terraform plan fails when no cluster is existent

[tillepille]

Terraform, Provider, Kubernetes and Helm Versions

Terraform version: 0.14.3
Provider version: 2.0.2
Kubernetes version: 1.19.3
Helm version: 3.5.0

Affected Resource(s)

• helm_release

Terraform Configuration Files

Provider config

provider "helm" {
  kubernetes {
    host     = module.kubernetes.host
    username = module.kubernetes.username
    password = module.kubernetes.password

    client_certificate     = base64decode(module.kubernetes.client_certificate)
    client_key             = base64decode(module.kubernetes.client_key)
    cluster_ca_certificate = base64decode(module.kubernetes.cluster_ca_certificate)
  }
}

Module which has one helm_release inside

module "externaldns" {
  source      = "../../modules/k8s-deployments/externaldns"
  namespace   = "kube-system"
  values_file = file("helm_values/externaldns.yaml")

  azure_dns_tenantId        = var.azure_dns_tenantId
  azure_dns_subscriptionId  = var.azure_dns_subscriptionId
  azure_dns_resourceGroup   = var.azure_dns_resourceGroup
  azure_dns_aadClientId     = var.azure_dns_aadClientId
  azure_dns_aadClientSecret = var.azure_dns_aadClientSecret
}

the helm_release inside this module

resource "helm_release" "exernaldns" {
  name             = "externaldns"
  repository       = "https://charts.bitnami.com/bitnami"
  chart            = "external-dns"
  version          = "3.6.0"
  create_namespace = true
  namespace        = var.namespace

  values = [var.values_file]

  set {
    name  = "azure.tenantId"
    value = var.azure_dns_tenantId
  }
  set {
    name  = "azure.subscriptionId"
    value = var.azure_dns_subscriptionId
  }
  set {
    name  = "azure.resourceGroup"
    value = var.azure_dns_resourceGroup
  }
  set {
    name  = "azure.aadClientId"
    value = var.azure_dns_aadClientId
  }
  set {
    name  = "azure.aadClientSecret"
    value = var.azure_dns_aadClientSecret
  }
}

Debug Output

NOTE: In addition to Terraform debugging, please set HELM_DEBUG=1 to enable debugging info from helm.

Debug captured with TF_LOG=trace HELM_DEBUG=1 terraform plan > debug_output.txt 2>&1

https://gist.github.com/tillepille/39e4983ba588f399f9203e2375133b6a

Panic Output

Steps to Reproduce

Create new environment, state only contains a network at that point.

1. terraform plan

Expected Behavior

plan should successfully run and produce the output that x helm releases have to be added.

Actual Behavior

terraform plan fails with

error: code = Unavailable desc = transport is closing

When there is a K8S Cluster already in place, the plan runs successfully.
When I comment out the modules (helm_releases), plan also runs successfully.

Important Factoids

References

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

[aareet]

2021/01/21 09:04:36 [TRACE] LoadSchemas: retrieving schema for provider type "registry.terraform.io/hashicorp/kubernetes-alpha"

It looks like you're using the kubernetes-alpha provider somewhere in your config? I think this may be the hidden issue. In its current state the kubernetes-alpha provider relies on server side planning, so it requires the cluster to be available at plan time. We are working to address this in hashicorp/terraform-provider-kubernetes-alpha#41, but it will take some time.

[tillepille]

@aareet You're totally right, thank you for that hint!

[ghost]

I'm going to lock this issue because it has been closed for 30 days ⏳. This helps our maintainers find and focus on the active issues.

If you feel this issue should be reopened, we encourage creating a new issue linking back to this one for added context. If you feel I made an error 🤖 🙉 , please reach out to my human friends 👉 hashibot-feedback@hashicorp.com. Thanks!