Error: Planned version is different from configured version

[anthosz]

Terraform, Provider, Kubernetes and Helm Versions

Tofu version: v1.10.3
Provider version: 3.0.2
Kubernetes version: 1.33

Affected Resource(s)

• helm_release

Terraform Configuration Files

data "gitlab_project_tags" "tags" {
  project = "project/test"
}

resource "helm_release" "test" {
  name          = "test"
  repository    = "oci://XXXXXX"
  chart         = var.project
  version       = data.gitlab_project_tags.tags.tags[0].name
  namespace     = kubernetes_namespace_v1.project.metadata[0].name
  wait_for_jobs = true
  timeout       = 60
  force_update  = true
}

Debug Output

NC

Panic Output

│ Error: Planned version is different from configured version
│ 
│   with module.test.helm_release.test,
│   on .terraform/modules/test/test.tf line 20, in resource "helm_release" "test":
│   20: resource "helm_release" "test" {
│ 
│ The version in the configuration is "" but the planned version is
│ "2025.6.1". 
│ You should update the version in your configuration to "2025.6.1", or
│ remove the version attribute from your configuration.

Steps to Reproduce

1. tofu plan

Expected Behavior

Tag 2025.6.1 is well updated in the plan

Actual Behavior

Plan fail for strange reason

Important Factoids

• if I set the version to "2025.6.1", it works.
• If I use the datasource, I have this error
• If I try an output of the datasource.tags[0].name, I have well "2025.6.1"
• If I try to force it via tostring() function, same error
• the release is not installed at this time, the plan is to install it so I don't really understand the error

References

NC

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

[anthosz]

In addition, like workaround, there is the way to use the replace_triggered_by but it uninstall/install the helm release every time even if no new tag was released.

data "gitlab_project_tags" "tags" {
  project = "project/test"
}

resource "terraform_data" "tags" {
  input = data.gitlab_project_tags.tags.tags[0].name
}

resource "helm_release" "test" {
  name            = "test"
  repository     = "oci://XXXXXX"
  chart             = var.project
  namespace    = kubernetes_namespace_v1.project.metadata[0].name
  wait_for_jobs = true
  timeout         = 60
  force_update = true
  lifecycle {
    replace_triggered_by = [terraform_data.tags]
  }
  depends_on = [data.gitlab_project_tags.tags]
}

[julianfbeck]

I have the same isssue when using a version like
~0.189

The old 2.x provider version would always update to the latest 0.189.xxx version using the correct semver syntax

[jaylonmcshan19-x]

Hi @anthosz this pr > #1713 Should fix your issue. Can you please confirm :)

[anthosz]

@jaylonmcshan19-x , thank you! Will back to you in the next days the time to test/reproduce it

[igorbrites]

I'm moving my workspace to v3 and started facing this error as well, the version comes from a data source.

╷
│ Error: Planned version is different from configured version
│ 
│   with module.prometheus.module.addon.helm_release.this[0],
│   on .terraform/modules/prometheus.addon/main.tf line 14, in resource "helm_release" "this":
│   14: resource "helm_release" "this" {
│ 
│ The version in the configuration is "" but the planned version is
│ "27.44.0". 
│ You should update the version in your configuration to "27.44.0", or remove
│ the version attribute from your configuration.
╵

How can I give this change a try? Will it become the new patch version?

EDIT:

If I pass the version directly, I start seeing this other error:

╷
│ Error: Provider produced invalid plan
│ 
│ Provider "registry.terraform.io/hashicorp/helm" planned an invalid value
│ for
│ module.prometheus.module.kube_state_metrics.helm_release.this[0].keyring:
│ planned value cty.NullVal(cty.String) does not match config value
│ cty.StringVal("").
│ 
│ This is a bug in the provider, which should be reported in the provider's
│ own issue tracker.
╵

I've tried passing it null and "", and verify is false, so no idea what I can do about it as well. My provider is on 3.1.0.

[anthosz]

Hi @anthosz this pr > #1713 Should fix your issue. Can you please confirm :)

Could you release it? I cannot test right now

[igorbrites]

@anthosz and @jaylonmcshan19-x, I just tested the new 3.1.1 version, and that error is gone for me. I'll test it out with more charts and post my findings here. Thanks @jaylonmcshan19-x for the fix!

[igorbrites]

OK, never mind, still seeing those errors. Tried to even remove the version, keyring, and verify fields (and passing their respective default values) from the helm_releases, but no-go:

╷
│ Error: Provider produced invalid plan
│ 
│ Provider "registry.terraform.io/hashicorp/helm" planned an invalid value
│ for module.vault_config_operator.module.addon.helm_release.this[0].keyring:
│ planned value cty.NullVal(cty.String) does not match config value
│ cty.StringVal("").
│ 
│ This is a bug in the provider, which should be reported in the provider's
│ own issue tracker.
╵
╷
│ Error: Planned version is different from configured version
│ 
│   with module.openwebui.module.addon.helm_release.this[0],
│   on .terraform/modules/openwebui.addon/main.tf line 14, in resource "helm_release" "this":
│   14: resource "helm_release" "this" {
│ 
│ The version in the configuration is "" but the planned version is "8.14.0".
│ You should update the version in your configuration to "8.14.0", or remove
│ the version attribute from your configuration.
╵