Create a step scaling policy using the Amazon CLI
You can create a step scaling policy for Application Auto Scaling by using the Amazon CLI for the following configuration tasks.
Tasks
Register a scalable target
If you haven't already done so, register the scalable target. Use the register-scalable-target command to register a specific resource in the target service as a scalable target. The following example registers an Amazon ECS service with Application Auto Scaling. Application Auto Scaling can scale the number of tasks at a minimum of 2 tasks and a maximum of 10.
Linux, macOS, or Unix
aws application-autoscaling register-scalable-target --service-namespace
ecs
\ --scalable-dimensionecs:service:DesiredCount
\ --resource-idservice/default/sample-app-service
\ --min-capacity2
--max-capacity10
Windows
aws application-autoscaling register-scalable-target --service-namespace
ecs
--scalable-dimensionecs:service:DesiredCount
--resource-idservice/default/sample-app-service
--min-capacity2
--max-capacity10
If successful, this command returns the ARN of the scalable target.
{
"ScalableTargetARN": "arn:aws-cn:application-autoscaling:region
:account-id
:scalable-target/1234abcd56ab78cd901ef1234567890ab123"
}
Note
For brevity, the examples in this topic illustrate CLI commands for an Amazon ECS service. To
specify a different scalable target, specify its namespace in --service-namespace
,
its scalable dimension in --scalable-dimension
, and its resource ID in
--resource-id
. For more information and examples for each service, see the topics
in Amazon services that you can use with Application Auto Scaling.
Create a step scaling policy
The following is an example step configuration with an adjustment type of
ChangeInCapacity
that increases the capacity of the scalable target based on the
following step adjustments (assuming a CloudWatch alarm threshold of 70 percent):
-
Increase capacity by 1 when the value of the metric is greater than or equal to 70 percent but less than 85 percent
-
Increase capacity by 2 when the value of the metric is greater than or equal to 85 percent but less than 95 percent
-
Increase capacity by 3 when the value of the metric is greater than or equal to 95 percent
Save this configuration in a file named config.json
.
{ "AdjustmentType": "ChangeInCapacity", "MetricAggregationType": "Average", "Cooldown":
60
, "StepAdjustments": [ { "MetricIntervalLowerBound":0
, "MetricIntervalUpperBound":15
, "ScalingAdjustment":1
}, { "MetricIntervalLowerBound":15
, "MetricIntervalUpperBound":25
, "ScalingAdjustment":2
}, { "MetricIntervalLowerBound":25
, "ScalingAdjustment":3
} ] }
Use the following put-scaling-policy command, along with the config.json
file that
you created, to create a scaling policy named my-step-scaling-policy
.
Linux, macOS, or Unix
aws application-autoscaling put-scaling-policy --service-namespace
ecs
\ --scalable-dimensionecs:service:DesiredCount
\ --resource-idservice/default/sample-app-service
\ --policy-namemy-step-scaling-policy
--policy-type StepScaling \ --step-scaling-policy-configurationfile://config.json
Windows
aws application-autoscaling put-scaling-policy --service-namespace
ecs
--scalable-dimensionecs:service:DesiredCount
--resource-idservice/default/sample-app-service
--policy-namemy-step-scaling-policy
--policy-type StepScaling --step-scaling-policy-configurationfile://config.json
The output includes the ARN that serves as a unique name for the policy. You need it to create CloudWatch alarms.
{
"PolicyARN": "arn:aws-cn:autoscaling:region
:123456789012
:scalingPolicy:ac542982-cbeb-4294-891c-a5a941dfa787
:resource/ecs/service/default/sample-app-service
:policyName/my-step-scaling-policy"
}
Create an alarm that triggers the scaling policy
Finally, use the following CloudWatch put-metric-alarm command to create an alarm to use with your step scaling policy. In
this example, you have an alarm based on average CPU utilization. The alarm is configured to be
in an ALARM state if it reaches a threshold of 70 percent for at least two consecutive
evaluation periods of 60 seconds. To specify a different CloudWatch metric or use your own custom
metric, specify its name in --metric-name
and its namespace in
--namespace
.
