Missing darwin_arm64 binary for the hashicorp/template provider (Apple Silicon, M1)

[larstobi]

If I use the darwin_arm64 binary of terraform, then I cannot run terraform init if I use the template_file provider.

(A quick terraform -version will reveal if you run amd64 on Rosetta or on arm64.)

The releases show that the darwin_arm64 is in fact not published:

• https://releases.hashicorp.com/terraform-provider-template/2.2.0/

I have checked out the source code for this and built it on my M1 macbook, and it compiles and runs without issue, so it should be a quick fix for Hashicorp to publish this binary for the same 2.2.0 version.

[jbardin]

Hi @larstobi,

The template provider was deprecated before darwin_arm64 providers were released, and the repository is now archived and read-only. The functionality from this provider is now handled by the templatefile function in Terraform.

Thanks!

[larstobi]

@jbardin
Thanks. I was able to migrate most of our uses of template_file to templatefile(), however some uses are not possible to migrate, like when using the output of a resource as an input to template_file:

data "aws_s3_bucket_object" "example" {
  bucket = "example-bucket-name"
  key    = "example.yaml"
}

data "template_file" "example" {
  template = data.aws_s3_bucket_object.example.body
  vars = {
    foo  = "bar"
  }
}

Since the templatefile() function doesn't support string input, but requires a file name, it can't be converted. Is this use case supported by a different function?

If I can't make it work by other means, I can see there is a fork of your old template provider repo that has published a binary for darwin_arm64 here: https://github.com/gxben/terraform-provider-template/tree/v2.2.0-m1

[larstobi]

#26838

It seems that the only way forward is to use a self-compiled darwin_arm64 provider or a the third party one from gxben, as the above use case is impossible to achieve without using a deprecated provider running on terraform with Intel emulation.

[github-actions]

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 have found a problem that seems similar to this, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.