Terraform-aws-eks: module error when using loops to define node_groups

[swapzero]

I have issues

module fails with generated node_groups variable that depends on another module outputs(vpc)

I'm submitting a...

• [x] bug report
• [ ] feature request
• [ ] support request - read the FAQ first!
• [ ] kudos, thank you, warm fuzzy

What is the current behavior?

I am using for loops to define the worker_groups based on vpc module outputs. While this works, it fails if I try to apply the same logic for node_groups.

If this is a bug, how to reproduce? Please include a code sample if relevant.

• clean run
• no need to upgrade from versions prior to 8.0.0

node_groups variable:

module "eks" {
...
  node_groups = merge(
          {
            for idx in range(length(module.vpc.private_subnets)) :
              "node-${module.vpc.azs[idx]}-${idx}" => {
                min_capacity     = 1
                desired_capacity = 3
                max_capacity     = 5
                subnets = ["${module.vpc.private_subnets[idx]}"]
              }
          }, {}
          )
...

$ terraform plan
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.eks.data.aws_caller_identity.current: Refreshing state...
module.eks.data.aws_iam_policy_document.cluster_assume_role_policy: Refreshing state...
module.eks.data.aws_iam_policy_document.workers_assume_role_policy: Refreshing state...
module.eks.data.aws_ami.eks_worker: Refreshing state...
module.eks.data.aws_ami.eks_worker_windows: Refreshing state...

------------------------------------------------------------------------

Error: Invalid count argument

  on .terraform/modules/eks/terraform-aws-modules-terraform-aws-eks-a9db852/aws_auth.tf line 46, in data "template_file" "node_group_arns":
  46:   count    = var.create_eks ? length(module.node_groups.aws_auth_roles) : 0

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.



md5-8977ae78cf5821a222e632c61b9e2a63



$ terraform version
Terraform v0.12.19
+ provider.aws v2.44.0
+ provider.kubernetes v1.10.0
+ provider.local v1.4.0
+ provider.null v2.1.2
+ provider.random v2.2.1
+ provider.template v2.1.2

Any other relevant info

[dpiddockcmp]

The node_groups module makes heavy use of for_each. Thus the keys given in the node_groups map must be known to Terraform at plan time.

Drop the merge and your sample plan above should apply. I was able to do a terraform plan.

You will likely have to do some gymnastics if you want to do more than this single for loop.

[swapzero]

I've done some "gymnastics" to be able to solve my issue (configure multiple node groups, each scoped to a single AZ), here is an ugly version, maybe it could help someone:

  node_groups = {
    for ng in concat(
        [ for idx in range(length(module.vpc.public_subnets)):
          {
            subnets = [module.vpc.public_subnets[idx]]
            name = "rick-${module.vpc.azs[idx]}"
            min_capacity     = 1
            desired_capacity = 3
            max_capacity     = 5
            k8s_labels = {
               "nodelabel" = "rick"
            }
          }
        ],
        [ for idx in range(length(module.vpc.public_subnets)):
          {
            subnet = module.vpc.public_subnets[idx]
            name = "morty-${module.vpc.azs[idx]}"
            desired_capacity = 2
            k8s_labels = {
               "nodelabel" = "morty"
            }
          }
        ],
    ):
    "${ng.name}" => ng
  }

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[stale[bot]]

This issue has been automatically closed because it has not had recent activity since being marked as stale.