Faster auto scaling with high-resolution metrics

On the Clusters page, choose the cluster you want to create the service in.

In the Services section, choose Create.

In the Create Service form, in the Monitoring configuration section, choose Add metric configuration.

For Metric name, select the metric you want to collect at high resolution. Choose CPUUtilization or MemoryUtilization.

For Resolution (seconds), select 20.

(Optional) To add both metrics at high resolution, choose Add metric configuration again and repeat the previous two steps for the other metric.

In the Service auto scaling section, choose Use service auto scaling.

For Minimum number of tasks, enter the lower limit of the number of tasks for service auto scaling to use. The desired count will not go below this count.

For Maximum number of tasks, enter the upper limit of the number of tasks for service auto scaling to use. The desired count will not go above this count.

For Scaling policy type, choose Target tracking.

For Metric type, select the corresponding high-resolution metric. For example, choose ECSServiceAverageCPUUtilizationHighResolution if you configured CPUUtilization at 20-second resolution, or ECSServiceAverageMemoryUtilizationHighResolution if you configured MemoryUtilization at 20-second resolution.

For Target value, enter the target utilization percentage. Service auto scaling scales out your capacity until the average utilization is at the target value, or until it reaches the maximum number of tasks you specified.

Under Additional Settings, configure the following:

1. For Scale-out cooldown period, enter the amount of time in seconds to wait for a previous scale-out activity to take effect.

2. For Scale-in cooldown period, enter the amount of time in seconds after a scale-in activity completes before another scale-in activity can start.

Choose Create to create the service.

Create the service with monitoring configuration using the create-service command.

aws ecs create-service --region us-east-1 \
    --cluster my-cluster \
    --service-name my-service \
    --task-definition my-app:1 \
    --desired-count 2 \
    --monitoring "metricConfigurations=[{metricNames=[CPUUtilization,MemoryUtilization],resolutionSeconds=20}]"

Register the service as a scalable target using the register-scalable-target command.

aws application-autoscaling register-scalable-target \
    --service-namespace ecs \
    --resource-id service/my-cluster/my-service \
    --scalable-dimension ecs:service:DesiredCount \
    --min-capacity 1 \
    --max-capacity 10 \
    --region us-east-1

Create a target tracking scaling policy with a high-resolution predefined metric using the put-scaling-policy command.

aws application-autoscaling put-scaling-policy \
    --service-namespace ecs \
    --resource-id service/my-cluster/my-service \
    --scalable-dimension ecs:service:DesiredCount \
    --policy-name my-highres-cpu-policy \
    --policy-type TargetTrackingScaling \
    --target-tracking-scaling-policy-configuration '{
        "TargetValue": 50.0,
        "PredefinedMetricSpecification": {
            "PredefinedMetricType": "ECSServiceAverageCPUUtilizationHighResolution"
        }
    }' \
    --region us-east-1

Verify the policy was created using the describe-scaling-policies command.

aws application-autoscaling describe-scaling-policies \
    --service-namespace ecs \
    --resource-id service/my-cluster/my-service \
    --scalable-dimension ecs:service:DesiredCount \
    --region us-east-1

On the Clusters page, choose the cluster that contains the service you want to update.

In the Services section, select the service and choose Update.

In the Update Service form, in the Monitoring configuration section, choose Add metric configuration.

For Metric name, select the metric you want to collect at high resolution. Choose CPUUtilization or MemoryUtilization.

For Resolution (seconds), select 20.

(Optional) To add both metrics at high resolution, choose Add metric configuration again and repeat the previous two steps for the other metric.

Choose Update to submit the change. This triggers a service deployment.

Wait for the deployment to complete. You can monitor progress on the service details page in the Deployments tab.

Update the service with monitoring configuration using the update-service command.

aws ecs update-service --region us-east-1 \
    --cluster my-cluster \
    --service my-service \
    --monitoring "metricConfigurations=[{metricNames=[CPUUtilization,MemoryUtilization],resolutionSeconds=20}]"

Wait for the deployment to complete. You can monitor progress using the describe-services command and checking the deployment status.

aws ecs describe-services --region us-east-1 \
    --cluster my-cluster \
    --services my-service

After the deployment completes, verify the monitoring configuration is active by describing the service revision. Get the active service revision ARN from the describe-services output, then run:

aws ecs describe-service-revisions --region us-east-1 \
    --service-revision-arns arn:aws:ecs:us-east-1:012345678910:service-revision/my-cluster/my-service/1234567890123456789

Confirm the monitoring configuration appears in the response:

{
    "serviceRevisions": [
        {
            ...
            "monitoring": {
                "metricConfigurations": [
                    {
                        "metricNames": ["CPUUtilization", "MemoryUtilization"],
                        "resolutionSeconds": 20
                    }
                ]
            }
        }
    ]
}

Register the service as a scalable target if not already registered, using the register-scalable-target command.

aws application-autoscaling register-scalable-target \
    --service-namespace ecs \
    --resource-id service/my-cluster/my-service \
    --scalable-dimension ecs:service:DesiredCount \
    --min-capacity 1 \
    --max-capacity 10 \
    --region us-east-1

Create a target tracking scaling policy with a high-resolution predefined metric using the put-scaling-policy command. The following example targets 50% CPU utilization:

aws application-autoscaling put-scaling-policy \
    --service-namespace ecs \
    --resource-id service/my-cluster/my-service \
    --scalable-dimension ecs:service:DesiredCount \
    --policy-name my-highres-cpu-policy \
    --policy-type TargetTrackingScaling \
    --target-tracking-scaling-policy-configuration '{
        "TargetValue": 50.0,
        "PredefinedMetricSpecification": {
            "PredefinedMetricType": "ECSServiceAverageCPUUtilizationHighResolution"
        }
    }' \
    --region us-east-1

Verify the policy was created:

aws application-autoscaling describe-scaling-policies \
    --service-namespace ecs \
    --resource-id service/my-cluster/my-service \
    --scalable-dimension ecs:service:DesiredCount \
    --region us-east-1