upgrade_mode fails if the cluster is being built

Terraform, Provider, Kubernetes and Helm Versions

Terraform version: 1.5.3
Provider version: 2.15.0
Kubernetes version: 1.30.4

Affected Resource(s)

helm_release

Terraform Configuration Files

variable "region" {
  default = "us-east1"
}

provider "google" {
  region = var.region
}

resource "google_container_cluster" "default" {
  name               = "helm-test-cluster"
  location           = var.region
  initial_node_count = 1
  project            = "my-test-project"

  master_auth {
    client_certificate_config {
      issue_client_certificate = false
    }
  }

  node_config {
    oauth_scopes = [
      "https://www.googleapis.com/auth/logging.write",
      "https://www.googleapis.com/auth/monitoring",
    ]
    metadata = {
      disable-legacy-endpoints = "true"
    }
  }
}

data "google_client_config" "provider" {}

provider "helm" {
  kubernetes {
    host                   = "https://${google_container_cluster.default.endpoint}"
    token                  = data.google_client_config.provider.access_token
    cluster_ca_certificate = base64decode(google_container_cluster.default.master_auth[0].cluster_ca_certificate)
  }
}

resource "helm_release" "redis" {
  name            = "redis"
  repository      = "https://kubernetes-charts.storage.googleapis.com/"
  chart           = "redis"
  atomic          = true
  wait            = true
  namespace       = "default"
  upgrade_install = true
}

Debug Output

https://gist.github.com/n-oden/033ea29f1156c779df95556197b23287

Steps to Reproduce

terraform plan

Expected Behavior

A successful terraform plan. Since the cluster has not yet been created, the helm_resource release should be planned for creation.

Actual Behavior

Because the upgrade_install logic attempts to create a helm history client in order to determine the state of the release, it errors out trying to query a nonexistent cluster:

│ Error: Kubernetes cluster unreachable: invalid configuration: no configuration has been provided, try setting KUBERNETES_MASTER environment variable
│
│   with helm_release.redis,
│   on main.tf line 46, in resource "helm_release" "redis":
│   46: resource "helm_release" "redis" {

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

hashicorp / terraform-provider-helm

upgrade_mode fails if the cluster is being built #1495