3.1.0: Upgrade from 3.0.2 caused all helm charts to be redeployed (updated) even thought there were no changes

[MichalKoziorowski-TomTom]

Terraform, Provider, Kubernetes and Helm Versions

Terraform version: OpenTofu v1.9.0
Provider version: 3.1.0 upgraded from 3.0.2
Kubernetes version: 1.32.x

Affected Resource(s)

• helm_release

Terraform Configuration Files

We are using helm_release with ignore_changes = all lifecycle

  # Do not update this release after it is deployed
  #
  # This fixes issues where the deployment has scaled out and by updating
  # this, without making any real changes like in the case of a new (updated) stack
  # is being deployed, the number of replicas are reverted to the original value.
  lifecycle {
    ignore_changes = all
  }

# Copy-paste your Terraform configurations here - for large Terraform configs,
# please use a service like Dropbox and share a link to the ZIP file. For
# security, you can also encrypt the files using our GPG public key.

Steps to Reproduce

1. Deploy helm chart using 3.0.2 provider
2. Update provider to 3.1.0
3. Don't make any configuration change to helm values, just terraform apply.

Expected Behavior

Apply should report no changes and do nothing

Actual Behavior

Helm deployment was updated. I expect that it might be related to introduction of new take_ownership field. We are using default value for it.

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

[air3ijai]

Same story

+       take_ownership             = false
+       upgrade_install            = false
#         (28 unchanged attributes hidden)

These fields were added in the release v3.1.0.

[AdamS3x]

We have the same issue, had to pin the provider to v3.0.2 to prevent this behaviour