从解码 Json 文件中获取属性并将它们作为字符串添加到 terraform 上的变量中

问题描述

我有解码 JSON 和 CSV 文件的 terraform 代码

provider "azurerm" {
    features{}
}

locals {
      resource_groupname = csvdecode(file("./ResourceTypes.csv"))
      json_data = jsondecode(file("${path.module}/Monitor.json"))
    }

    resource "azurerm_resource_group" "Main" {
      count    =  length(local.resource_groupname)
      name     =  local.resource_groupname[count.index].resourcetype
      location = "north europe"
    } 
     resource "azurerm_resource_group" "name" {
       # (other settings)
       name = local.json_data.name
       location = "north europe"
     }

      output "local_json_data" {
        value = local.json_data
      } 

当我运行 terraform 代码时，我的输出如下：

  # azurerm_resource_group.Main[212] will be created
  + resource "azurerm_resource_group" "Main" {
      + id       = (kNown after apply)
      + location = "northeurope"
      + name     = "Wandisco.Fusion"
    }

  # azurerm_resource_group.name will be created
  + resource "azurerm_resource_group" "name" {
      + id       = (kNown after apply)
      + location = "northeurope"
      + name     = "Azure-apimanagement-FailedRequests"
    }

Plan: 214 to add,0 to change,0 to destroy.

Changes to Outputs:
  + local_json_data = {
      + classification = "anomaly"
      + id             = 16020941
      + message        = <<-EOT
            Azure - API Management - Failed Requests
            {{name.name}} exceeds the prevIoUsly estimated average.

            Please refer to the following reaction process:
            https://apptemetry/kNowledgebase/Article.aspx?id=54321

            Alerts generate an aim ticket,viewable here (search via CI or Alert Name):
            https://apptemetry/aim/alertsearch.aspx
        EOT
      + name           = "Azure-apimanagement-FailedRequests"
      + options        = {
          + escalation_message  = ""
          + include_tags        = true
          + locked              = false
          + new_host_delay      = 300
          + no_data_timeframe   = null
          + notify_audit        = false
          + notify_no_data      = false
          + renotify_interval   = 0
          + require_full_window = true
          + silenced            = {}
          + threshold_windows   = {
              + recovery_window = "last_15m"
              + trigger_window  = "last_15m"
            }
          + thresholds          = {
              + critical          = 1
              + critical_recovery = 0
            }
          + timeout_h           = 1
        }
      + priority       = null
      + query          = "avg(last_4h):anomalies(avg:azure.apimanagement_service.Failed_requests{*} by {name},'agile',2,direction='above',alert_window='last_15m',interval=60,count_default_zero='true',seasonality='hourly') >= 1"
      + tags           = [
          + "Sev:54",]
      + type           = "query alert"
    }

有没有办法可以保存 JSON 和 CSV 文件中的属性，并将其作为 terraform 中的字符串保存到变量中？然后从变量中创建一个新的 terraform 脚本？

我的最终目标是尝试从 JSON 文件创建一个 terraform 脚本。

解决方法

如果我没记错的话。您想在另一个 Terraform 脚本中引用从 JSON 文件加载值的变量。所以你可以使用 Terraform 模块来实现它。例子如下：

./modules/monitor.json

{
  "resource_group_name": "testGroup"
}

./modules/main.tf

locals {
  json_data = jsondecode(file("./Monitor.json"))
}

output "group_name" {
  value = local.json_data.resource_group_name
}

ma​​in.tf

modules "groups" {
  source = "./modules/"
}

resource "azurerm_resource_group" "group" {
  name = module.groups.group_name
  ...
}

但是为什么不直接使用 locals 块来引用 JSON 文件中的数据？它更容易并且符合逻辑。当然，这是一个建议，您可以根据需要使用该方法。