如何在Terraform中用Foreach循环遍历数组生成Datadog监控？

步骤1：定义命名空间与Slack通道的映射 #

先声明一个map类型变量，存储命名空间和对应Slack告警通道的键值对：

variable "namespace_slack_channels" {
  type = map(string)
  default = {
    "production"  = "#prod-alerts"
    "staging"     = "#staging-alerts"
    "development" = "#dev-alerts"
  }
}

步骤2：用for_each遍历映射生成监控实例 #

通过datadog_monitor资源的for_each参数遍历上述map，动态替换命名空间和Slack通道配置：

resource "datadog_monitor" "kubernetes_failed_pods" {
  for_each = var.namespace_slack_channels

  name               = "Kubernetes: Failed pods in ${each.key} namespace"
  type               = "query alert"
  query              = "sum(last_5m):sum(kubernetes.pod.status.phase{namespace:${each.key},phase:failed}) by {pod_name} > 0"
  message            = <<EOT
Failed pods detected in *${each.key}* namespace.
Notify: ${each.value}
EOT
  tags               = ["environment:${each.key}", "monitor:kubernetes-failed-pods"]
  priority           = 3

  notification {
    channel = each.value
    message = "Failed pods in ${each.key} - check Datadog dashboard for details"
  }

  notify_no_data = false
  renotify_interval = 60
}

核心逻辑说明 #

• for_each = var.namespace_slack_channels：遍历定义的映射集合，每个键值对对应一个独立的监控实例
• each.key：调用当前遍历项的键（即目标命名空间名称）
• each.value：调用当前遍历项的值（即对应的Slack告警通道）
• 监控查询通过namespace:${each.key}精准过滤指定命名空间的失败Pod
• 告警标题、消息、通知通道均与当前命名空间的配置动态绑定

内容的提问来源于stack exchange，提问作者jssteele89