Skip to content

Latest commit

 

History

History
99 lines (79 loc) · 2.89 KB

README.md

File metadata and controls

99 lines (79 loc) · 2.89 KB

Profiles Management Library

This module aims to be a standalone python module focused around:

  1. Exposing the ProfileManagementRepresentation (PMR) classes
  2. Functions that manipulate the cluster, based on a PMR

Overview

The PMR consists of a list of classes for defining the Profiles and their Contributors in an agnostic way. Specifically it is a custom data structure consisting of a mapping of Profile objects.

With this representation it's then possible to write functions that manipulate a Kubeflow cluster based on a PMR.

The structure of the PMR, in YAML can be seen in the following section:

profiles:
- name: ml-engineers
  owner:
    kind: User
    name: [email protected]
  resources:
    hard:
      limits.cpu = "1"
  contributors:
  - name: [email protected]
    role: admin
  - name: [email protected]
    role: edit
  - name: [email protected]
    role: view
- name: data-engineers
  owner:
    kind: User
    name: [email protected]
  contributors:
  - name: [email protected]
    role: edit
  - name: [email protected]
    role: edit
  resources: {}

Consume from Charms

This library is meant to be used by other Charms that will construct a PMR and run the functions of this module to update the cluster's state.

Here's is some psudo-code that can show how a potential charm could interact with the code of this library:

from profiles_management.pmr import classes
from profiles_management import create_or_update_profiles


class ConsumerCharm(ops.CharmBase):

    def __init__(self, framework: ops.Framework):
        super().__init__(framework)

        self.framework.observe(self.on["container"].pebble_custom_notice, self.reconcile)

    def reconcile(self, _: ops.PebbleNoticeEvent):
        # create a PMR and call the library's functions
        pmr = classes.ProfilesManagementRepresentation()

       # Iterate over a list of groups
        for profile_name in ["ml-engineers"]:
            # create the contributors of the Profile
            contributors = []
            # Iterate over a list of users
            for user, role in [("[email protected]", classes.ContributorRole.EDIT)]:
                contributors.append(classes.Contributor(
                    name=user,
                    role=self.pmr_role_from_oidc_role(role)
                ))

            # resource quota
            quota = classes.ResourceQuotaSpecModel.model_validate({"hard": {"limits.cpu": "1"}})

            # owner
            owner = classes.Owner(name="[email protected]"
                                  kind=classes.UserKind.USER)
            # create the Profile
            profile = classes.Profile(
                name=group_name,
                resources=quota,
                contributors=contributors,
                owner=owner
            )

            # add Profile to PMR
            pmr.add_profile(profile)


        create_or_update_profiles(pmr)