Help required in addressing invalid metric name error when provisioning a cloud watch metric alarm

[d-saravanan]

I am creating an AWS cloud watch alarm,for EBS volume of an EC2 server using terraform, when running in terraform, i get the below error

Error: creating CloudWatch Metric Alarm (ec2-ebs-disk-read-high-avg-read-latency-alarm): ValidationError: Invalid metrics list
status code: 400, request id: 019e6d14-0d1b-4f3e-bb50-4a6f0f4a592c

The below script has the metric parameters and query, which failed and to tryout using an expression, I tried with the next script given below

resource "aws_cloudwatch_metric_alarm" "ec2-ebs-disk-high-avg-read-latency-alarm" {
  alarm_name          = "ebs-disk-read-high-avg-read-latency-alarm"
  alarm_description   = "Alarm triggered when Avg read latency metric is greater than or equal to 1.5 s for a 3-minute period for ${var.server_name}"
  comparison_operator = "GreaterThanOrEqualToThreshold"
  evaluation_periods  = 1
  threshold           = 1500
  insufficient_data_actions = []

  metric_query {
    id          = "AvgReadLatency"
    expression  = "IF(rmq2 !=0, (rmq1 / rmq2) * 1000, 0)"
    label       = "Average Read Latency"
    return_data = "true"
  }

  metric_query {
    id = "rmq1"
    metric {
      metric_name = "VolumeTotalReadTime"
      namespace   = "AWS/EBS"
      period      = 180
      stat        = "Sum"
#      unit        = "Count"
      dimensions  = {
        VolumeId = var.root_volume_id
      }
    }
  }

  metric_query {
    id = "rmq2"
    metric {
      metric_name = "VolumeReadOps"
      namespace   = "AWS/EBS"
      period      = 180
      stat        = "Sum"
#      unit        = "Count"
      dimensions  = {
        VolumeId = var.root_volume_id
      }
    }
  }

  alarm_actions = [var.alarm_actions]
  ok_actions    = [var.alarm_actions]
}

** Testing with expression model without metrics similar to that of cloud formation json exported from aws **

resource "aws_cloudwatch_metric_alarm" "ec2-ebs-disk-high-avg-read-latency-alarm" {
  alarm_name                = "ebs-disk-read-high-avg-read-latency-alarm"
  alarm_description         = "Alarm triggered when Avg read latency metric is greater than or equal to 1.5 s for a 3-minute period "
  comparison_operator       = "GreaterThanOrEqualToThreshold"
  evaluation_periods        = 1
  threshold                 = 1500
  insufficient_data_actions = []
  actions_enabled           = true

  metric_query {
    id          = "avg_read_latency"
    expression  = "((m1 / m2) * 1000)"
#    label       = "Average Read Latency"
    return_data = true
  }

  metric_query {
    id          = "m1"
    expression  = "SELECT SUM(VolumeTotalWriteTime)\nFROM SCHEMA(\"AWS/EBS\", VolumeId)"
#    label       = "sum of total write time, all volumes"
    return_data = false
    period      = 180
  }

  metric_query {
    id          = "m2"
    expression  = "SELECT SUM(VolumeReadOps)\nFROM SCHEMA(\"AWS/EBS\", VolumeId)"
#    label       = "Sum of read ops volume time, all volumes"
    return_data = false
    period      = 180
  }

  alarm_actions = [var.alarm_actions]
  ok_actions    = [var.alarm_actions]
}

[peterwoodworth]

Does this have anything to do with CDK? If not, I think you might find a better place to report this, like the terraform repo

[d-saravanan]

I found that there is a terraform repo stating that the provider has raised the validation error and that terraform has no issues since this has been raised by aws provider. Hence requested for help here

[peterwoodworth]

Terraform doesn't have anything to do with CDK - if you are running into AWS issues, you would need to raise an issue with premium support

[github-actions[bot]]

Hello! Reopening this discussion to make it searchable.