customized_metric_specification metrics is missing #225
Description
Description
While trying to add a custom metric for one of my project, I was not able to find how to specify... well the metrics.
It looks like this field is not handle. Here the definition of customized_metric_specification, and here where this module build it. If I’m not wrong the metrics attribute is missing.
- ✋ I have searched the open/closed issues and my issue is not listed.
Versions
-
Module version [Required]:
-
Terraform version: 1.5
-
Provider version(s): v5.63.1
Reproduction Code [Required]
module "web_service" {
source = "terraform-aws-modules/ecs/aws//modules/service"
version = "5.11.4"
name = "api"
...
autoscaling_min_capacity = var.autoscaling_min_capacity
autoscaling_max_capacity = var.autoscaling_max_capacity
autoscaling_policies = {
cpu = {
policy_type = "TargetTrackingScaling"
target_tracking_scaling_policy_configuration = {
target_value = 60
customized_metric_specification = {
metric_name = "AdjustedCpuUtilization"
namespace = "ECS/Custom"
statistic = "Average"
unit = "Percent"
dimensions = [
{
name = "ClusterName"
value = local.cluster_name
},
{
name = "ServiceName"
value = local.service_name
}
]
metrics = [
{
id = "m1"
label = "CpuUtilized"
metric_name = "CpuUtilized"
namespace = "ECS/ContainerInsights"
stat = "Average"
dimensions = [
{
name = "ClusterName"
value = local.cluster_name
},
{
name = "ServiceName"
value = local.service_name
}
]
return_data = false
},
{
id = "m2"
label = "CpuReserved"
metric_name = "CpuReserved"
namespace = "ECS/ContainerInsights"
stat = "Average"
dimensions = [
{
name = "ClusterName"
value = local.cluster_name
},
{
name = "ServiceName"
value = local.service_name
}
]
return_data = false
},
{
id = "m3"
label = "Running Task Count"
metric_name = "RunningTaskCount"
namespace = "ECS/ContainerInsights"
stat = "Average"
dimensions = [
{
name = "ClusterName"
value = local.cluster_name
},
{
name = "ServiceName"
value = local.service_name
}
]
return_data = false
},
{
id = "m4"
label = "Mitigated Host Count"
metric_name = "MitigatedHostCount"
namespace = "AWS/ApplicationELB"
stat = "Average"
dimensions = [
{
name = "LoadBalancer"
value = data.aws_alb.external[0].arn_suffix
},
{
name = "TargetGroup"
value = aws_lb_target_group.web[0].arn_suffix
}
]
return_data = false
},
# Expression 1: CpuUtilization = (100 * CpuUtilized) / CpuReserved
{
id = "e1"
expression = "(100 * m1) / m2"
label = "CpuUtilization"
return_data = false
},
# Expression 2: AdjustedCpuUtilization = (CpuUtilization * NumberOfTasks) / (NumberOfTasks - MitigatedHostCount)
{
id = "e2"
expression = "(e1 * m3) / (m3 - m4)"
label = "AdjustedCpuUtilization"
return_data = true
}
]
}
scale_in_cooldown = 300
scale_out_cooldown = 60
}
}
}
Expected behavior
Be able to create an autoscalling policy based on a custom metrics.
Actual behavior
metrics is not handle.