Ressource aws_s3_bucket_lifecycle_configuration > object_size_greater_than doesn’t work when used at Filter level

[flibustier]

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 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
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Terraform CLI and Terraform AWS Provider Version

Terraform 0.14 but reproduced with 0.15 and 1.0
AWS provider 4.4.0 and 4.3.0

Affected Resource(s)

• aws_s3_bucket_lifecycle_configuration (filter field)

Terraform Configuration Files

This code doesn’t update the real infrastructure, it keeps planning the same change over and over

resource "aws_s3_bucket_lifecycle_configuration" "my_bucket_lifecycles" {
  bucket = aws_s3_bucket.my_bucket.id

  rule {
    status = "Enabled"
    id     = "to-intelligent-tiering"

    filter {
      object_size_greater_than = 5242880
    }

    transition {
      days          = 1
      storage_class = "INTELLIGENT_TIERING"
    }
  }

Debug Output

The output produced

  ~ resource "aws_s3_bucket_lifecycle_configuration" "my_bucket_lifecycles" {
        id     = "my-bucket"
        # (1 unchanged attribute hidden)

      ~ rule {
            id     = "to-intelligent-tiering"
            # (1 unchanged attribute hidden)

          ~ filter {
              + object_size_greater_than = "5242880"
            }

            # (1 unchanged block hidden)
        }
        # (3 unchanged blocks hidden)
    }

The output stay the same when applying multiple times (as it doesn’t change the real infrastructure)

Panic Output

If I manually modify the corresponding field in the AWS console (I put 2MB of object size minimal), I got the following panic when terraform is planning the changes :

Error: error setting rule: rule.0.filter.0.object_size_greater_than: '' expected type 'string', got unconvertible type 'int', value: '2097152'

Quick fix

It’s possible to avoid this using the and clause :

filter {
      and {
        object_size_greater_than = 5242880
      }
}

It produced a working output (we can observe that the int type is here, instead of string in the faulty case)

  ~ resource "aws_s3_bucket_lifecycle_configuration" "my_bucket_lifecycles" {
        id     = "my-bucket"
        # (1 unchanged attribute hidden)

      ~ rule {
            id     = "to-intelligent-tiering"
            # (1 unchanged attribute hidden)

          ~ filter {
              + and {
                  + object_size_greater_than = 5242880
                }
            }

            # (1 unchanged block hidden)
        }
        # (3 unchanged blocks hidden)
    }

And the real infrastructure is changed 🎉

References

I guess this code seems incriminated :

terraform-provider-aws/internal/service/s3/bucket_lifecycle_configuration.go

Line 95 in 1d268ac

Elem: &schema.Resource{

[anGie44]

Hi @flibustier , thank you for raising this issue. Just wanted to quickly note that I think this PR (#23441) should fix this as it addresses the panic your are seeing. We weren't handling the data conversion (int -> string) in both rule.filter.object_size_greater_than and rule.filter.object_size_less_than

[flibustier]

Awesome ! I will track the progress of the PR then :) Thanks !!

[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.