Terraform (0.12.29) 导入未按预期工作；导入成功，但计划显示破坏和重新创建 [英] Terraform (0.12.29) import not working as expected; import succeeded but plan shows destroy & recreate
问题描述
一些背景:我们有 terraform 代码来创建各种 AWS 资源.其中一些资源是按 AWS 账户创建的，因此被构造为存储在我们项目的 account-scope 文件夹中.当时我们只有一个 AWS 区域.现在我们的应用程序是多区域的，因此将为每个 AWS 账户的每个区域创建这些资源.
Some Background:
We have terraform code to create various AWS resources. Some of these resources are created per AWS account and hence are structured to be stored in a account-scope folder in our project. This was when we were only having one AWS region. Now our application is made multi-region and hence these resources are to be created per region for each AWS account.
为了做到这一点，我们现在已将这些 TF 脚本移至 region-scope 文件夹，该文件夹将按区域运行.由于这些资源不再属于帐户范围"，我们已将它们从帐户范围Terraform 状态中删除.现在当我尝试导入这些资源时
In order to do that we have now moved these TF scripts to region-scope folder which will be run per region. Since these resources are no longer part of 'account scope' we have removed them from the account scope Terraform state.
Now when I try to import these resources
通过从 xyz-region-scope 目录运行此导入资源:
Imported the resources by running this from xyz-region-scope directory:
terraform import -var-file=config/us-west-2/default.tfvars -var-file=variables.tfvars -var-file=../globals.tfvars -var profile=xyz-stage -var region=us-west-2 -var tfstate_bucket=ab-xyz-stage-tfstate-5b8873b8 -no-color <RESOURCE_NAME> <RESOURCE_ID>
资源的一个例子是:
RESOURCE_NAME=module.buckets.aws_s3_bucket.cloudtrail_logging_bucket
RESOURCE_ID="ab-xyz-stage-cloudtrail-logging-72a2c5cd"
我希望导入会更新本地计算机上 terraform 状态文件中的资源，但在 xyz-region-scope/state/xyz-stage/terraform.tfstate 下创建的 terraform 状态文件没有更新.
I was expecting the imports would update the resources in the terraform state file on my local machine but the terraform state file created under xyz-region-scope/state/xyz-stage/terraform.tfstate is not updated.
通过以下方式验证导入:
terraform show
运行 terraform 计划:
terraform plan -var-file=config/us-west-2/default.tfvars -var-file=variables.tfvars -var-file=../globals.tfvars -var profile=xyz-stage -var region=us-west-2 -var tfstate_bucket=ab-xyz-stage-tfstate-5b8873b8 -no-color
但是 terraform plan 输出显示 Plan: 6 to add, 0 to change, 5 to destroy. 那就是那些资源将被销毁并重新创建.
But the terraform plan output shows Plan: 6 to add, 0 to change, 5 to destroy. that is those resources will be destroyed and recreated.
我不清楚为什么会这样，是我遗漏了什么并且没有做对吗?
I am not clear why so, am I missing something and not doing it right?
请注意，我们将远程状态存储在 S3 存储桶中但我目前没有在 S3 存储桶中为 区域范围 创建远程 TF 状态文件(我有一个用于 帐户范围).我期待 Import..Plan..Apply 过程也会为区域范围创建一个.
Please note we store the remote state in S3 bucket but I currently do not have the remote TF state file created in S3 bucket for region scope (I do have one for account scope though). I was expecting that the Import..Plan..Apply process will create one for region scope as well.
