🧠 Metric #4: Task Restart Rate (Failure Monitoring) Use TaskStoppedCount metric to detect frequent task restarts:

restart_rate.tf

resource "awscc_applicationsignals_service_level_objective" "task_restart_slo" {
  name        = "ecs-task-restart-slo"
  description = "Monitor ECS task restart frequency"

  sli = {
    comparison_operator = "LessThan"
    metric_threshold    = 1.0 # 1 or fewer per 5 min

    sli_metric = {
      metric_data_queries = [{
        id = "restarts"
        metric_stat = {
          metric = {
            namespace   = "ECS/ContainerInsights"
            metric_name = "TaskStoppedCount"
            dimensions = [
              {
                name  = "ClusterName"
                value = "my-cluster-name"
              },
              {
                name  = "ServiceName"
                value = "my-service-name"
              }
            ]
          }
          period = 300
          stat   = "Sum"
        }
      }]
    }
  }

  goal = {
    attainment_goal = 99.0
    interval = {
      rolling_interval = {
        duration      = 7
        duration_unit = "DAY"
      }
    }
  }

  burn_rate_configurations = [
    { look_back_window_minutes = 60 },
    { look_back_window_minutes = 1440 }
  ]
}