kubernetes_manifest - Error: Failed to determine GroupVersionResource for manifest - cannot select exact GV from REST mapper

[MrCoffey]

terraform --version
Terraform v1.2.0
on darwin_amd64
+ provider registry.terraform.io/hashicorp/aws v3.45.0
+ provider registry.terraform.io/hashicorp/kubernetes v2.3.2
+ provider registry.terraform.io/hashicorp/kubernetes-alpha v0.5.0
+ provider registry.terraform.io/hashicorp/null v3.1.0

Affected Resource(s)

kubernetes_manifest

Terraform Configuration Files

resource "kubernetes_namespace" "mwaa" {
  metadata {
    annotations = {
      name = "mwaa"
    }
    name = "mwaa"
  }
}

resource "kubernetes_manifest" "mwaa_role" {

  manifest = {
    "apiVersion" = "rbac.authorization.k8s.io/v1"
    "kind"       = "Role"
    "metadata" = {
      "name"      = "mwaa-role"
      "namespace" = "mwaa"
    }
    "rules" = [
      {
        "apiGroups" = [
          "",
          "apps",
          "batch",
          "extensions",
        ]
        "resources" = [
          "jobs",
          "pods",
          "pods/attach",
          "pods/exec",
          "pods/log",
          "pods/portforward",
          "secrets",
          "services",
        ]
        "verbs" = [
          "create",
          "delete",
          "describe",
          "get",
          "list",
          "patch",
          "update",
        ]
      },
    ]
  }

  depends_on = [
    kubernetes_namespace.mwaa
  ]
}

Debug Output

TF_LOG=TRACE

[ERROR] vertex "kubernetes_manifest.mwaa_role_binding" error: Failed to determine GroupVersionResource for manifest
[TRACE] vertex "kubernetes_manifest.mwaa_role_binding": visit complete, with errors
[DEBUG] provider.terraform-provider-kubernetes-alpha_v0.5.0_x5:     "get"
[ERROR] vertex "kubernetes_manifest.mwaa_role (expand)" error: Failed to determine GroupVersionResource for manifest
[TRACE] vertex "kubernetes_manifest.mwaa_role_binding (expand)": dynamic subgraph encountered errors: Failed to determine GroupVersionResource for manifest
[TRACE] vertex "kubernetes_manifest.mwaa_role (expand)": visit complete, with errors
[ERROR] vertex "kubernetes_manifest.mwaa_role_binding (expand)" error: Failed to determine GroupVersionResource for manifest
[DEBUG] provider.terraform-provider-kubernetes-alpha_v0.5.0_x5:    ]
[TRACE] vertex "kubernetes_manifest.mwaa_role_binding (expand)": visit complete, with errors

Panic Output

Steps to Reproduce

1. terraform apply -->

Expected Behavior

Kubernetes object should be created.

Actual Behavior

Provider fails with the following error:

│ Error: Failed to determine GroupVersionResource for manifest
│ 
│   with kubernetes_manifest.mwaa_role_binding,
│   on main.tf line 77, in resource "kubernetes_manifest" "mwaa_role_binding":
│   77: resource "kubernetes_manifest" "mwaa_role_binding" {
│ 
│ cannot select exact GV from REST mapper

Api versions are available.

kubectl get apiservices  | grep authorization

v1.authorization.k8s.io                Local     True        2d2h
v1.rbac.authorization.k8s.io           Local     True        2d2h

Important Factoids

References

• kubernetes_manifest: Error: Failed to determine GroupVersionResource for manifest #1583

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

[BBBmau]

Hello! Thank you for opening this issue, could you explain why k8s alpha is present in the provider list? Could you also attempt this again with the latest version of the k8s provider.

[MrCoffey]

Hi, thanks for taking a look.

At that time we were using the two Kubernetes providers it looked like a duplication so we removed and used only kubernetes provider v2.3.2 instead of kubernetes-alpha.

Neither of them worked in the end, even after deleting the cache. We decided to replace resource "kubernetes_manifest" with "resource "kubectl_manifest" to resolve our issue.

[arybolovlev]

Hi @MrCoffey,

v2.3.2 is quite old. Please try the latest version v2.16.0.

Thanks.

[samiamoura]

I @arybolovlev I facing the same issue with the v2.16.0 and the Traefik application
The problem is when the CRD doesn't exist in the kubernetes cluster, the terraform kubernetes terraform provider fails beacause it try to read the relevant CRD before creates the CR despite the depends_on field.
The correct behavior should be, first create the helm traefik release (and appropriate CRD) and then read and create the CR

[github-actions]

Marking this issue as stale due to inactivity. If this issue receives no comments in the next 30 days it will automatically be closed. If this issue was automatically closed and you feel this issue should be reopened, we encourage creating a new issue linking back to this one for added context. This helps our maintainers find and focus on the active issues. Maintainers may also remove the stale label at their discretion. Thank you!

[Monachawla1712]

Hi, thanks for taking a look.

At that time we were using the two Kubernetes providers it looked like a duplication so we removed and used only kubernetes provider v2.3.2 instead of kubernetes-alpha.

Neither of them worked in the end, even after deleting the cache. We decided to replace resource "kubernetes_manifest" with "resource "kubectl_manifest" to resolve our issue.

i want to know if you used kubectl_manifest resource, it properly worked or not ?