如何在 terraform 中循环嵌套对象地图？

Question

我希望下面的资源能够摄取包含嵌套对象映射的变量/本地。

资源：

resource "azurerm_postgresql_flexible_server_firewall_rule" "pgsql_fw_rules" {
  for_each = ???
  server_id        = module.pgsql[each.key].id
  name             = ???
  start_ip_address = ???
  end_ip_address   = ???
}

变量/局部：

  + output1 = {
      + aa1 = {
          + azure_services = {
              + end_ip   = "0.0.0.0"
              + start_ip = "0.0.0.0"
            }
          + rule2          = {
              + end_ip   = "10.255.255.255"
              + start_ip = "10.0.0.1"
            }
          + ruleabc        = {
              + end_ip   = "1.1.1.2"
              + start_ip = "1.1.1.1"
            }
          + tf_user        = {
              + end_ip   = "127.0.0.2"
              + start_ip = "127.0.0.1"
            }
        }
      + aa2 = {
          + azure_services = {
              + end_ip   = "0.0.0.0"
              + start_ip = "0.0.0.0"
            }
          + rulexyz        = {
              + end_ip   = "1.1.1.2"
              + start_ip = "1.1.1.1"
            }
          + tf_user        = {
              + end_ip   = "127.0.0.2"
              + start_ip = "127.0.0.1"
            }
        }
    }

正如您从上面的资源中看到的，我不确定如何访问每个对象映射的内容，这就是我想要的帮助。

这里有一些限制：

• 资源中的

  each.key

  语句必须是客户的名称（aa1、aa2等）-第一个地图的键
• 我无法更改名为 output1 的原始变量/本地，因此必须从 for_each 资源（最好）或创建另一个本地来解决该问题

我假设：

• 带有嵌套

  for

  的 for_each 可以实现技巧，尽管构建它的逻辑我不明白
• 可能有一种简单（神奇的:-/）方法来引用这些内部地图字符串，例如每个.[*].value.start_ip

有什么想法吗？

Answer 1

为了让事情变得更容易，您可以：

1. 扁平化所有规则
2. 使用自定义键创建地图，例如

   ${server_name}-${rule_name}

3. 最后，迭代您要管理的资源中的地图

   for_each

工作示例：

locals {
  output1 = {
    aa1 = {
      azure_services = {
        end_ip   = "0.0.0.0"
        start_ip = "0.0.0.0"
      }
      rule2 = {
        end_ip   = "10.255.255.255"
        start_ip = "10.0.0.1"
      }
      ruleabc = {
        end_ip   = "1.1.1.2"
        start_ip = "1.1.1.1"
      }
      tf_user = {
        end_ip   = "127.0.0.2"
        start_ip = "127.0.0.1"
      }
    }
    aa2 = {
      azure_services = {
        end_ip   = "0.0.0.0"
        start_ip = "0.0.0.0"
      }
      rulexyz = {
        end_ip   = "1.1.1.2"
        start_ip = "1.1.1.1"
      }
      tf_user = {
        end_ip   = "127.0.0.2"
        start_ip = "127.0.0.1"
      }
    }
  }

  flattened_rules = flatten([
    for server_name, rules in local.output1 : [
      for rule_name, details in rules : {
        server_name = server_name
        rule_name   = rule_name
        start_ip    = details.start_ip
        end_ip      = details.end_ip
      }
    ]
  ])

  # Create a map with key as "${server_name}-${rule_name}" and value as the rule details
  rules_map = {
    for rule in local.flattened_rules : "${rule.server_name}-${rule.rule_name}" => rule
  }
}

resource "null_resource" "example" {
  for_each = local.rules_map

  triggers = {
    server_name = each.value.server_name
    rule_name   = each.value.rule_name
    start_ip    = each.value.start_ip
    end_ip      = each.value.end_ip
  }
}

跑步

terraform plan

：

Terraform used the selected providers to generate the following execution
plan. Resource actions are indicated with the following symbols:
  + create

Terraform will perform the following actions:

  # null_resource.example["aa1-azure_services"] will be created
  + resource "null_resource" "example" {
      + id       = (known after apply)
      + triggers = {
          + "end_ip"      = "0.0.0.0"
          + "rule_name"   = "azure_services"
          + "server_name" = "aa1"
          + "start_ip"    = "0.0.0.0"
        }
    }

  # null_resource.example["aa1-rule2"] will be created
  + resource "null_resource" "example" {
      + id       = (known after apply)
      + triggers = {
          + "end_ip"      = "10.255.255.255"
          + "rule_name"   = "rule2"
          + "server_name" = "aa1"
          + "start_ip"    = "10.0.0.1"
        }
    }

  # null_resource.example["aa1-ruleabc"] will be created
  + resource "null_resource" "example" {
      + id       = (known after apply)
      + triggers = {
          + "end_ip"      = "1.1.1.2"
          + "rule_name"   = "ruleabc"
          + "server_name" = "aa1"
          + "start_ip"    = "1.1.1.1"
        }
    }

  # null_resource.example["aa1-tf_user"] will be created
  + resource "null_resource" "example" {
      + id       = (known after apply)
      + triggers = {
          + "end_ip"      = "127.0.0.2"
          + "rule_name"   = "tf_user"
          + "server_name" = "aa1"
          + "start_ip"    = "127.0.0.1"
        }
    }

  # null_resource.example["aa2-azure_services"] will be created
  + resource "null_resource" "example" {
      + id       = (known after apply)
      + triggers = {
          + "end_ip"      = "0.0.0.0"
          + "rule_name"   = "azure_services"
          + "server_name" = "aa2"
          + "start_ip"    = "0.0.0.0"
        }
    }

  # null_resource.example["aa2-rulexyz"] will be created
  + resource "null_resource" "example" {
      + id       = (known after apply)
      + triggers = {
          + "end_ip"      = "1.1.1.2"
          + "rule_name"   = "rulexyz"
          + "server_name" = "aa2"
          + "start_ip"    = "1.1.1.1"
        }
    }

  # null_resource.example["aa2-tf_user"] will be created
  + resource "null_resource" "example" {
      + id       = (known after apply)
      + triggers = {
          + "end_ip"      = "127.0.0.2"
          + "rule_name"   = "tf_user"
          + "server_name" = "aa2"
          + "start_ip"    = "127.0.0.1"
        }
    }

Plan: 7 to add, 0 to change, 0 to destroy.