在运行导入后，我看到在 S3 中为区域范围创建的远程 TF 状态文件.我看到的这个新的区域范围 tf 状态文件与旧帐户范围一之间的一个区别是:新文件在任何资源 resources 下都没有任何 depends_on" 块[] >实例[]
I see the remote TF state file created in the S3 for region scope after running imports. One difference that I see between this new region-scope tf state file from old account-scope one is: the new file does not have any "depends_on" block under any of the resources resources[] > instances[]
环境:
Local machine: macOS v10.14.6
Terraform v0.12.29
+ provider.aws v3.14.1
+ provider.null v2.1.2
+ provider.random v2.3.1
+ provider.template v2.1.2
编辑 2:
这是我的导入和 terraform 计划:
terraform import module.buckets.random_id.cloudtrail_bucket_suffix cqLFzQ
terraform import module.buckets.aws_s3_bucket.cloudtrail_logging_bucket "ab-xyz-stage-cloudtrail-logging-72a2c5cd"
terraform import module.buckets.aws_s3_bucket_policy.cloudtrail_logging_bucket "ab-xyz-stage-cloudtrail-logging-72a2c5cd"
terraform import module.buckets.module.access_logging_bucket.aws_s3_bucket.default "ab-xyz-stage-access-logging-9d8e94ff"
terraform import module.buckets.module.access_logging_bucket.random_id.bucket_suffix nY6U_w
terraform import module.encryption.module.data_key.aws_iam_policy.decrypt "arn:aws:iam::123412341234:policy/ab_data_key_xyz_stage_decrypt"
terraform import module.encryption.module.data_key.aws_iam_policy.encrypt "arn:aws:iam::123412341234:policy/ab_data_key_xyz_stage_encrypt"
mymachine:xyz-region-scope kuldeepjain$ ../scripts/terraform.sh xyz-stage plan -no-color
+ set -o posix
+ IFS='
'
++ blhome
+ BASH_LIB_HOME=/usr/local/lib/mycompany/ab/bash_library/0.0.1-SNAPSHOT
+ source /usr/local/lib/mycompany/ab/bash_library/0.0.1-SNAPSHOT/s3/bucket.sh
+ main xyz-stage plan -no-color
+ '[' 3 -lt 2 ']'
+ local env=xyz-stage
+ shift
+ local command=plan
+ shift
++ get_region xyz-stage
++ local env=xyz-stage
++ shift
+++ aws --profile xyz-stage configure get region
++ local region=us-west-2
++ '[' -z us-west-2 ']'
++ echo us-west-2
+ local region=us-west-2
++ _get_bucket xyz-stage xyz-stage-tfstate
++ local env=xyz-stage
++ shift
++ local name=xyz-stage-tfstate
++ shift
+++ _get_bucket_list xyz-stage xyz-stage-tfstate
+++ local env=xyz-stage
+++ shift
+++ local name=xyz-stage-tfstate
+++ shift
+++ aws --profile xyz-stage --output json s3api list-buckets --query 'Buckets[?contains(Name, `xyz-stage-tfstate`) == `true`].Name'
++ local 'bucket_list=[
"ab-xyz-stage-tfstate-5b8873b8"
]'
+++ _count_buckets_in_json '[
"ab-xyz-stage-tfstate-5b8873b8"
]'
+++ local 'json=[
"ab-xyz-stage-tfstate-5b8873b8"
]'
+++ shift
+++ echo '[
"ab-xyz-stage-tfstate-5b8873b8"
]'
+++ jq '. | length'
++ local number_of_buckets=1
++ '[' 1 == 0 ']'
++ '[' 1 -gt 1 ']'
+++ echo '[
"ab-xyz-stage-tfstate-5b8873b8"
]'
+++ jq -r '.[0]'
++ local bucket_name=ab-xyz-stage-tfstate-5b8873b8
++ echo ab-xyz-stage-tfstate-5b8873b8
+ local tfstate_bucket=ab-xyz-stage-tfstate-5b8873b8
++ get_config_file xyz-stage us-west-2
++ local env=xyz-stage
++ shift
++ local region=us-west-2
++ shift
++ local config_file=config/us-west-2/xyz-stage.tfvars
++ '[' '!' -f config/us-west-2/xyz-stage.tfvars ']'
++ config_file=config/us-west-2/default.tfvars
++ echo config/us-west-2/default.tfvars
+ local config_file=config/us-west-2/default.tfvars
+ export TF_DATA_DIR=state/xyz-stage/
+ TF_DATA_DIR=state/xyz-stage/
+ terraform get
+ terraform plan -var-file=config/us-west-2/default.tfvars -var-file=variables.tfvars -var-file=../globals.tfvars -var profile=xyz-stage -var region=us-west-2 -var tfstate_bucket=ab-xyz-stage-tfstate-5b8873b8 -no-color
Refreshing Terraform state in-memory prior to plan...
