我正在尝试配置2个公共IP地址和2个网络接口。到目前为止,我写的是:
resource "azurerm_public_ip" "example" {
name = "test-pip${count.index}"
count = 2
location = "${azurerm_resource_group.rc.location}"
resource_group_name = "${azurerm_resource_group.rc.name}"
allocation_method = "Dynamic"
idle_timeout_in_minutes = 30
}
output "public_ip_address" {
value = "${azurerm_public_ip.example.*.id}"
}
resource "azurerm_network_interface" "main" {
name = "test${count.index}"
count = 2
location = "${azurerm_resource_group.rc.location}"
resource_group_name = "${azurerm_resource_group.rc.name}"
ip_configuration {
name = "testconfiguration1${count.index}"
subnet_id = "${azurerm_subnet.internal.id}"
private_ip_address_allocation = "Dynamic"
public_ip_address_id = "${azurerm_public_ip.example[count.index].id}"
}
}
稍后,我将使用这两个IP和NI将它们分配给2台VM计算机。
当我运行terraform plan
时,出现一条错误消息:
Terraform版本为"v0.12.3"
,Azure提供者版本为"v1.40.0"
答案 0 :(得分:1)
鉴于这是Terraform 0.12,而不是问题语法所暗示的Terraform 0.11,所以实际错误出在特定的导出属性中。要访问azurerm_public_ip.example
资源导出的ip地址,我们需要使用ip_address
导出的属性,而不是id
。这就是为什么为无效密钥抛出错误的原因,尽管错误中的特定引用确实确实具有误导性。
我们可以通过以下方式更新您的代码以解决此问题:
ip_configuration {
name = "testconfiguration1${count.index}"
subnet_id = "${azurerm_subnet.internal.id}"
private_ip_address_allocation = "Dynamic"
public_ip_address_id = "${azurerm_public_ip.example[count.index].ip_address}"
}
答案 1 :(得分:0)
你可以尝试像这样更改代码,它对我有用:
public_ip_address_id = "${element(azurerm_public_ip.example.*.id, count.index)}"