[Bug]: aws_servicecatalog_provisioning_artifact - modifying timeouts doesn't work after initialisation

[boxrick]

Terraform Core Version

1.5.0

AWS Provider Version

5.4.0

Affected Resource(s)

aws_servicecatalog_provisioning_artifact

Expected Behavior

After setting timeouts upon first initialisation, then updating the timeouts later I would expect the timeouts to be updated.

Actual Behavior

However it only appears the initial values are used and then never updated again. I would expect when timeouts are modified it would use the updated value

Relevant Error/Panic Output Snippet

Terraform shows no changes

Terraform Configuration Files

# Product
resource "aws_servicecatalog_product" "this" {
  name        = "iamaproduct"
  owner       = "iamanowner"
  description = "iamadescription"
  type        = "CLOUD_FORMATION_TEMPLATE"

  provisioning_artifact_parameters {
    template_url = "https://gist.githubusercontent.com/boxrick/6ff96c96a611c54f3cdbf4de61e1a174/raw/1dcc426188008b5fbbdf773369863145a9a1ff25/cfn?_sm_au_=iVV6RFrLNpW0VrFRpGsWvKttvN1NG"
    name         = "1.0.0"
    type         = "CLOUD_FORMATION_TEMPLATE"
  }
}

# Artifact
resource "aws_servicecatalog_provisioning_artifact" "this" {
  name         = "1.2.3"
  product_id   = aws_servicecatalog_product.this.id
  type         = "CLOUD_FORMATION_TEMPLATE"
  template_url = "https://gist.githubusercontent.com/boxrick/6ff96c96a611c54f3cdbf4de61e1a174/raw/1dcc426188008b5fbbdf773369863145a9a1ff25/cfn?_sm_au_=iVV6RFrLNpW0VrFRpGsWvKttvN1NG"

  timeouts {
    create = "60s"
    read   = "60s"
    update = "60s"
    delete = "60s"
  }
}

Steps to Reproduce

Create the above resources, modify the 'timeouts' section in the aws_servicecatalog_provisioning_artifact. Then re-apply and it will not show any changes.

The only way to update is to delete the aws_servicecatalog_provisioning_artifact and re-create.

Debug Output

No response

Panic Output

No response

Important Factoids

No response

References

No response

Would you like to implement a fix?

No

[github-actions[bot]]

Community Note

Voting for Prioritization

• Please vote on this issue by adding a 👍 reaction to the original post to help the community and maintainers prioritize this request.
• Please see our prioritization guide for information on how we prioritize.
• Please do not leave "+1" or other comments that do not add relevant new information or questions, they generate extra noise for issue followers 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.

[boxrick]

When I have delved further into it the 'private' section within the state file when base64 decoded seems to store the timeouts in addition to them being stored in the resource itself. Manually adjusting the resource timeouts in the state file doesn't seem to do anything, but deleting or updating the private metadata you can hack it in.

"private": "eyJlMmJmYjczMC1lY2FhLTExZTYtOGY4OC0zNDM2M2JjN2M0YzAiOnsiY3JlYXRlIjozMDAwMDAwMDAwMDAsImRlbGV0ZSI6MzAwMDAwMDAwMDAwLCJyZWFkIjo2MDAwMDAwMDAwMDAsInVwZGF0ZSI6MzAwMDAwMDAwMDAwfX0="

echo 'eyJlMmJmYjczMC1lY2FhLTExZTYtOGY4OC0zNDM2M2JjN2M0YzAiOnsiY3JlYXRlIjozMDAwMDAwMDAwMDAsImRlbGV0ZSI6MzAwMDAwMDAwMDAwLCJyZWFkIjo2MDAwMDAwMDAwMDAsInVwZGF0ZSI6MzAwMDAwMDAwMDAwfX0=' | base64 -d

{"e2bfb730-ecaa-11e6-8f88-34363bc7c4c0":{"create":300000000000,"delete":300000000000,"read":600000000000,"update":300000000000}}

[justinretzolk]

Hey @boxrick 👋 Thank you for taking the time to raise this! I wouldn't expect that adjusting the custom timeouts for the resource would trigger an infrastructure change, since they're not actually modifying the resource, but rather adjusting the amount of time that the provider allows for particular parts of the CRUD operations. Are you seeing issues where you've extended one of the timeouts and the adjusted time is not being respected?

[boxrick]

Hey @boxrick 👋 Thank you for taking the time to raise this! I wouldn't expect that adjusting the custom timeouts for the resource would trigger an infrastructure change, since they're not actually modifying the resource, but rather adjusting the amount of time that the provider allows for particular parts of the CRUD operations. Are you seeing issues where you've extended one of the timeouts and the adjusted time is not being respected?

Yes, exactly this. Once I have adjusted the custom timeouts, then updated the resource with other values the custom timeouts do not get changed or used with the resource. The only way to update these values is to destroy the resource entirely, or to change it in the state file directly as I list above.