If the version tag for the module is updated, is there a way to check the state file or some other means to see what version of the module was used?
Thank you for any help
1 reply
I'm trying to use Typescript + CDKTF in a project. I've imported the Resource"function" from cdktf. Whenever I create an S3 bucket, there appears to be a random UUID appended to the resource name (from a statefile perspective).
Any idea how to debug why? I'd like to remove that..
2 replies
Howdy, y'all.
I'm thinking through possible solutions for Terraform's "feature" of not planning and applying when it does not understand how many resources need to be created. E.g.:
The "count" value depends on resource attributes that cannot be determined until apply, so Terraform cannot predict how many instances will be created. To work around this, use the -target argument to first apply only the resources that the count depends on.
I understand the design decision probably stems from the plan portion of Terraform's functionality. It makes sense but it is very much a hurdle to teams trying to move the needle in automated infrastructure management.
I reject the idea of using -target. It only masks the issue and creates the need for human intervention or greater opportunity for human error with deployment scripts that target specific resources. There are clever workarounds to give TF the length of the list prior to knowing the details of the values, and I've used such hacks a lot. However, each time I think, "This is a waste of time." Adding to this, the issue crops up in popular TF modules on a regular basis because people do not know to test thoroughly for this problem or do not care. It eliminates my team's ability to use such modules because we must be able to stand up and tear down environments without human intervention.
In the past, I have written small Terraform wrappers that gather data for Terraform and either provide that data to Terraform as inputs or straight up generates the Terraform code for it to execute. But there's no point in maintaining a custom and private project for each employer; that's not helpful for the bottom line and its not helpful for the team(s) in the company that have the added burden of custom code to maintain.
So I have looked for public solutions, either paid or OSS, but have found nothing tenable for my employer. They should not have to learn the nuances of JSONNET or DHALL or similar, and they typically do not know TypeScript for CDKTF. Golang CDKTF thus far is not an option because it adds cognitive overhead, noise and toil with wrapping types for conversions to JSII types. CDKTF also was not GA until this week.
What is the communities solution to this Terraform shortcoming? The hacks and -target are increasingly a time sink and do not move us towards automated, reliable and self-healing deployments that may be destroyed with ease when needed.
module "app" {
source = "./imaged"
node_group = "app"
name_prefix = "app"
name_suffix = var.name_suffix
project = var.project
zone = var.zone
source_disk_name = var.source_disk_name
source_image_name = var.source_image_name
machine_type = var.app_machine_type
}
module "pgbouncer" {
source = "./imaged"
node_group = "pgbouncer"
name_prefix = "pgbouncer"
name_suffix = var.name_suffix
project = var.project
zone = var.zone
source_disk_name = var.source_disk_name
source_image_name = var.source_image_name
machine_type = var.pgbouncer_machine_type
}
module "sidekiq" {
source = "./imaged"
node_group = "sidekiq"
name_prefix = "sidekiq"
name_suffix = var.name_suffix
project = var.project
zone = var.zone
source_disk_name = var.source_disk_name
source_image_name = var.source_image_name
machine_type = var.sidekiq_machine_type
}
Hi everyone,
I need to add multiple cross zonal GCP instances to UIG. But, below code is attaching only few instances to UIG.
It is working as expected (creating UIG and attaching instances to terraform created UIG) if var.num_instances count is greater than or equal to length(data.google_compute_zones.available.names).
But , It does not attach instance in case length(data.google_compute_zones.available.names) is less than var.num_instances.
`data "google_compute_zones" "available" {
project = var.project
region = var.region
}
module "google_uig" {
source = "C:/Users/xxxx/OneDrive/Desktop/New folder/terraform-google-core/modules/compute_engine_uig"
depends_on = [
module.google_vm
]
project = var.project
count = var.num_instances
count = length(data.google_compute_zones.available.names) < var.num_instances ? length(data.google_compute_zones.available.names) : var.num_instances
zone = var.zone == null ? data.google_compute_zones.available.names[count.index % length(data.google_compute_zones.available.names)] : var.zone
name = "use1${[for zone in (data.google_compute_zones.available.names): substr(zone,-1,-1)][count.index]}-oss94-mta99-app-ig01"
instances = element((module.google_vm[*].google_instance_id), count.index)
}`
project = "gcp-test01"
region = "us-central1"
num_instances = "6"
image_project = "gcp-kcfn01"
image_name = "centos7-kent-v20220105"
instance_name = "kcfn01-cfn01-apoc01-app"
tags = ["oss01-mta-app", "usc1-sss-oss-snet01-local"]
machine_type = "n1-standard-4"
disk_size = "50"
disk_type = "pd-standard"
boot_disk_size = "50"
boot_disk_type = "pd-standard"
network = "cfn01-sss-net"
subnetwork = "cfn01-usc1-sss-oss-snet01"
subnetwork_project = "gcp-kcfn01"
`$ terraform import aws_dynamodb_table.raw-dora raw-dora-prod
Acquiring state lock. This may take a few moments...
aws_dynamodb_table.raw-dora: Importing from ID "raw-dora-prod"...
aws_dynamodb_table.raw-dora: Import prepared!
