runners-scale-up fails with 'AuthFailure.ServiceLinkedRoleCreationNotPermitted'

[compiaffe]

Summary

When following the readme, using the example configuration and adjusting the Github app permissions as per https://github.com/philips-labs/terraform-aws-github-runner/issues/100#issuecomment-665923045 the scale-up lambda fails to create the EC2 instance due to ServiceLinkedRoleCreationNotPermitted

Steps to reproduce

• Do step 1 of Github app setup
• Checkout terraform-aws-github-runner repo, cd into example folder
• Download lambda zips
• Create terraform.tfvars file with Github App credentials
• run terraform init && terraform apply
• Trigger a build on Github

What is the current bug behavior?

Github app sends webhook, webhook lambda forwards it, scaleup-lambda throws error:

...
ERROR	AuthFailure.ServiceLinkedRoleCreationNotPermitted: The provided credentials do not have permission to create the service-linked role for EC2 Spot Instances.
    at Request.extractError (/var/task/index.js:41424:35)
    at Request.callListeners (/var/task/index.js:47771:20)
    at Request.emit (/var/task/index.js:47743:10)
    at Request.emit (/var/task/index.js:18467:14)
    at Request.transition (/var/task/index.js:17801:10)
    at AcceptorStateMachine.runTo (/var/task/index.js:26145:12)
    at /var/task/index.js:26157:10
    at Request.<anonymous> (/var/task/index.js:17817:9)
    at Request.<anonymous> (/var/task/index.js:18469:12)
    at Request.callListeners (/var/task/index.js:47781:18) {
  code: 'AuthFailure.ServiceLinkedRoleCreationNotPermitted',
  time: 2020-07-30T15:03:24.631Z,
  requestId: 'c7bab39e-b75c-4e7d-bc29-6622b3d4ddb1',
  statusCode: 403,
  retryable: false,
  retryDelay: 68.19342592727871
}

What is the expected correct behavior?

Scale up lambda should create EC2 instance

Possible fixes

I'm sure this is a IAM permissions issue. I am rather new to both AWS and terraform and am not sure in which of them this needs to be solved and how to go about it.
Would be great to get some pointers.

[cmcconnell1]

Hey @compiaffe ,

Assume that you are using the v0.2.0 tag?
And, curious if you always see this error now?

• i.e.: can you terraform destroy, delete ./.terraform dir, and then terraform init/plan/apply again a few times and validate that you always get this error?

• I don't see the error you noted--although the effect is the same for both of us, as we are not seeing any EC2 spot instances deployed (ref: https://github.com/philips-labs/terraform-aws-github-runner/issues/100#issuecomment-665995200 ). And I definitely have terraform destroyed/applied probably ten or so times in the past couple of days.

• For the record, I'm using terraform 0.12.28 on macos, curious what TF version you are using?

I ended up importing the project as I needed to make changes to some submodules code with things like the hard coded tags, etc. in case its helpful, here is how I have my internal module structure, based on the default/example from this project

tree
.
├── README.md
├── lambdas-download
│   ├── main.tf
│   ├── runner-binaries-syncer.zip
│   ├── runners.zip
│   ├── terraform.tfstate
│   └── webhook.zip
├── main.tf
├── outputs.tf
├── providers.tf
├── terraform-aws-github-runner.tfvars
├── terraform.secret.auto.tfvars
├── terraform.tf
└── variables.tf

1 directory, 13 files
cmcc@cmcc:default $ cat ../../.terraform-version
0.12.28

I found this curious seeing that this has apparently been an intermittent issue for folks in the past with other projects:
LeanerCloud/AutoSpotting#187

This issue notes how they apparently took their fix and made it less permissive and were able to get it resolved:
LeanerCloud/AutoSpotting#416 (comment)

[npalm]

I am quite sure it is an IAM issue. In this repo you find a bit more infor how to create the required service linked roles. When time I will add them also to this repo.

See, you only will need the one for spot https://github.com/npalm/terraform-aws-gitlab-runner#service-linked-roles

Via terraform you create the services link role as follow:

resource "aws_iam_service_linked_role" "spot" {
  aws_service_name = "spot.amazonaws.com"
}
``

Another quick fix would be first to create manaully via the AWS console a spot instance and remove it. You will see that AWS will create the required role for you.

[compiaffe]

@npalm Thanks for the info, will try it out. Due to upcoming holidays might not report back before October.

[toots]

I was having the same issue and adding

resource "aws_iam_service_linked_role" "spot" {
  aws_service_name = "spot.amazonaws.com"
}

Did indeed fix it. Thanks!

[npalm]

Docs are updated via #229