TRGRND 0.68.4/0.72.0 timeout while waiting for plugin to start

[RazvanGherlea]

Describe the bug

I am trying to integrate the latest version of terragurnt into our pipeline. The version upgrade is from 0.58.2 to 0.72.0. The code works as expected on the older version and when executing terragrunt prun-all plan on the whole AWS Account level everything works as expected. After upgrading to the latest version 0.72.0 I can see the following error on some resources

Error: failed to read provider configuration schema for registry.terraform.io/hashicorp/tls: failed to instantiate provider "registry.terraform.io/hashicorp/tls" to obtain schema: timeout while waiting for plugin to start

If I run then the terragrunt plan with --terragrunt-working-dir argument on the directory that failed, everything is fine and it is running well. It fails also for resources that are already created and have no change to be performed at the time of the plan execution.

If I run terragrunt plan on all the failed directory, it will work as expected and resume an account level run-all plan everything is green and there is no failure

Steps To Reproduce

Steps to reproduce the behavior, code snippets and examples which can be used to reproduce the issue.

Be sure that the maintainers can actually reproduce the issue. Bug reports that are too vague or hard to reproduce are hard to troubleshoot and fix.

If I switch back to the old version 0.58.2, clean local cache and lock files and re run run-all plan on whole account level everything is green. There are no failures

// paste code snippets here

Expected behavior

Terragrunt plan will work as expected and there are no timeout issues.

Nice to haves

• Terminal output
• Screenshots

Versions

• Terragrunt version: old working as expected (0.58.2), new version 0.72.0/0.68.4 are failing
• OpenTofu/Terraform version: terraform version 1.5.1
• Environment details (Ubuntu 24.04, Windows 10, etc.): Ubuntu 24.04 LTS

Additional context

I tend to believe that the latest version of terragrunt requires more memory and CPU. But at the same time I am executed the test from MAC OS (M1 PRO) as well and got the same results.

[RazvanGherlea]

I also created a gruntwork discussion if this will help. Jus to make sure that I add as much details as possible

Discussion: https://github.com/orgs/gruntwork-io/discussions/854

[RazvanGherlea]

There was another issue similar to this. The reported mentioned it was related to HW requirements. but I am running the lower verison of terragrunt on the same target and everything works as expected. So cold it be that the HW requirements would be so major from version 0.58.2 to the latest version ?

• ISSUE: terragrunt run-all : timeout while waiting for plugin to start #2252

[tolikkostin]

Hi @RazvanGherlea. I have the same issue. And I use Mac with M1 and I noticed that the same operation works fine within my CI/CD pipeline which uses linux/amd64 architecture (terragrunt v0.64.0, tofu v1.7.3). I also noticed that it happens only in case OpenTofu v1.7.x on Mac, because Terraform v1.7.2 works well on my Mac. I checked Terragrunt v0.55.0 and v0.64.0 with both versions Terraform binary works well, in case OpenTofu works only on first module and hungs on next. It raises errors like these:

Planning failed. OpenTofu encountered an error while generating this plan.

│ Error: timeout while waiting for plugin to start
│ 

P.S. I updated OpenTofu to v1.8.8 and it works well finally