The refreshed state will be used to calculate this plan, but will not be
persisted to local or remote state storage.
module.encryption.module.data_key.data.null_data_source.key: Refreshing state...
module.buckets.data.template_file.dependencies: Refreshing state...
module.buckets.module.access_logging_bucket.data.template_file.dependencies: Refreshing state...
module.encryption.module.data_key.data.aws_region.current: Refreshing state...
module.buckets.module.access_logging_bucket.data.aws_caller_identity.current: Refreshing state...
data.aws_caller_identity.current: Refreshing state...
module.buckets.module.access_logging_bucket.data.aws_kms_alias.encryption_key_alias: Refreshing state...
module.buckets.data.aws_caller_identity.current: Refreshing state...
module.encryption.module.data_key.data.aws_caller_identity.current: Refreshing state...
module.encryption.module.data_key.data.aws_kms_alias.default: Refreshing state...
module.buckets.module.access_logging_bucket.data.template_file.encryption_configuration: Refreshing state...
module.encryption.module.data_key.data.aws_iam_policy_document.decrypt: Refreshing state...
module.encryption.module.data_key.data.aws_iam_policy_document.encrypt: Refreshing state...
module.buckets.module.access_logging_bucket.random_id.bucket_suffix: Refreshing state... [id=nY6U_w]
module.encryption.module.data_key.aws_iam_policy.decrypt: Refreshing state... [id=arn:aws:iam::123412341234:policy/ab_data_key_xyz_stage_decrypt]
module.encryption.module.data_key.aws_iam_policy.encrypt: Refreshing state... [id=arn:aws:iam::123412341234:policy/ab_data_key_xyz_stage_encrypt]
module.buckets.module.access_logging_bucket.aws_s3_bucket.default: Refreshing state... [id=ab-xyz-stage-access-logging-9d8e94ff]
module.buckets.random_id.cloudtrail_bucket_suffix: Refreshing state... [id=cqLFzQ]
module.buckets.aws_s3_bucket.cloudtrail_logging_bucket: Refreshing state... [id=ab-xyz-stage-cloudtrail-logging-72a2c5cd]
module.buckets.data.aws_iam_policy_document.restrict_access_cloudtrail: Refreshing state...
module.buckets.aws_s3_bucket_policy.cloudtrail_logging_bucket: Refreshing state... [id=ab-xyz-stage-cloudtrail-logging-72a2c5cd]
------------------------------------------------------------------------
An execution plan has been generated and is shown below.
Resource actions are indicated with the following symbols:
+ create
-/+ destroy and then create replacement
<= read (data resources)
Terraform will perform the following actions:
# module.buckets.data.aws_iam_policy_document.restrict_access_cloudtrail will be read during apply
# (config refers to values not yet known)
<= data "aws_iam_policy_document" "restrict_access_cloudtrail" {
+ id = (known after apply)
+ json = (known after apply)
+ statement {
+ actions = [
+ "s3:GetBucketAcl",
]
+ effect = "Allow"
+ resources = [
+ (known after apply),
]
+ sid = "AWSCloudTrailAclCheck"
+ principals {
+ identifiers = [
+ "cloudtrail.amazonaws.com",
]
+ type = "Service"
}
}
+ statement {
+ actions = [
+ "s3:PutObject",
]
+ effect = "Allow"
+ resources = [
+ (known after apply),
]
+ sid = "AWSCloudTrailWrite"
+ condition {
+ test = "StringEquals"
+ values = [
+ "bucket-owner-full-control",
]
+ variable = "s3:x-amz-acl"
}
+ principals {
+ identifiers = [
+ "cloudtrail.amazonaws.com",
]
+ type = "Service"
}
}
}
# module.buckets.aws_s3_bucket.cloudtrail_logging_bucket must be replaced
-/+ resource "aws_s3_bucket" "cloudtrail_logging_bucket" {
