This repository does not contain any runs of the UATs. As part of our vulnerability response implementation, and for being able to catch regressions, there has to be a workflow that runs the MLflow specific UATs.
Please note some of the notebooks that test MLflow also require components from the CKF bundle, such as kserve, and in the specific case of the e2e wine example, pipelines.
What needs to get done
Enable a scheduled workflow that runs MLflow UATs, similar to this - it can be AKS, EKS or even Microk8s on an AWS instance. The important part is to test the bundle, the infrastructure does not have impact on this.
Definition of Done
There is a scheduled workflow that runs the MLflow UATs.
Context
This repository does not contain any runs of the UATs. As part of our vulnerability response implementation, and for being able to catch regressions, there has to be a workflow that runs the MLflow specific UATs. Please note some of the notebooks that test MLflow also require components from the CKF bundle, such as kserve, and in the specific case of the e2e wine example, pipelines.
What needs to get done
Enable a scheduled workflow that runs MLflow UATs, similar to this - it can be AKS, EKS or even Microk8s on an AWS instance. The important part is to test the bundle, the infrastructure does not have impact on this.
Definition of Done
There is a scheduled workflow that runs the MLflow UATs.