glance/doc/source/admin/policies.rst
Ghanshyam Mann c107629f90 [goal] Deprecate the JSON formatted policy file
As per the community goal of migrating the policy file
the format from JSON to YAML[1], we need to do two things:

1. Change the default value of '[oslo_policy] policy_file''
config option from 'policy.json' to 'policy.yaml' with
upgrade checks.

2. Deprecate the JSON formatted policy file on the project side
via warning in doc and releasenotes.

Also convert the ./glance/tests/etc/policy.json to policy.yaml
file. Replace policy.json to policy.yaml ref from doc and tests.

[1]https://governance.openstack.org/tc/goals/selected/wallaby/migrate-policy-format-from-json-to-yaml.html

Depends-On: https://review.opendev.org/c/openstack/nova/+/773192
Change-Id: I17d0374dd4223688e5f95253802a4ae87377953a
2021-01-29 15:31:47 -08:00

6.1 KiB

Policies

Warning

JSON formatted policy file is deprecated since Glance 22.0.0 (Wallaby). This oslopolicy-convert-json-to-yaml tool will migrate your existing JSON-formatted policy file to YAML in a backward-compatible way.

Glance's public API calls may be restricted to certain sets of users using a policy configuration file. This document explains exactly how policies are configured and what they apply to.

A policy is composed of a set of rules that are used by the policy "Brain" in determining if a particular action may be performed by the authorized tenant.

Constructing a Policy Configuration File

A policy configuration file is a simply JSON object that contain sets of rules. Each top-level key is the name of a rule. Each rule is a string that describes an action that may be performed in the Glance API.

The actions that may have a rule enforced on them are:

  • get_images - List available image entities
    • GET /v1/images
    • GET /v1/images/detail
    • GET /v2/images
  • get_image - Retrieve a specific image entity
    • HEAD /v1/images/<IMAGE_ID>
    • GET /v1/images/<IMAGE_ID>
    • GET /v2/images/<IMAGE_ID>
  • download_image - Download binary image data
    • GET /v1/images/<IMAGE_ID>
    • GET /v2/images/<IMAGE_ID>/file
  • upload_image - Upload binary image data
    • POST /v1/images
    • PUT /v1/images/<IMAGE_ID>
    • PUT /v2/images/<IMAGE_ID>/file
  • copy_from - Copy binary image data from URL
    • POST /v1/images
    • PUT /v1/images/<IMAGE_ID>
  • add_image - Create an image entity
    • POST /v1/images
    • POST /v2/images
  • modify_image - Update an image entity
    • PUT /v1/images/<IMAGE_ID>
    • PUT /v2/images/<IMAGE_ID>
  • publicize_image - Create or update public images
    • POST /v1/images with attribute is_public = true
    • PUT /v1/images/<IMAGE_ID> with attribute is_public = true
    • POST /v2/images with attribute visibility = public
    • PUT /v2/images/<IMAGE_ID> with attribute visibility = public
  • communitize_image - Create or update community images
    • POST /v2/images with attribute visibility = community
    • PUT /v2/images/<IMAGE_ID> with attribute visibility = community
  • delete_image - Delete an image entity and associated binary data
    • DELETE /v1/images/<IMAGE_ID>
    • DELETE /v2/images/<IMAGE_ID>
  • add_member - Add a membership to the member repo of an image
    • POST /v2/images/<IMAGE_ID>/members
  • get_members - List the members of an image
    • GET /v1/images/<IMAGE_ID>/members
    • GET /v2/images/<IMAGE_ID>/members
  • delete_member - Delete a membership of an image
    • DELETE /v1/images/<IMAGE_ID>/members/<MEMBER_ID>
    • DELETE /v2/images/<IMAGE_ID>/members/<MEMBER_ID>
  • modify_member - Create or update the membership of an image
    • PUT /v1/images/<IMAGE_ID>/members/<MEMBER_ID>
    • PUT /v1/images/<IMAGE_ID>/members
    • POST /v2/images/<IMAGE_ID>/members
    • PUT /v2/images/<IMAGE_ID>/members/<MEMBER_ID>
  • manage_image_cache - Allowed to use the image cache management API

To limit an action to a particular role or roles, you list the roles like so :

{
  "delete_image": ["role:admin", "role:superuser"]
}

The above would add a rule that only allowed users that had roles of either "admin" or "superuser" to delete an image.

Writing Rules

Role checks are going to continue to work exactly as they already do. If the role defined in the check is one that the user holds, then that will pass, e.g., role:admin.

To write a generic rule, you need to know that there are three values provided by Glance that can be used in a rule on the left side of the colon (:). Those values are the current user's credentials in the form of:

  • role
  • tenant
  • owner

The left side of the colon can also contain any value that Python can understand, e.g.,:

  • True
  • False
  • "a string"
  • &c.

Using tenant and owner will only work with images. Consider the following rule:

tenant:%(owner)s

This will use the tenant value of the currently authenticated user. It will also use owner from the image it is acting upon. If those two values are equivalent the check will pass. All attributes on an image (as well as extra image properties) are available for use on the right side of the colon. The most useful are the following:

  • owner
  • protected
  • is_public

Therefore, you could construct a set of rules like the following:

{
    "not_protected": "False:%(protected)s",
    "is_owner": "tenant:%(owner)s",
    "is_owner_or_admin": "rule:is_owner or role:admin",
    "not_protected_and_is_owner": "rule:not_protected and rule:is_owner",

    "get_image": "rule:is_owner_or_admin",
    "delete_image": "rule:not_protected_and_is_owner",
    "add_member": "rule:not_protected_and_is_owner"
}

Examples

Example 1. (The default policy configuration)

{
    "default": ""
}

Note that an empty JSON list means that all methods of the Glance API are callable by anyone.

Example 2. Disallow modification calls to non-admins

{
    "default": "",
    "add_image": "role:admin",
    "modify_image": "role:admin",
    "delete_image": "role:admin"
}