[Bug]: aws_security_group destruction stucks on HTTP 400 response, masking issues with an incorrect plan

[ei-grad]

Terraform Core Version

1.3.7

AWS Provider Version

4.51.0

Affected Resource(s)

aws_security_group

Expected Behavior

1. When removing the security group along the rules referencing it, at first the rules should be deleted, then the group. Dependent resource must be updated to be independent from dependency before its dependency could be destroyed.
2. During this operation, if terraform triggers a request which results in an HTTP 400 response from AWS API - the operation should be cancelled and considered failed.

Actual Behavior

Terraform tries to destroy the security group before removing the rules referencing it, and stucks in a loop receiving HTTP 400 from AWS for continuously retrying incorrect requests.

Relevant Error/Panic Output Snippet

No response

Terraform Configuration Files

Initial state:

resource "aws_security_group" "test1" {
  name = "test1"
  ingress {
    protocol  = "-1"
    from_port = 0
    to_port   = 0
    security_groups = [
      aws_security_group.test2.id,
    ]
  }
}

resource "aws_security_group" "test2" {
  name = "test2"
}

Target state:

resource "aws_security_group" "test1" {
  name = "test1"
  ingress {
    protocol  = "-1"
    from_port = 0
    to_port   = 0
    cidr_blocks = ["0.0.0.0/0"]
  }
}

# aws_security_group.test2 removed

Steps to Reproduce

1. Create test1 and test2 security rules, where test1 group has the rule referencing test2
2. Try to remove test2 along with the rule referencing it in a single terraform apply.

Debug Output

No response

Panic Output

No response

Important Factoids

No response

References

Similar issue - #265, it points on a wrong (IMO) facts, and contains a long discussion with some unrelated things, so I decided to start a new one.

Would you like to implement a fix?

None

[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.

[justinretzolk]

Similar to #27476