AWS Auto Scaling
API Reference (API Version 2018-01-06)


Represents a CloudWatch metric of your choosing that can be used for predictive scaling.

For predictive scaling to work with a customized load metric specification, AWS Auto Scaling needs access to the Sum and Average statistics that CloudWatch computes from metric data.

When you choose a load metric, make sure that the required Sum and Average statistics for your metric are available in CloudWatch and that they provide relevant data for predictive scaling. The Sum statistic must represent the total load on the resource, and the Average statistic must represent the average load per capacity unit of the resource. For example, there is a metric that counts the number of requests processed by your Auto Scaling group. If the Sum statistic represents the total request count processed by the group, then the Average statistic for the specified metric must represent the average request count processed by each instance of the group.

For information about terminology, available metrics, or how to publish new metrics, see Amazon CloudWatch Concepts in the Amazon CloudWatch User Guide.



The dimensions of the metric.

Conditional: If you published your metric with dimensions, you must specify the same dimensions in your customized load metric specification.

Type: Array of MetricDimension objects

Required: No


The name of the metric.

Type: String

Required: Yes


The namespace of the metric.

Type: String

Required: Yes


The statistic of the metric. Currently, the value must always be Sum.

Type: String

Valid Values: Average | Minimum | Maximum | SampleCount | Sum

Required: Yes


The unit of the metric.

Type: String

Required: No

See Also

For more information about using this API in one of the language-specific AWS SDKs, see the following:

On this page: