Horizon Policy Enforcement (RBAC: Role Based Access Control)

Introduction

Horizon’s policy enforcement builds on the oslo_policy engine. The basis of which is openstack_auth/policy.py. Services in OpenStack use the oslo policy engine to define policy rules to limit access to APIs based primarily on role grants and resource ownership.

The implementation in Horizon is based on copies of policy files found in the service’s source code.

The service rules files are loaded into the policy engine to determine access rights to actions and service APIs.

Horizon Settings

There are a few settings that must be in place for the Horizon policy engine to work.

  • POLICY_CHECK_FUNCTION

  • POLICY_DIRS

  • POLICY_FILES_PATH

  • POLICY_FILES

  • DEFAULT_POLICY_FILES

For more detail, see Settings Reference.

How user’s roles are determined

Each policy check uses information about the user stored on the request to determine the user’s roles. This information was extracted from the scoped token received from Keystone when authenticating.

Entity ownership is also a valid role. To verify access to specific entities like a project, the target must be specified. See the section rule targets later in this document.

How to Utilize RBAC

Django: Table action

The primary way to add role based access control checks to panels is in the definition of table actions. When implementing a derived action class, setting the policy_rules attribute to valid policy rules will force a policy check before the horizon.tables.Action.allowed() method is called on the action. These rules are defined in the policy files pointed to by POLICY_PATH and POLICY_FILES. The rules are role based, where entity owner is also a role. The format for the policy_rules is a list of two item tuples. The first component of the tuple is the scope of the policy rule, this is the service type. This informs the policy engine which policy file to reference. The second component is the rule to enforce from the policy file specified by the scope. An example tuple is:

("identity", "identity:get_user")

x tuples can be added to enforce x rules.

Note

If a rule specified is not found in the policy file, the policy check will return False and the action will not be allowed.

Django: policy check function

The secondary way to add a role based check is to directly use the check() method. The method takes a list of actions, same format as the policy_rules attribute detailed above; the current request object; and a dictionary of action targets. This is the method that horizon.tables.Action class utilizes. Examples look like:

from openstack_dashboard import policy

allowed = policy.check((("identity", "identity:get_user"),
                       ("identity", "identity:get_project"),), request)

can_see = policy.check((("identity", "identity:get_user"),), request,
                       target={"domain_id": domainId})

Note

Any time multiple rules are specified in a single policy.check method call, the result is the logical and of each rule check. So, if any rule fails verification, the result is False.

Angular: ifAllowed method

The third way to add a role based check is in javascript files. Use the method ‘ifAllowed()’ in file ‘openstack_dashboard.static.app.core.policy.service.js’. The method takes a list of actions, similar format with the policy_rules attribute detailed above. An Example looks like:

angular
.module('horizon.dashboard.identity.users')
.controller('identityUsersTableController', identityUsersTableController);

identityUsersTableController.$inject = [
  'horizon.app.core.openstack-service-api.policy',
];

function identityUsersTableController(toast, gettext, policy, keystone) {
  var rules = [['identity', 'identity:list_users']];
  policy.ifAllowed({ rules: rules }).then(policySuccess, policyFailed);
}

Angular: hz-if-policies

The fourth way to add a role based check is in html files. Use angular directive ‘hz-if-policies’ in file ‘openstack_dashboard/static/app/core/cloud-services/hz-if-policies.directive.js’. Assume you have the following policy defined in your angular controller:

ctrl.policy = { rules: [["identity", "identity:update_user"]] }

Then in your HTML, use it like so:

<div hz-if-policies='ctrl.policy'>
  <span>I am visible if the policy is allowed!</span>
</div>

Rule Targets

Some rules allow access if the user owns the entity. Policy check targets specify particular entities to check for user ownership. The target parameter to the check() method is a simple dictionary. For instance, the target for checking access a project looks like:

{"project_id": "0905760626534a74979afd3f4a9d67f1"}

If the value matches the project_id to which the user’s token is scoped, then access is allowed.

When deriving the horizon.tables.Action class for use in a table, if a policy check is desired for a particular target, the implementer should override the horizon.tables.Action.get_policy_target() method. This allows a programmatic way to specify the target based on the current datum. The value returned should be the target dictionary.

