ScalingInstruction
Describes a scaling instruction for a scalable resource in a scaling plan. Each scaling instruction applies to one resource.
AWS Auto Scaling creates target tracking scaling policies based on the scaling instructions. Target tracking scaling policies adjust the capacity of your scalable resource as required to maintain resource utilization at the target value that you specified.
AWS Auto Scaling also configures predictive scaling for your Amazon EC2 Auto Scaling groups using a subset of parameters, including the load metric, the scaling metric, the target value for the scaling metric, the predictive scaling mode (forecast and scale or forecast only), and the desired behavior when the forecast capacity exceeds the maximum capacity of the resource. With predictive scaling, AWS Auto Scaling generates forecasts with traffic predictions for the two days ahead and schedules scaling actions that proactively add and remove resource capacity to match the forecast.
We recommend waiting a minimum of 24 hours after creating an Auto Scaling group to configure predictive scaling. At minimum, there must be 24 hours of historical data to generate a forecast. For more information, see Best Practices for AWS Auto Scaling in the AWS Auto Scaling User Guide.
Types
Properties
The customized load metric to use for predictive scaling. This parameter or a PredefinedLoadMetricSpecification is required when configuring predictive scaling, and cannot be used otherwise.
Controls whether dynamic scaling by AWS Auto Scaling is disabled. When dynamic scaling is enabled, AWS Auto Scaling creates target tracking scaling policies based on the specified target tracking configurations.
The maximum capacity of the resource. The exception to this upper limit is if you specify a non-default setting for PredictiveScalingMaxCapacityBehavior.
The minimum capacity of the resource.
The predefined load metric to use for predictive scaling. This parameter or a CustomizedLoadMetricSpecification is required when configuring predictive scaling, and cannot be used otherwise.
Defines the behavior that should be applied if the forecast capacity approaches or exceeds the maximum capacity specified for the resource. The default value is SetForecastCapacityToMaxCapacity.
The size of the capacity buffer to use when the forecast capacity is close to or exceeds the maximum capacity. The value is specified as a percentage relative to the forecast capacity. For example, if the buffer is 10, this means a 10 percent buffer, such that if the forecast capacity is 50, and the maximum capacity is 40, then the effective maximum capacity is 55.
The predictive scaling mode. The default value is ForecastAndScale. Otherwise, AWS Auto Scaling forecasts capacity but does not create any scheduled scaling actions based on the capacity forecast.
The ID of the resource. This string consists of the resource type and unique identifier.
The scalable dimension associated with the resource.
Controls whether a resource's externally created scaling policies are kept or replaced.
The amount of time, in seconds, to buffer the run time of scheduled scaling actions when scaling out. For example, if the forecast says to add capacity at 10:00 AM, and the buffer time is 5 minutes, then the run time of the corresponding scheduled scaling action will be 9:55 AM. The intention is to give resources time to be provisioned. For example, it can take a few minutes to launch an EC2 instance. The actual amount of time required depends on several factors, such as the size of the instance and whether there are startup scripts to complete.
The namespace of the AWS service.
The target tracking configurations (up to 10). Each of these structures must specify a unique scaling metric and a target value for the metric.