an Aurora replica to a DB cluster

  1. Open the Amazon RDS console at https://console.aws.amazon.com/rds/.

  2. In the navigation pane, choose Databases, and then select your DB cluster.

  3. Make sure that both the cluster and the primary instance are in the Available state.

  4. Choose Actions, Add reader.

  5. On the Add reader page, specify options for your new Aurora replica.

  6. Choose Add reader.

To increase the provisioned read and write capacity of a DynamoDB table or global secondary index

  1. Open the DynamoDB console at https://console.aws.amazon.com/dynamodb/.

  2. In the navigation pane, choose Tables, and then choose the name of your table from the list.

  3. On the Additional settings tab, choose Read/write capacity, Edit.

  4. On the Edit read/write capacity page, for Read capacity, Provisioned capacity units, increase the provisioned read capacity of the table.

  5. (Optional) If you want your global secondary indexes to use the same read capacity settings as the base table, then select the Use the same read capacity settings for all global secondary indexes check box.

  6. For Write capacity, Provisioned capacity units, increase the provisioned write capacity of the table.

  7. (Optional) If you want your global secondary indexes to use the same write capacity settings as the base table, then select the Use the same write capacity settings for all global secondary indexes check box.

  8. If you didn't select the check boxes in steps 5 or 7, then scroll down the page to update the read and write capacity of any global secondary indexes.

  9. Choose Save changes to continue.

To increase the running task count for your Amazon ECS service

  1. Open the console at https://console.aws.amazon.com/ecs/v2.

  2. In the navigation pane, choose Clusters, and then choose the name of your cluster from the list.

  3. In the Services section, select the check box next to the service, and then choose Update.

  4. For Desired tasks, enter the number of tasks that you want to run for the service.

  5. Choose Update.

To increase the capacity of a Spot Fleet

  1. Open the Amazon EC2 console at https://console.aws.amazon.com/ec2/.

  2. In the navigation pane, choose Spot Requests, and then select your Spot Fleet request.

  3. Choose Actions, Modify target capacity.

  4. In Modify target capacity, enter the new target capacity and On-Demand Instance portion.

  5. Choose Submit.

Step 6: Reactivate dynamic scaling

Reactivate dynamic scaling by creating target tracking scaling policies.

When you create a target tracking scaling policy for an Auto Scaling group, you add it directly to the group. When you create a target tracking scaling policy for other scalable resources, you first register the resource as a scalable target and then you add a target tracking scaling policy to the scalable target.

Create target tracking scaling policies for Auto Scaling groups

To create target tracking scaling policies for Auto Scaling groups

  1. In a JSON file, create a PredefinedMetricSpecification or CustomizedMetricSpecification using the equivalent settings from the scaling plan.

    The following are examples of a target tracking configuration. In these examples, replace each user input placeholder with your own information.

    With predefined metrics
    {
  "TargetValue": 50.0,
  "PredefinedMetricSpecification": 
    {
      "PredefinedMetricType": "ASGAverageCPUUtilization"
    }
}

    For more information, see PredefinedMetricSpecification in the Amazon EC2 Auto Scaling API Reference.

    With custom metrics
    {
  "TargetValue": 100.0,
  "CustomizedMetricSpecification": {
    "MetricName": "MyBacklogPerInstance",
    "Namespace": "MyNamespace",
    "Dimensions": [{
      "Name": "MyOptionalMetricDimensionName",
      "Value": "MyOptionalMetricDimensionValue"
    }],
    "Statistic": "Average",
    "Unit": "None"
  }
}

    For more information, see CustomizedMetricSpecification in the Amazon EC2 Auto Scaling API Reference.

  2. To create your scaling policy, use the put-scaling-policy command, along with the JSON file that you created in the previous step. In the following example, replace each user input placeholder with your own information.

    aws autoscaling put-scaling-policy --policy-name my-target-tracking-scaling-policy \
  --auto-scaling-group-name my-asg --policy-type TargetTrackingScaling \
  --target-tracking-configuration file://config.json

  3. Repeat this process for each scaling plan-based scaling policy that you're migrating to an Amazon EC2 Auto Scaling-based target tracking scaling policy.

Create target tracking scaling policies for other scalable resources

Next, create target tracking scaling policies for other scalable resources by performing the following configuration tasks.

