Skip to content

sagemaker studio - github templates for pipeline - failing at SageMakerModelDeploySeedCodeCheckinProjectTriggerLambdaInvoker.

0

I am trying to create a new project in sagemaker studio and I sleected the default template for github, however it is failing to build. I started by creating a CONNECTION in codepipeline and noted my ARN.

I followed the instruction in : https://docs.aws.amazon.com/sagemaker/latest/dg/sagemaker-projects-walkthrough-3rdgit.html

Enter image description here

The error log says :

Enter image description here

Is this caused by codestar deprecation? I can see my ARN has codeconnection whereas the cloudformation templates still use codestar

1 Answer
0

Hi, I think I have the same problem. This is the auth problem I get in CloudFormation:

errorCF

This is the AmazonSageMakerAdmin-ServiceCatalogProductsServiceRolePolicy policy, the one attached to the Amazon SageMaker ServiceCatalog ProductsLaunch Role.

AmazonSageMakerAdmin-ServiceCatalogProductsServiceRolePolicy

As you can see, it only allows actions from codestar-connections to codestar-connections resources; and from codeconnections to codeconnections resources. However, the problem is that the action triggered in Cloudformation from the Project template links a codestar-connections action to a codeconnection resource: it mixes both names. And, of course, the action is denied, because the policy does not have an explicit allow for that mix.

I think those policies should be reviewed, because codepipeline seems to keep providing the codestar source connection (https://docs.aws.amazon.com/codepipeline/latest/userguide/action-reference-CodestarConnectionSource.html). So, if you provide a connection created recently (whose ARN now includes codeconnection), you will get that mixed reference when the template is deployed.

Until a solution arrives, you can attach a custom policy to the Amazon SageMaker ServiceCatalog ProductsLaunch Role:

Enter image description here

Hope it helps!!

answered a year ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.