Dualstack prefix forcibly removed from ALIAS records

[jdubeau123]

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Terraform Version

Terraform v0.11.10

• provider.aws v1.44.0
• provider.terraform v1.0.2

Affected Resource(s)

• aws_route53_record

Terraform Configuration Files

resource "aws_route53_record" "big_happy_alias_record" {
  zone_id = "${var.zone_id}"
  name    = "${var.name}"
  type    = "${var.type}"

  alias {
    name                   = "${var.alias_name}"
    zone_id                = "${var.alias_zone_id}"
    evaluate_target_health = "${var.alias_evaluate_target_health}"
  }

  allow_overwrite = "${var.allow_overwrite}"
}

Expected Behavior

I imported an existing DNS record into terraform, which was an ALIAS record that contained a "dualstack" prefix in its alias target. I then ran a trivial apply operation on the resource to change the "allow_overwrite" value to false. This should not have silently changed the ALIAS target in the record.

Actual Behavior

The initial import worked fine, but when I run any apply whatsoever on the resource it ends up silently stripping off the "dualstack" prefix without saying anything in the plan.

Steps to Reproduce

1. terraform apply with any changes after importing an ALIAS record whose target has a dualstack prefix

Important Factoids

I think I narrowed down why this is happening. I inspected the debug output and found that the prefix did indeed seem to be stripped off halfway through the apply operation, so I checked the provider code and found this function. I haven't read this in any detail but it appears this would cause any ALIAS record in terraform to have its dualstack prefix stripped off 100% of the time. Can anyone provide any insight into why this would be the desired/appropriate behavior? We actually do want many of our ALIAS records to use dualstack, so this pretty much means the route 53 resources are largely not useful at all in our case.

References

[paulcdejean]

So it's just impossible to have a dualstack alias record if you use terraform???

[paulcdejean]

A perpetual changed loop was "fixed" by just forcing all alias records to non dualstack!

🤦‍♂️

95ce334

[ngaya-ll]

I'm having a related issue with terraform 0.12.12 / aws provider 2.33. If I create an alias record and then add/remove the dualstack prefix in the terraform configuration, Terraform detects no changes to apply so the record does not get updated.

[alvarolinarescabre]

Any news about it????????

[nmofonseca]

I am having the same issue. Is this going to be picked up eventually?

[guilatrova]

Updating the code reference mentioned: https://github.com/hashicorp/terraform-provider-aws/blob/main/internal/service/route53/record.go#L956

[himalayKondekar]

We are facing the same issue.
system info:
Terraform v1.1.9
on darwin_arm64

• provider registry.terraform.io/hashicorp/aws v4.2.0

When we can expect to be resolved

[github-actions]

This functionality has been released in v4.24.0 of the Terraform AWS Provider. Please see the Terraform documentation on provider versioning or reach out if you need any assistance upgrading.

For further feature requests or bug reports with this functionality, please create a new GitHub issue following the template. Thank you!

[jtsoi]

For anyone looking finding this issue post 4.24.0 release:
If you have a state diff due to dualstack.ALB_NAME.*.elb.amazonaws.com -> ALB_NAME.*.elb.amazonaws.com

Here is what I found:

• AWS Added IPv6 support for VPC ALBs in 2021 (https://aws.amazon.com/blogs/networking-and-content-delivery/dual-stack-ipv6-architectures-for-aws-and-hybrid-networks/)
• With ELB-Classic you did need the dualstack prefix when configuring the alias record.
• When creating an alias record via the AWS Console, the dualstack prefix is added automatically (as of 2022-08)
• The ALB page in AWS console does not have the dualstack prefix for the ALB DNS Name field.
• Both dualstack.ALB_NAME.*.elb.amazonaws.com and ALB_NAME.*.elb.amazonaws.com respond to AAAA queries and can be visited using IPv6

My conclusion is that for recently created VPC dualstack ALBs, the dualstack prefix is redundant and can be omitted.
Apply at your own risk.

[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.