Prepared aws_dynamodb_table for import
aws_dynamodb_table.raw-dora: Refreshing state... [id=raw-dora-prod]
╷
│ Error: Cannot import non-existent remote object
│
│ While attempting to import an existing object to "aws_dynamodb_table.raw-dora", the provider detected that no object exists with the given id. Only pre-existing objects can be imported; check that the id is correct and that it
│ is associated with the provider's configured region or endpoint, or use "terraform apply" to create a new remote object for this resource.
╵
Releasing state lock. This may take a few moments...`
I am getting this error while trying to import dynamoDB table.
Here is the Code:resource "aws_dynamodb_table" "raw-dora" {
name = "raw-dora-${terraform.workspace}
This part too?
is associated with the provider's configured region or endpoint
When I apply a destroy, the file is one of the first to be destroyed. This makes destroy fails in a later step as the kubeconfig is missing.
How to make `resource "local_file" "kubeconfig" the last item to be destroyed ?
Curious about development workflows/timelines. When might this PR be merged and included in a release of the AWS provider? (Our migration to AWS-SSO has been blocked by the lack of this feature.)
dynamic at the top level to define modules however
i'm trying to allow for a little more than one eks managed node group in the eks module, so i have a data structure in a variable like this
managed_node_groups = [
{
group_name = "small"
instance_type = "t2.small"
max_size = 2
min_size = 1
desired_size = 1
},
{
group_name = "medium"
instance_type = "t2.medium"
max_size = 2
min_size = 1
desired_size = 1
}
]but in the resource it's a map, not a repeated block (so i dont think i can use a dynamic block, right ?). So what i have now is like this:
eks_managed_node_groups = {
"${var.cluster_node_groups[0]["group_name"]}" = {
subnet_ids = module.vpc.private_subnets
min_size = var.cluster_node_groups[0]["min_size"]
max_size = var.cluster_node_groups[0]["max_size"]
desired_size = var.cluster_node_groups[0]["desired_size"]
instance_types = [var.cluster_node_groups[0]["instance_type"]]
capacity_type = "ON_DEMAND"
}
}but that's obviously just for the first of the two entries. I want to get rid of the [0] and probably have some kind of index.count, or a loop ? I know i'm looking to end up with eks_managed_node_groups = { "small" = {}, "medium" = {}} in terms of its structure. Not sure how i do that in TF, though.
1 reply
Hey all,
Getting the following error when trying to apply a helm release with a bunch of manifests that exceed 4MB
The plugin returned an unexpected error from plugin.(*GRPCProvider).PlanResourceChange: rpc error: code = ResourceExhausted desc = grpc: received message larger than max (4196495 vs.
│ 4194304)Would anyone know how to overcome this? using helm provider v2.5.1
TerraformModule class. The error is Warning: Provider aws is undefined
│
│ on cdk.tf.json line 109, in module.sso.providers:
│ 109: "aws": "aws.mgmt"
│
│ Module module.sso does not declare a provider named aws. # Setup providers
mgmt_provider = AwsProvider(
self,
"aws_mgmt",
alias="mgmt",
profile="sso",
region="us-west-2",
default_tags=AwsProviderDefaultTags(
tags={
"CreatedBy": "Terraform"
}
)
)
deployment_provider = AwsProvider(
self,
"aws_deployment",
alias="deployment",
profile="sso-deployment",
region="us-west-2",
default_tags=AwsProviderDefaultTags(
tags={
"CreatedBy": "Terraform"
}
)
)
# Modules
StateModule(self, "state", providers=[deployment_provider])
SSOModule(self, "sso", providers=[mgmt_provider])
invalid function argument error from inside a module, this module is pulled in multiple times and i'm not sure which instance is giving the problem
Error: Invalid function argument
on ../../../modules012/server/alb-attachment.tf line 2, in locals:
2: hostname = lookup(var.tags, "hs:alb_hostname", false) != false ? lookup(var.tags, "hs:alb_hostname") : lookup(var.tags, "cluster")
|----------------
| var.tags is object with 6 attributes
Invalid value for "inputMap" parameter: the given object has no attribute
"cluster".
Put an end to your financial problems,now you can start earning with your PC or mobile phone, PAY INSTANTLY!
So far I’ve made over €5,500 in last 12days of doing this!(HOW) info inbox me the directly Telegram
👇👇
https://t.me/+klV1ANnp3q02YjU0