  • Register a scalable target for auto scaling with the Application Auto Scaling service.

  • Add a target tracking scaling policy on the scalable target.

To create target tracking scaling policies for other scalable resources

  1. Use the register-scalable-target command to register the resource as a scalable target and define the scaling limits for the scaling policy.

    In the following example, replace each user input placeholder with your own information. For the command options, provide the following information:

    • --service-namespace – A namespace for the target service (for example, ecs). To obtain service namespaces, see the RegisterScalableTarget reference.

    • --scalable-dimension – A scalable dimension associated with the target resource (for example, ecs:service:DesiredCount). To obtain scalable dimensions, see the RegisterScalableTarget reference.

    • --resource-id – A resource ID for the target resource (for example, service/my-cluster/my-service). For information about the syntax and examples of specific resource IDs, see the RegisterScalableTarget reference.

    aws application-autoscaling register-scalable-target --service-namespace namespace \
  --scalable-dimension dimension \
  --resource-id identifier \
  --min-capacity 1 --max-capacity 10

    If successful, this command returns the ARN of the scalable target.

    {
    "ScalableTargetARN": "arn:aws:application-autoscaling:region:account-id:scalable-target/1234abcd56ab78cd901ef1234567890ab123"
}

  2. In a JSON file, create a PredefinedMetricSpecification or CustomizedMetricSpecification using the equivalent settings from the scaling plan.

    The following are examples of a target tracking configuration.

    With predefined metrics
    {
  "TargetValue": 70.0,
  "PredefinedMetricSpecification": 
    {
      "PredefinedMetricType": "ECSServiceAverageCPUUtilization"
    }
}

    For more information, see PredefinedMetricSpecification in the Application Auto Scaling API Reference.

    With custom metrics
    {
  "TargetValue": 70.0,
  "CustomizedMetricSpecification": {
    "MetricName": "MyUtilizationMetric",
    "Namespace": "MyNamespace",
    "Dimensions": [{
      "Name": "MyOptionalMetricDimensionName",
      "Value": "MyOptionalMetricDimensionValue"
    }],
    "Statistic": "Average",
    "Unit": "Percent"
  }
}

    For more information, see CustomizedMetricSpecification in the Application Auto Scaling API Reference.

  3. To create your scaling policy, use the put-scaling-policy command, along with the JSON file that you created in the previous step.

    aws application-autoscaling put-scaling-policy --service-namespace namespace \
  --scalable-dimension dimension \
  --resource-id identifier \
  --policy-name my-target-tracking-scaling-policy --policy-type TargetTrackingScaling \
  --target-tracking-scaling-policy-configuration file://config.json

  4. Repeat this process for each scaling plan-based scaling policy that you're migrating to an Application Auto Scaling-based target tracking scaling policy.

Step 7: Reactivate predictive scaling

If you don't use predictive scaling, then skip this step.

Reactivate predictive scaling by switching predictive scaling to forecast and scale.

To make this change, update the JSON files that you created in Step 2: Create predictive scaling policies and change the value of the Mode option to ForecastAndScale as in the following example:

"Mode":"ForecastAndScale"

Then, update each predictive scaling policy with the put-scaling-policy command. In this example, replace each user input placeholder with your own information.

aws autoscaling put-scaling-policy --policy-name my-predictive-scaling-policy \
  --auto-scaling-group-name my-asg --policy-type PredictiveScaling \
  --predictive-scaling-configuration file://my-predictive-scaling-config.json

Alternatively, you can make this change from the Amazon EC2 Auto Scaling console by turning on the Scale based on forecast setting. For more information, see Predictive scaling for Amazon EC2 Auto Scaling in the Amazon EC2 Auto Scaling User Guide.

Amazon EC2 Auto Scaling reference for migrating target tracking scaling policies

For reference purposes, the following table lists all the target tracking configuration properties in the scaling plan with their corresponding property in the Amazon EC2 Auto Scaling PutScalingPolicy API operation.

