Unsupported attribute - Region

[rhysmcn]

Description

I have been working internally within my company deploying resources, and today I have been hit with an issue directly related to the latest changes merged this morning.

│ Error: Unsupported attribute
│ 
│   on .terraform/modules/iriusrisk_environment.iriusrisk_network.vpc/vpc-flow-logs.tf line 28, in locals:
│   28:     "arn:${data.aws_partition.current[0].partition}:logs:${data.aws_region.current[0].region}:${data.aws_caller_identity.current[0].account_id}:log-group:${log_group.name}:*"
│ 
│ This object has no argument, nested block, or exported attribute named
│ "region".
╵

Please, can you fix this ASAP, as we (like many others) pin to use the latest minor release of this module.

I believe the issue here is that we didnt consider backwards compatibility when making the change.

If your request is for a new feature, please use the Feature request template.

• ✋ I have searched the open/closed issues and my issue is not listed.

⚠️ Note

Before you submit an issue, please perform the following first:

1. Remove the local .terraform directory (! ONLY if state is stored remotely, which hopefully you are following that best practice!): rm -rf .terraform/
2. Re-initialize the project root to pull down modules: terraform init
3. Re-attempt your terraform plan or apply and check if the issue still persists

Versions

• Module version [Required]: v5.21.1 <-- this is the issue.
• Tofu version: v1.9.1
• Provider version(s): 5.x.x

Reproduction Code [Required]

We use the latest version of the module, and any time you try to terraform plan with it, it fails

Steps to reproduce the behavior:

We run TF/Tofu in pipelines so it is a clean slate each run. But, essentially:

1. Run terraform init
2. terraform plan

We get this error, which is above.

Expected behavior

plans to work

Actual behavior

plans failing

Terminal Output Screenshot(s)

Additional context

[geoL86]

Looks like it relates to 5.21.1 version

[rhysmcn]

It looks like this issue is the same as:
#1201
#1202

[Nisarg-Satani]

try with v5.21.0
worked for me
There is backward compatibility issue in v5.21.1 (latest)

[robcoward]

Please update the module's dependencies to reflect that this latest version 5.21.1 requires v6.0 of the aws provider

[rhysmcn]

It works with 5.21.0, the issue is the latest release.

@robcoward Of course, updating the provider will fix the issue. However, its not a simple change - In fact, going from v5 to v6 is a major change. These changes in this module need to be backwards compatible and contributors should be testing both versions before merging otherwise these issues occur. v5 and v6 are both fully supported and maintained versions of the AWS provider, therefore these changes need to account for both versions.

[antonbabenko]

Version 5.21.1 has been removed, and version 6.0.0 has just been released in #1205.

[rhysmcn]

Hi @antonbabenko,

Thank you for making that change! I face another issue, perhaps you have some idea on it?

When we do an terraform init, it still fetches 5.21.1 - Do you know why?

│ error downloading
│ 'https://github.com/terraform-aws-modules/terraform-aws-vpc?ref=v5.21.1':
│ /usr/bin/git exited with 1: error: pathspec 'v5.21.1' did not match any
│ file(s) known to git

On here it refs 5.21.0 and on the GH tags the same. But why does TF still try to fetch the 5.21.1?

Btw we run tf commands in an isolated pipeline so each build it creates all files from scratch (lock files etc). I also have tried terraform init --upgrade but it persists

[Yantrio]

Hi from the OpenTofu team 👋. We've got a PR ready to remove this from the opentofu registry here: opentofu/registry#2320

I would love it if @antonbabenko could signoff on this just to say it's okay to remove.

[rhysmcn]

Hi @Yantrio!

Looks like it has been approved by @antonbabenko, so we can merge and deploy when you are ready 🚀

Thank you for making this change as quickly as you did - How could we improve the registry by dynamically fetching the tags from GH? Is this something the Tofu team have explored? I don't know the breadth and depth of the technicalities but I would be up for lending a hand to implement something like this, if needed.

[Yantrio]

I think this is something I can discuss with the rest of the opentofu team and its not something we need to discuss here as this can quickly devolve into a security debate :D.

Removing module versions is an extremely rare case and for now a manual process has served us very well.

[rhysmcn]

Agreed - It was just to see if it has already been discussed/explored.

Thanks again. @Yantrio

Just an FYI (This is probably not the right place to report this) - On the registry, even after the PR merge it still shows 5.21.1, is there some pipeline that we need to run on Tofu side to deploy the changes to remove the version?

Edit: it is solved, must have needed to sync