+ acceleration_status = (known after apply)
+ acl = "private"
~ arn = "arn:aws:s3:::ab-xyz-stage-cloudtrail-logging-72a2c5cd" -> (known after apply)
~ bucket = "ab-xyz-stage-cloudtrail-logging-72a2c5cd" -> (known after apply) # forces replacement
~ bucket_domain_name = "ab-xyz-stage-cloudtrail-logging-72a2c5cd.s3.amazonaws.com" -> (known after apply)
~ bucket_regional_domain_name = "ab-xyz-stage-cloudtrail-logging-72a2c5cd.s3.us-west-2.amazonaws.com" -> (known after apply)
+ force_destroy = false
~ hosted_zone_id = "Z3BJ6K6RIION7M" -> (known after apply)
~ id = "ab-xyz-stage-cloudtrail-logging-72a2c5cd" -> (known after apply)
~ region = "us-west-2" -> (known after apply)
~ request_payer = "BucketOwner" -> (known after apply)
tags = {
"mycompany:finance:accountenvironment" = "xyz-stage"
"mycompany:finance:application" = "ab-platform"
"mycompany:finance:billablebusinessunit" = "my-dev"
"name" = "Cloudtrail logging bucket"
}
+ website_domain = (known after apply)
+ website_endpoint = (known after apply)
~ lifecycle_rule {
- abort_incomplete_multipart_upload_days = 0 -> null
enabled = true
~ id = "intu-lifecycle-s3-int-tier" -> (known after apply)
- tags = {} -> null
transition {
days = 32
storage_class = "INTELLIGENT_TIERING"
}
}
- logging {
- target_bucket = "ab-xyz-stage-access-logging-9d8e94ff" -> null
- target_prefix = "logs/cloudtrail-logging/" -> null
}
+ logging {
+ target_bucket = (known after apply)
+ target_prefix = "logs/cloudtrail-logging/"
}
~ versioning {
~ enabled = false -> (known after apply)
~ mfa_delete = false -> (known after apply)
}
}
# module.buckets.aws_s3_bucket_policy.cloudtrail_logging_bucket must be replaced
-/+ resource "aws_s3_bucket_policy" "cloudtrail_logging_bucket" {
~ bucket = "ab-xyz-stage-cloudtrail-logging-72a2c5cd" -> (known after apply) # forces replacement
~ id = "ab-xyz-stage-cloudtrail-logging-72a2c5cd" -> (known after apply)
~ policy = jsonencode(
{
- Statement = [
- {
- Action = "s3:GetBucketAcl"
- Effect = "Allow"
- Principal = {
- Service = "cloudtrail.amazonaws.com"
}
- Resource = "arn:aws:s3:::ab-xyz-stage-cloudtrail-logging-72a2c5cd"
- Sid = "AWSCloudTrailAclCheck"
},
- {
- Action = "s3:PutObject"
- Condition = {
- StringEquals = {
- s3:x-amz-acl = "bucket-owner-full-control"
}
}
- Effect = "Allow"
- Principal = {
- Service = "cloudtrail.amazonaws.com"
}
- Resource = "arn:aws:s3:::ab-xyz-stage-cloudtrail-logging-72a2c5cd/*"
- Sid = "AWSCloudTrailWrite"
},
]
- Version = "2012-10-17"
}
) -> (known after apply)
}
# module.buckets.random_id.cloudtrail_bucket_suffix must be replaced
-/+ resource "random_id" "cloudtrail_bucket_suffix" {
~ b64 = "cqLFzQ" -> (known after apply)
~ b64_std = "cqLFzQ==" -> (known after apply)
~ b64_url = "cqLFzQ" -> (known after apply)
byte_length = 4
~ dec = "1923270093" -> (known after apply)
~ hex = "72a2c5cd" -> (known after apply)
~ id = "cqLFzQ" -> (known after apply)
+ keepers = {
+ "aws_account_id" = "123412341234"
+ "env" = "xyz-stage"
} # forces replacement
}
# module.buckets.module.access_logging_bucket.aws_s3_bucket.default must be replaced
-/+ resource "aws_s3_bucket" "default" {
+ acceleration_status = (known after apply)
+ acl = "log-delivery-write"
~ arn = "arn:aws:s3:::ab-xyz-stage-access-logging-9d8e94ff" -> (known after apply)
~ bucket = "ab-xyz-stage-access-logging-9d8e94ff" -> (known after apply) # forces replacement
~ bucket_domain_name = "ab-xyz-stage-access-logging-9d8e94ff.s3.amazonaws.com" -> (known after apply)