Policy-in-Code and deprecated rules

As the effort of policy-in-code, most OpenStack projects define their default policies in their codes. All projects (except swift) covered by horizon supports “policy-in-code”. (Note that swift is an exception as it has its own mechanism to control RBAC.)

“oslo.policy” provides a way to deprecate existing policy rules like renaming rule definitions (“check_str”) and renaming rule names. They are defined as part of python codes in back-end services. horizon cannot import python codes of back-end services, so we need a way to restore policies defined by “policy-in-code” including deprecated rules.

To address the above issue, horizon adopts the following two-step approach:

  • The first step scans policy-in-code of back-end services and and dump the loaded default policies into YAML files per service including information of deprecated rules. This step is executed as part of the development process per release cycle and these YAML files are shipped per release.

    Note that oslopolicy-sample-generator does not output deprecated rules in a structured way, so we prepare a dedicated script for this purpose in the horizon repo.

  • The horizon policy implementation loads the above YAML file into a list of RuleDefault and registers the list as the default rules to the policy enforcer. The default rules and operator-defined rules are maintained separately, so operators still can edit the policy files as oslo.policy does in back-end services.

This approach has the following merits:

  • All features supported by oslo.policy can be supported in horizon as default rules in back-end services are restored as-is. Horizon can evaluate deprecated rules.

  • The default rules and operator defined rules are maintained separately. Operators can use the same way to maintain policy files of back-end services.

The related files in the horizon codebase are:

  • openstack_dashboard/conf/<service>_policy.yaml: operator-defined policies. These files are generated by oslopolicy-sample-generator.

  • openstack_dashboard/conf/default_policies/<service>.yaml YAML files contain default policies.

  • openstack_dashboard/management/commands/dump_default_policies.py: This script scans policy-in-code of a specified namespace under oslo.policy.policies entrypoints and dump them into the YAML file under openstack_dashboard/conf/default_policies.

  • openstack_auth/policy.py: _load_default_rules function loads the YAML files with default rules and call register_defautls method of the policy enforcer per service.

Policy file maintenance

  • YAML files for default policies

    Run the following command after installing a corresponding project. You need to run it for keystone, nova, cinder, neutron, glance.

    python3 manage.py dump_default_policies \
  --namespace $PROJECT \
  --output-file openstack_dashboard/conf/default_policies/${PROJECT}.yaml

  • Sample policy files

    Run the following commands after installing a corresponding project. You need to run it for keystone, nova, cinder, neutron, glance.

    oslopolicy-sample-generator --namespace $PROJECT \
  --output-file openstack_dashboard/conf/${PROJECT}_policy.yaml
sed -i 's/^"/#"/' openstack_dashboard/conf/${PROJECT}_policy.yaml

    Note

    We now use YAML format for sample policy files now. “oslo.policy” can accept both YAML and JSON files. We now support default policies so there is no need to define all policies using JSON files. YAML files also allows us to use comments, so we can provide good sample policy files. This is the same motivation as the Wallaby community goal Migrate RBAC Policy Format from JSON to YAML.

    Note

    The second “sed” command is to comment out rules for rule renames. oslopolicy-sample-generator does not comment out them, but they are unnecessary in horizon usage. A single renaming rule can map to multiple rules, so it does not work as-is. In addition, they trigger deprecation warnings in horizon log if these sample files are used in horizon as-is. Thus, we comment them out by default.

After syncing policies from back-end services, you need to check what are changed. If a policy referred by horizon has been changed, you need to check and modify the horizon code base accordingly.

Warning

After the support of default policies, the following tool does not work. It is a future work to make it work again or evaluate the need itself.

To summarize which policies are removed or added, a convenient tool is provided:

$ cd openstack_dashboard/conf/
$ python ../../tools/policy-diff.py --help
usage: policy-diff.py [-h] --old OLD --new NEW [--mode {add,remove}]

optional arguments:
-h, --help           show this help message and exit
--old OLD            Current policy file
--new NEW            New policy file
--mode {add,remove}  Diffs to be shown

# Show removed policies
# The default is "--mode remove". You can omit --mode option.
$ python ../../tools/policy-diff.py \
    --old keystone_policy.json --new keystone_policy.json.new --mode remove
default
identity:change_password
identity:get_identity_provider