Amazon connect instance ID now outputting ARN

[jfarchergamma]

Terraform and AWS Provider Version

Provider version >5.2.0
aws_connect_instance.connect.id outputs just the ID
But with newer versions it outputs the ARN which is now breaking my TF code references where the ID is used/required.

Affected Resource(s) or Data Source(s)

No response

Expected Behavior

.id should return the ID not the ARN

Actual Behavior

Getting ARN with both .arn and .id

Sample Terraform Configuration

when using any of the instance storage resource types that require the ID as an input we get an error. e.g.

resource "aws_connect_instance_storage_config" "CTR" {
  instance_id   = aws_connect_instance.connect.id
  resource_type = "CONTACT_TRACE_RECORDS"

  storage_config {
    kinesis_stream_config {
      stream_arn = aws_kinesis_stream.contact_trace_record.arn
    }
    storage_type = "KINESIS_STREAM"
  }
}

this results in the error:

Error: reading Connect Instance Storage Config (ARN REDACTED:CONTACT_TRACE_RECORDS): operation error Connect: DescribeInstanceStorageConfig, https response error StatusCode: 400, RequestID: , InvalidRequestException: Invalid input for field InstanceId

instance_id = regex("instance/(.*)", aws_connect_instance.connect.id)[0] will help, but less than ideall

Steps to Reproduce

Run plan/apply with my code and it now wants to change all references of an ID to the ARN.

[github-actions]

Community Guidelines

This comment is added to every new Issue to provide quick reference to how the Terraform AWS Provider is maintained. Please review the information below, and thank you for contributing to the community that keeps the provider thriving! 🚀

Voting for Prioritization

• Please vote on this Issue by adding a 👍 reaction to the original post to help the community and maintainers prioritize it.
• Please see our prioritization guide for additional information on how the maintainers handle prioritization.
• Please do not leave +1 or other comments that do not add relevant new information or questions; they generate extra noise for others following the Issue and do not help prioritize the request.

Volunteering to Work on This Issue

• If you are interested in working on this issue, please leave a comment.
• If this would be your first contribution, please review the contribution guide.
• For new resources and data sources, use skaff to generate scaffolding with comments detailing common expectations.

[justinretzolk]

Hey @jfarchergamma 👋 Thank you for taking the time to raise this! I took a look over the resource definition and noted that the ID is set to the id output of the API call, which uses CreateInstance. Based on that, I would expect that the .id output of the resource would still be the ID, but it's possible that something has changed on the AWS side. Are you able to provide any additional logging? Unfortunately my test account isn't able to interact with this particular resource.

[jfarchergamma]

Sorry for delayed response, I think I may have stumbled onto something but working to confirm it.
If the connect instance was created using the Terraform config I'm working with all is good, but if the instance was created but imported later to the terraform state then the ID presents as the ARN.

A terraform state show on the resource for both created and imported correctly returns the ID, but when referring to the ID in other resources such as when creating the associated storage configs the imported version errors complaining about the ID structure and viewing the output shows the ARN.

[D2G-TimP]

I can confirm that I encountered the same problem when importing an existing Connect Instance.