~ bucket_regional_domain_name = "ab-xyz-stage-access-logging-9d8e94ff.s3.us-west-2.amazonaws.com" -> (known after apply)
+ force_destroy = false
~ hosted_zone_id = "Z3BJ6K6RIION7M" -> (known after apply)
~ id = "ab-xyz-stage-access-logging-9d8e94ff" -> (known after apply)
~ region = "us-west-2" -> (known after apply)
~ request_payer = "BucketOwner" -> (known after apply)
tags = {
"mycompany:finance:accountenvironment" = "xyz-stage"
"mycompany:finance:application" = "ab-platform"
"mycompany:finance:billablebusinessunit" = "my-dev"
"name" = "Access logging bucket"
}
+ website_domain = (known after apply)
+ website_endpoint = (known after apply)
- grant {
- permissions = [
- "READ_ACP",
- "WRITE",
] -> null
- type = "Group" -> null
- uri = "http://acs.amazonaws.com/groups/s3/LogDelivery" -> null
}
- grant {
- id = "0343271a8c2f184152c171b223945b22ceaf5be5c9b78cf167660600747b5ad8" -> null
- permissions = [
- "FULL_CONTROL",
] -> null
- type = "CanonicalUser" -> null
}
- lifecycle_rule {
- abort_incomplete_multipart_upload_days = 0 -> null
- enabled = true -> null
- id = "intu-lifecycle-s3-int-tier" -> null
- tags = {} -> null
- transition {
- days = 32 -> null
- storage_class = "INTELLIGENT_TIERING" -> null
}
}
+ logging {
+ target_bucket = (known after apply)
+ target_prefix = "logs/access-logging/"
}
~ versioning {
~ enabled = false -> (known after apply)
~ mfa_delete = false -> (known after apply)
}
}
# module.buckets.module.access_logging_bucket.random_id.bucket_suffix must be replaced
-/+ resource "random_id" "bucket_suffix" {
~ b64 = "nY6U_w" -> (known after apply)
~ b64_std = "nY6U/w==" -> (known after apply)
~ b64_url = "nY6U_w" -> (known after apply)
byte_length = 4
~ dec = "2643367167" -> (known after apply)
~ hex = "9d8e94ff" -> (known after apply)
~ id = "nY6U_w" -> (known after apply)
+ keepers = {
+ "aws_account_id" = "123412341234"
+ "env" = "xyz-stage"
} # forces replacement
}
Plan: 6 to add, 0 to change, 5 to destroy.
cloudtrail_bucket_suffix 的当前远程 TF 状态(LEFT)与旧帐户范围(RIGHT)的差异片段:
Snippet of Diff of my current remote TF state(LEFT) vs old account-scope(RIGHT) for cloudtrail_bucket_suffix:
推荐答案
方案显示bucket名称不同(bucket强制替换).
The plan shows a difference in the name of the bucket (bucket forces replacement).
这会触发存储桶本身和相关资源的重新创建.
This triggers recreation of the bucket itself and dependent resources.
您需要将存储桶名称设为稳定状态，然后其余部分也将稳定.由于您对存储桶名称使用随机后缀，我怀疑您忘记导入它.random_id 资源允许这样的导入:
You need to get the bucket name to a stable state, then the rest will be stable as well. As you are using a random suffix for the bucket name, I suspect you forget to import this. The random_id resource allows imports like this:
terraform import module.buckets.random_id.cloudtrail_bucket_suffix cqLFzQ
但是，您需要删除 keepers，因为它们会触发 random_id 资源的替换.keepers 用于在其他资源发生变化时触发依赖资源的重新创建.
However, you will need to remove the keepers as they trigger the replacement of the random_id resource. keepers are used to trigger the recreation of dependent resources whenever other resources change.
我认为这不是您想要的存储桶，因为您定义的保持器似乎是稳定/静态的:account_id 和 env 都不太可能因此而改变部署.如果你真的需要它们，你可以尝试手动操作状态.
I assume this is not what you want for your buckets as the keepers you defined seem to be stable/static: account_id and env are both unlikely to change for this deployment. If you really need them you can try to manipulate the state either manually.
这篇关于Terraform (0.12.29) 导入未按预期工作；导入成功，但计划显示破坏和重新创建的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
