使用模块从aws导入一个现有的s3

Question

我有一个模块要创建s3，从模块中我创建了一个桶，在aws中我手动创建了另一个桶，我试图导入它。

main.tf - of the module that is called "s3"

resource "aws_s3_bucket" "s3" {
count         = var.create_bucket && length(var.s3_bucket_names) > 0 ? length(var.s3_bucket_names) : 0
bucket        = "${var.s3_bucket_names[count.index]}-bucket"
force_destroy = var.force_destroy
}
resource "aws_s3_bucket_public_access_block" "s3_access" {
count  = var.create_bucket && length(var.s3_bucket_names) > 0 ? length(var.s3_bucket_names) : 0
bucket = aws_s3_bucket.s3[count.index].id
block_public_acls       = var.block_public_acls
block_public_policy     = var.block_public_policy
ignore_public_acls      = var.ignore_public_acls
restrict_public_buckets = var.restrict_public_buckets
}

main.tf

module "s3" {
count = var.create_bucket ? 1 : 0
source          = "./modules/s3"
s3_bucket_names = var.s3_bucket_names
create_bucket   = var.create_bucket
force_destroy = var.force_destroy
block_public_acls       = var.block_public_acls
block_public_policy     = var.block_public_policy
ignore_public_acls      = var.ignore_public_acls
restrict_public_buckets = var.restrict_public_buckets
}

在我的tfvar中有我最初创建的s3，我认为我想要导入的新s3应该添加到tfvar中…

tfvars

create_bucket = true
s3_bucket_names = [
"my-k-frontend-test",
"my-k-frontend-test-2"
]
block_public_acls       = true
block_public_policy     = true
ignore_public_acls      = true
restrict_public_buckets = true

然后我运行命令导入，它说正确完成了…

$ terraform import -var-file=test.tfvars module.s3.aws_s3_bucket.s3 my-k-frontend-test-2-bucket

data.aws_subnet.subnets_id["subnet-xxxxxxxxxx"]: Read complete after 1s [id=subnet-xxxxxxxxxx]
data.aws_subnet.subnets_id["subnet-xxxxxxxxxx"]: Read complete after 1s [id=subnet-xxxxxxx]
data.aws_subnet.subnets_id["subnet-xxxxxxxxx"]: Read complete after 1s [id=subnet-xxxxxxxx]
Import successful!
The resources that were imported are shown above. These resources are now in
your Terraform state and will henceforth be managed by Terraform.

但是当执行计划时，它好像还没有在状态文件中..

data.aws_subnet.subnets_id["subnet-xxxxxx"]: Read complete after 1s [id=subnet-xxxxx]
data.aws_subnet.subnets_id["subnet-xxxxxxx"]: Read complete after 1s [id=subnet-xxxxxx]
module.s3[0].aws_s3_bucket_public_access_block.s3_access[0]: Refreshing state... [id=my-k-frontend-test-2-bucket]
Plan: 2 to add, 0 to change, 0 to destroy.

是my-k-front -test-2 bucket的aws_s3_bucket和aws_s3_bucket_public_access_block资源。

我觉得奇怪的是在plan中它确实显示了s3模块但在import中它只显示了项目中的数据

*当我创建s3时，我通过模块添加后缀-bucket，所以在导入时，我使用它在aws中创建的全名，它也具有后缀-bucket

Answer 1

在没有看到所有输出的情况下，我认为问题在于您已经在状态文件中导入了手动创建的桶到terrform创建的桶之上。(你可以自己打开并检查状态文件，看看发生了什么，它是JSON)。

为了得出这个结论，我假设my-k-frontend-test-bucket是您使用TF创建的桶，而my-k-frontend-test-2-bucket是您手动创建的桶。

要正确导入它，您需要正确索引到bucket资源列表中:

$ terraform import -var-file=test.tfvars module.s3[0].aws_s3_bucket.s3[1] my-k-frontend-test-2-bucket
^^^
Note the indexing here

aws_s3_bucket.s3资源被创建为一个列表，而不是一个单独的项目，因为您已经使用count参数指定了它。

您可能还需要将tf创建的原始bucket导入到该列表的第一个槽中。