Integration in the backend for Model Storage(s)

[tarilabs]

Is your feature request related to a problem? Please describe.
Following

• leverage upcoming helper methods of orchestration of Add upload_artifact_and_register_model method to python client #761
  • to S3 with boto3 as extra feat(py): Add support for storing models in S3 #765
  • to OCI as ModelCar by leveraging oci-layout Add save_to_oci_registry python client method #800

we should have a way for a Model Registry instance to indicate/point to Storage solution like S3 buckets, OCI registries, which the MR-admin designate for the MR-user to make use of for the actual storing of the model asset, prerequisite to metadata Registration in MR (including URI pointer, as it's done today).

Describe the solution you'd like
MR API can return to the MR py client the configured Storage solutions; then the MR py client allows the MR-user to select one, or defaults to the designated default one.

Describe alternatives you've considered
Alternative is to continue as-is, but requires the MR-user to ask the admins about which servers to be used.

Additional context
MR is not currently and should not become a proxy/abstraction/man-in-the-middle, the scope is to have "pointers" to the Storage solutions to be used, which is setup by the MR-admin, and consumed by the MR-user via the MR py client (or API).

Embeds also from:

• docs: initial proposal for OCI artifact registry #48
• https://github.com/kubeflow/community/pull/682/files#diff-aaf54745ecb36016135c83a5a41a03025574ecb492aec56ef6d2c7c902abfe17R180

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[tarilabs]

As we walked through the user story discovery, we're finding that K8s Secrets are more than enough to cover for Storage access; as part of this goal however, we realized that creating and Async K8s (vanilla) Job to convert into OCI/ModelCar to be most helpful, job which we intend to trigger from the UI; see also here.

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[github-actions]

This issue has been automatically closed because it has not had recent activity. Please comment "/reopen" to reopen it.