Scaling plan source property Amazon EC2 Auto Scaling target property
PolicyName PolicyName
PolicyType PolicyType
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.Dimensions.Name TargetTrackingConfiguration.CustomizedMetricSpecification.Dimensions.Name
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.Dimensions.Value TargetTrackingConfiguration.CustomizedMetricSpecification.Dimensions.Value
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.MetricName TargetTrackingConfiguration.CustomizedMetricSpecification.MetricName
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.Namespace TargetTrackingConfiguration.CustomizedMetricSpecification.Namespace
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.Statistic TargetTrackingConfiguration.CustomizedMetricSpecification.Statistic
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.Unit TargetTrackingConfiguration.CustomizedMetricSpecification.Unit
TargetTrackingConfiguration.DisableScaleIn TargetTrackingConfiguration.DisableScaleIn
TargetTrackingConfiguration.EstimatedInstanceWarmup TargetTrackingConfiguration.EstimatedInstanceWarmup¹
TargetTrackingConfiguration.PredefinedScalingMetricSpecification.PredefinedScalingMetricType TargetTrackingConfiguration.PredefinedMetricSpecification.PredefinedMetricType
TargetTrackingConfiguration.PredefinedScalingMetricSpecification.ResourceLabel TargetTrackingConfiguration.PredefinedMetricSpecification.ResourceLabel
TargetTrackingConfiguration.ScaleInCooldown Not available
TargetTrackingConfiguration.ScaleOutCooldown Not available
TargetTrackingConfiguration.TargetValue TargetTrackingConfiguration.TargetValue

¹ Instance warmup is a feature for Auto Scaling groups that helps to ensure that newly launched instances are ready to receive traffic before contributing their usage data to the scaling metric. While instances are still warming up, Amazon EC2 Auto Scaling slows down the process of adding or removing instances to the group. Instead of specifying a warmup time for a scaling policy, we recommend that you use the default instance warmup setting of your Auto Scaling group to ensure that all instance launches use the same instance warmup time. For more information, see Set the default instance warmup for an Auto Scaling group in the Amazon EC2 Auto Scaling User Guide.

Application Auto Scaling reference for migrating target tracking scaling policies

For reference purposes, the following table lists all the target tracking configuration properties in the scaling plan with their corresponding property in the Application Auto Scaling PutScalingPolicy API operation.

Scaling plan source property Application Auto Scaling target property
PolicyName PolicyName
PolicyType PolicyType
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.Dimensions.Name TargetTrackingScalingPolicyConfiguration.CustomizedMetricSpecification.Dimensions.Name
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.Dimensions.Value TargetTrackingScalingPolicyConfiguration.CustomizedMetricSpecification.Dimensions.Value
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.MetricName TargetTrackingScalingPolicyConfiguration.CustomizedMetricSpecification.MetricName
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.Namespace TargetTrackingScalingPolicyConfiguration.CustomizedMetricSpecification.Namespace
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.Statistic TargetTrackingScalingPolicyConfiguration.CustomizedMetricSpecification.Statistic
TargetTrackingConfiguration.CustomizedScalingMetricSpecification.Unit TargetTrackingScalingPolicyConfiguration.CustomizedMetricSpecification.Unit
TargetTrackingConfiguration.DisableScaleIn TargetTrackingScalingPolicyConfiguration.DisableScaleIn
TargetTrackingConfiguration.EstimatedInstanceWarmup Not available
TargetTrackingConfiguration.PredefinedScalingMetricSpecification.PredefinedScalingMetricType TargetTrackingScalingPolicyConfiguration.PredefinedMetricSpecification.PredefinedMetricType
TargetTrackingConfiguration.PredefinedScalingMetricSpecification.ResourceLabel TargetTrackingScalingPolicyConfiguration.PredefinedMetricSpecification.ResourceLabel
TargetTrackingConfiguration.ScaleInCooldown¹ TargetTrackingScalingPolicyConfiguration.ScaleInCooldown
TargetTrackingConfiguration.ScaleOutCooldown¹ TargetTrackingScalingPolicyConfiguration.ScaleOutCooldown
TargetTrackingConfiguration.TargetValue TargetTrackingScalingPolicyConfiguration.TargetValue

¹ Application Auto Scaling uses cooldown periods to slow down scaling when your scalable resource is scaling out (increasing capacity) and scaling in (reducing capacity). For more information, see Define cooldown periods in the Application Auto Scaling User Guide.

Additional information

To learn how to create new predictive scaling policies from the console, see the following topic:

To learn how to create new target tracking scaling policies using the console, see the following topics: