Terraform 1.3.4 - errors for deleted resources after partial destroy

[kirikors]

Terraform Version

Terraform v1.3.4
on windows_amd64
+ provider registry.terraform.io/aliyun/alicloud v1.190.0
+ provider registry.terraform.io/hashicorp/null v3.2.0

Terraform Configuration Files

Sample code to reproduce the issue

Debug Output

TF debug logs for the second destroy run

Expected Behavior

destroy operation doesn't validate references to already deleted resources

Actual Behavior

references to non-existing (already deleted) resources are validated during destroy operation, resulting in an exception

Steps to Reproduce

The issue happens on TF destroy operation after a failure (when previous destroy run managed to delete some of the resources, but not all).

To reproduce:

1. terraform apply
2. manually create a resource that will block VPC deletion
3. run first terraform destroy -> will delete everything except the VPC, VPC deletion will fail (see gist)
4. run second terraform destroy -> will fail because of references to non-existing resources (those deleted by the previous partial destroy run) - see gist

Additional Context

The issue doesn't happen with Terraform 1.2.9 - second destroy attempt is successful.

The issue is not limited to just outputs validation, we get similar errors in other places (see gist)

References

Same issue as #31975, but happens in Terraform 1.3.4 (that was supposed to fix it)

[jbardin]

Thanks @kirikors! Because root module outputs have a different use from child modules, this is a slightly different case than the linked issue.

[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.