awslabs / awslabs/aiops-modules

[FEATURE] Multi-account basic template - add ability to invoke SM build pipeline in target/dev environment from tooling

Open
#28 0 comments 0 reactions 0 assignees View on GitHub
enhancement
Dominant language
Python
Stars
105
Forks
36
PR merge metrics
No merged PRs in 30d

Description

**Is your feature request related to a problem? Please describe.**
Existing [basic multi-account template](https://github.com/awslabs/mlops-modules/tree/main/modules/sagemaker/sagemaker-templates-service-catalog/templates/multi_account_basic) during build phase, [calls SageMaker pipeline](https://github.com/awslabs/mlops-modules/blob/main/modules/sagemaker/sagemaker-templates-service-catalog/templates/multi_account_basic/seed_code/build_app/buildspec.yml#L15) in the same account where the CodeBuild runs. However, this does not cater to multi-account scenario where a separate tooling account hosts the infra, and development/experimentation account is used to run training and provides access to the data.

**Describe the solution you'd like**
- Add assume role to the build CodeBuild to run training pipeline in the target/dev/experimentation account.
- Add required IAM config/ S3 buckets/ KMS in dev to run the pipeline.
- Keep the template flexible to allow simple (current) scenario that runs the pipeline in the same account.

Contributor guide

Open the contributing guide

Research direction

Start with the multi_account_basic template and its seed_code/build_app/buildspec.yml, especially the existing SageMaker pipeline invocation. Trace the template's IAM, S3, and KMS configuration and compare it with the requested target/dev account flow. Done means the template supports both same-account and assumed-role execution with the required development-account resources.

Written by the indexing model from the issue text.

Assessment

Tech stack
aws
Domain
cloud, infrastructure, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.