Linux, macOS, or Unix
aws cloudwatch put-metric-alarm --alarm-name
Step-Scaling-AlarmHigh-ECS:service/default/sample-app-service
\ --metric-nameCPUUtilization
--namespaceAWS/ECS
--statistic Average \ --period60
--evaluation-periods2
--threshold70
\ --comparison-operator GreaterThanOrEqualToThreshold \ --dimensions Name=ClusterName,Value=default
Name=ServiceName,Value=sample-app-service
\ --alarm-actionsPolicyARN
Windows
aws cloudwatch put-metric-alarm --alarm-name
Step-Scaling-AlarmHigh-ECS:service/default/sample-app-service
--metric-nameCPUUtilization
--namespaceAWS/ECS
--statistic Average --period60
--evaluation-periods2
--threshold70
--comparison-operator GreaterThanOrEqualToThreshold --dimensions Name=ClusterName,Value=default
Name=ServiceName,Value=sample-app-service
--alarm-actionsPolicyARN
Describe step scaling policies
You can describe all scaling policies for the specified service namespace using the following describe-scaling-policies command.
aws application-autoscaling describe-scaling-policies --service-namespace
ecs
You can filter the results to just the step scaling policies using the --query
parameter. For more information about the syntax for query
, see Controlling command output from the Amazon CLI in
the Amazon Command Line Interface User Guide.
Linux, macOS, or Unix
aws application-autoscaling describe-scaling-policies --service-namespace
ecs
\ --query 'ScalingPolicies[?PolicyType==`StepScaling`]'
Windows
aws application-autoscaling describe-scaling-policies --service-namespace
ecs
--query "ScalingPolicies[?PolicyType==`StepScaling`]"
The following is example output.
[
{
"PolicyARN": "PolicyARN",
"StepScalingPolicyConfiguration": {
"MetricAggregationType": "Average",
"Cooldown": 60,
"StepAdjustments": [
{
"MetricIntervalLowerBound": 0.0,
"MetricIntervalUpperBound": 15.0,
"ScalingAdjustment": 1
},
{
"MetricIntervalLowerBound": 15.0,
"MetricIntervalUpperBound": 25.0,
"ScalingAdjustment": 2
},
{
"MetricIntervalLowerBound": 25.0,
"ScalingAdjustment": 3
}
],
"AdjustmentType": "ChangeInCapacity"
},
"PolicyType": "StepScaling",
"ResourceId": "service/default/sample-app-service",
"ServiceNamespace": "ecs",
"Alarms": [
{
"AlarmName": "Step-Scaling-AlarmHigh-ECS:service/default/sample-app-service",
"AlarmARN": "arn:aws-cn:cloudwatch:region:012345678910:alarm:Step-Scaling-AlarmHigh-ECS:service/default/sample-app-service"
}
],
"PolicyName": "my-step-scaling-policy",
"ScalableDimension": "ecs:service:DesiredCount",
"CreationTime": 1515024099.901
}
]
Delete a step scaling policy
When you no longer need a step scaling policy, you can delete it. To delete both the scaling policy and the CloudWatch alarm, complete the following tasks.
To delete your scaling policy
Use the following delete-scaling-policy command.
Linux, macOS, or Unix
aws application-autoscaling delete-scaling-policy --service-namespace
ecs
\ --scalable-dimensionecs:service:DesiredCount
\ --resource-idservice/default/sample-app-service
\ --policy-namemy-step-scaling-policy
Windows
aws application-autoscaling delete-scaling-policy --service-namespace
ecs
--scalable-dimensionecs:service:DesiredCount
--resource-idservice/default/sample-app-service
--policy-namemy-step-scaling-policy
To delete the CloudWatch alarm
Use the delete-alarms
command. You can delete one or more alarms at a time. For example, use the following command to
delete the Step-Scaling-AlarmHigh-ECS:service/default/sample-app-service
and
Step-Scaling-AlarmLow-ECS:service/default/sample-app-service
alarms.
aws cloudwatch delete-alarms --alarm-name
Step-Scaling-AlarmHigh-ECS:service/default/sample-app-service
Step-Scaling-AlarmLow-ECS:service/default/sample-app-service