Terraform 变量：如何以更扁平的方式重构 vnets/subnets/nsgs 的 Terraform 变量数据结构？

Question

我正在使用的以下 variables.vnet 数据结构在即将使用 count/for_each 和动态创建 subnets/vnets 方面造成了太多的复杂性。我想要一个更简单的解决方案。我仍然希望能够在各自的 vnet 中对相关子网进行分组，但也许可以摆脱过多的地图列表？

有人可以帮我制定更好的结构吗？

variable "vnets" {
  default = [
    {
      vnet_name     = "first-vnet"
      address_space = "10.250.0.0"
      network_size  = 16
      subnets = [
        {
          name                     = "test1-subnet"
          bitmask                  = 24
          netnum                   = 1
          service_endponts_enabled = false
          network_security_group   = "test1-nsg"
          security_group_rules = [
            {
              name                       = "test1-sg"
              priority                   = 100
              direction                  = "inbound"
              access                     = "allow"
              protocol                   = "tcp"
              source_port_range          = "*"
              destination_port_range     = "*"
              source_address_prefix      = "*"
              destination_address_prefix = "172.16.2.0/24"
            },
          ]
        },
        {
          name                     = "test2-subnet"
          bitmask                  = 24
          netnum                   = 2
          service_endponts_enabled = false
          network_security_group   = "test2-nsg"
          security_group_rules = [
            {
              name                       = "test2-sg"
              priority                   = 100
              direction                  = "inbound"
              access                     = "allow"
              protocol                   = "tcp"
              source_port_range          = "*"
              destination_port_range     = "*"
              source_address_prefix      = "172.16.0.0/24"
              destination_address_prefix = "172.16.4.0/24"
            },
          ]
        },
        {
          name                     = "test3-subnet"
          bitmask                  = 24
          netnum                   = 3
          service_endponts_enabled = true
          network_security_group   = "test3-nsg"
          security_group_rules = [
            {
              name                       = "test3-sg"
              priority                   = 100
              direction                  = "inbound"
              access                     = "allow"
              protocol                   = "tcp"
              source_port_range          = "*"
              destination_port_range     = "*"
              source_address_prefix      = "172.16.2.0/24"
              destination_address_prefix = "172.16.8.0/22"
            },
          ]
        }
      ]
    },
    {
      vnet_name     = "second-vnet"
      address_space = "10.251.0.0"
      network_size  = 16
      subnets       = []
    },
    {
      vnet_name     = "third-vnet"
      address_space = "10.252.0.0"
      network_size  = 16
      subnets       = []
    }
  ]
}

Answer 1

variables.vnet数据结构很好，除了NSG规则外，没必要把规则放在子网属性里面。要知道，一个子网只能附加一个 NSG。所以你只需要设置哪个 NSG 将附加到子网。您可以单独为 NSG 和规则使用另一个变量。

例如，对于子网，它可能会点击：

...
subnets = [
        {
          name                     = "test1-subnet"
          bitmask                  = 24
          netnum                   = 1
          service_endponts_enabled = false
          network_security_group   = "test1-nsg"
        },
        ...
      ]

Terraform 变量：如何以更扁平的方式重构 vnets/subnets/nsgs 的 Terraform 变量数据结构？

Terraform variables: How can I restructure my terraform variables data-structure for vnets/subnets/nsgs in a more flatter way?

terraform

terraform-provider-azure