Support customized metric specification in autoscaling predictive scaling policy

.changelog/22451.txt

@@ -0,0 +1,2 @@
+```release-note:enhancement
+resource/aws_autoscaling_policy: Add customized metric specification in `predictive_scaling_configuration` argument

website/docs/r/autoscaling_policy.html.markdown

@@ -159,9 +159,12 @@ The following arguments are supported:
 
 The following arguments are supported:
 
-* `predefined_load_metric_specification` - (Optional) The load metric specification.
+* `predefined_load_metric_specification` - (Optional) The predefined load metric specification.
 * `predefined_metric_pair_specification` - (Optional) The metric pair specification from which Amazon EC2 Auto Scaling determines the appropriate scaling metric and load metric to use.
-* `predefined_scaling_metric_specification` - (Optional) The scaling metric specification.
+* `predefined_scaling_metric_specification` - (Optional) The predefined scaling metric specification.
+* `customized_scaling_metric_specification` - (Optional) The customized scaling metric specification. 
+* `customized_load_metric_specification` - (Optional) The customized load metric specification.
+* `customized_capacity_metric_specification` - (Optional) The customized capacity metric specification. The field is only valid when you use `customized_load_metric_specification`
 
 ##### predefined_load_metric_specification
 
@@ -184,6 +187,89 @@ The following arguments are supported:
 * `predefined_metric_type` - (Required) Describes a scaling metric for a predictive scaling policy. Valid values are `ASGAverageCPUUtilization`, `ASGAverageNetworkIn`, `ASGAverageNetworkOut`, or `ALBRequestCountPerTarget`.
 * `resource_label` - (Required) A label that uniquely identifies a specific Application Load Balancer target group from which to determine the request count served by your Auto Scaling group.
 
+##### customized_scaling_metric_specification
+The following arguments are supported:
+
+* `metric_data_queries` - (Required) A list of up to 10 structure that defines custom scaling metric in predictive scaling policy
+
+##### customized_load_metric_specification
+The following arguments are supported:
+* `metric_data_queries` - (Required) A list of up to 10 structure that defines custom load metric in predictive scaling policy
+
+##### customized_capacity_metric_specification
+The following arguments are supported:
+* `metric_data_queries` - (Required) A list of up to 10 structure that defines custom capacity metric in predictive scaling policy
+
+##### metric_data_queries
+The following arguments are supported:
+* `id` - (Required) A short name for the metric used in predictive scaling policy.
+* `expression` - (Optional) The math expression used on the returned metric. You must specify either `expression` or `metric_stat`, but not both.
+* `metric_stat` - (Optional) A structure that defines CloudWatch metric to be used in predictive scaling policy. You must specify either `expression` or `metric_stat`, but not both.
+* `label` - (Optional) A human-readable label for this metric or expression.
+* `return_data` - (Optional) A boolean that indicates whether to return the timestamps and raw data values of this metric, the default it true
+
+##### metric_stat
+The following arguments are supported:
+* `metric` - (Required) A structure that defines the CloudWatch metric to return, including the metric name, namespace, and dimensions.
+* `stat` - (Required) The statistic of the metrics to return.
+* `unit` - (Optional) The unit of the metrics to return.
+
+##### metric
+* `metric_name` - (Required) The name of the metric.
+* `namespace` - (Required) The namespace of the metric.
+* `dimensions` - (Optional) The dimensions of the metric.
+
+### Example predictive scaling policy using custom metric
+
+```terraform
+resource "aws_autoscaling_policy" "example" {
+  # ... other configuration ...
+
+  predictive_scaling_configuration {
+    metric_specification {
+      target_value = 10
+      customized_scaling_metric_specification {
+        metric_data_queries { 
+          id = "scaling"
+          metric_stat { 
+            metric { 
+              metric_name = "CPUUtilization"
+              namespace = "AWS/EC2"
+              dimensions { 
+                name = "AutoScalingGroupName"
+                value = "ASG-myapp" 
+              } 
+            }
+            stat = "Average" 
+          }
+        }
+      }
+      customized_load_metric_specification {
+        metric_data_queries { 
+          id = "load_sum"
+          expression = "SUM(SEARCH('{AWS/EC2,AutoScalingGroupName} MetricName=\"CPUUtilization\" ASG-myapp', 'Sum', 3600))"        
+        }
+      }
+      customized_capacity_metric_specification {
+        metric_data_queries { 
+          id = "capacity_sum"
+          expression = "SUM(SEARCH('{AWS/AutoScaling,AutoScalingGroupName} MetricName=\"GroupInServiceIntances\" ASG-myapp', 'Average', 300))"
+          return_data = false
+        }
+        metric_data_queries { 
+          id = "load_sum"
+          expression = "SUM(SEARCH('{AWS/EC2,AutoScalingGroupName} MetricName=\"CPUUtilization\" ASG-myapp', 'Sum', 300))"
+          return_data = false
+        }
+        metric_data_queries { 
+          id = "weighted_average"
+          expression = "load_sum / capacity_sum"
+        }
+      }
+    }
+  }
+}
+```
 ## Attributes Reference
 
 In addition to all arguments above, the following attributes are exported: