The following sections will help you gain an understanding of how automation controller handles and lets you control file system security.
All playbooks are executed via the
awx file system user. For running jobs, automation controller offers job isolation via the use of Linux containers. This projection ensures jobs can only access playbooks, roles, and data from the Project directory for that job template.
For credential security, users may choose to upload locked SSH keys and set the unlock password to "ask". You can also choose to have the system prompt them for SSH credentials or sudo passwords rather than having the system store them in the database.
Automation controller's use of automation execution environments and Linux containers prevents playbooks from reading files outside of their project directory.
By default, the only data exposed to the ansible-playbook process inside the container is the current project being used.
You can customize this in the Job Settings and expose additional directories from the host into the container. Refer the next section, Isolation functionality and variables for more information.
Automation controller uses container technology to isolate jobs from each other. By default, only the current project is exposed to the container running a job template.
You may find that you need to customize your playbook runs to expose additional directories. To fine tune your usage of job isolation, there are certain variables that can be set.
By default, automation controller will use the system's
tmp directory (
/tmp by default) as its staging area. This can be changed in the Job Execution Path field of the Jobs settings screen, or in the REST API at
AWX_ISOLATION_BASE_PATH = "/opt/tmp"
If there are any additional directories that should specifically be exposed from the host to the container that playbooks run in, you can specify those in the Paths to Expose to Isolated Jobs
field of the Jobs setting scren, or in the REST API at
AWX_ISOLATION_SHOW_PATHS = ['/list/of/', '/paths']
The primary file you may want to add to
/var/lib/awx/.ssh, if your playbooks need to use keys or settings defined there.
The above fields can be found in the Jobs Settings window:
Role-Based Access Controls (RBAC) are built into automation controller and allow administrators to delegate access to server inventories, organizations, and more. Administrators can also centralize the management of various credentials, allowing end users to leverage a needed secret without ever exposing that secret to the end user. RBAC controls allow the controller to help you increase security and streamline management.
RBACs are easiest to think of in terms of Roles which define precisely who or what can see, change, or delete an "object" for which a specific capability is being set. RBAC is the practice of granting roles to users or teams.
There are a few main concepts that you should become familiar with regarding automation controller's RBAC design--roles, resources, and users. Users can be members of a role, which gives them certain access to any resources associated with that role, or any resources associated with "descendant" roles.
A role is essentially a collection of capabilities. Users are granted access to these capabilities and the controller's resources through the roles to which they are assigned or through roles inherited through the role hierarchy.
Roles associate a group of capabilities with a group of users. All capabilities are derived from membership within a role. Users receive capabilities only through the roles to which they are assigned or through roles they inherit through the role hierarchy. All members of a role have all capabilities granted to that role. Within an organization, roles are relatively stable, while users and capabilities are both numerous and may change rapidly. Users can have many roles.
Imagine that you have an organization named "SomeCompany" and want to allow two people, "Josie" and "Carter", access to manage all the settings associated with that organization. You should make both people members of the organization's
Often, you will have many Roles in a system and you will want some roles to include all of the capabilities of other roles. For example, you may want a System Administrator to have access to everything that an Organization Administrator has access to, who has everything that a Project Administrator has access to, and so on.
This concept is referred to as the 'Role Hierarchy':
Parent roles get all capabilities bestowed on any child roles
Members of roles automatically get all capabilities for the role they are a member of, as well as any child roles.
The Role Hierarchy is represented by allowing Roles to have "Parent Roles". Any capability that a Role has is implicitly granted to any parent roles (or parents of those parents, and so on).
Often, you will have many Roles in a system and you will want some roles to include all of the capabilities of other roles. For example, you may want a System Administrator to have access to everything that an Organization Administrator has access to, who has everything that a Project Administrator has access to, and so on. We refer to this concept as the 'Role Hierarchy' and it is represented by allowing Roles to have "Parent Roles". Any capability that a Role has is implicitly granted to any parent roles (or parents of those parents, and so on). Of course Roles can have more than one parent, and capabilities are implicitly granted to all parents.
RBAC controls also give you the capability to explicitly permit User and Teams of Users to run playbooks against certain sets of hosts. Users and teams are restricted to just the sets of playbooks and hosts to which they are granted capabilities. And, with automation controller, you can create or import as many Users and Teams as you require--create users and teams manually or import them from LDAP or Active Directory.
RBACs are easiest to think of in terms of who or what can see, change, or delete an "object" for which a specific capability is being determined.
The following sections cover how to apply automation controller's RBAC system in your environment.
When editing a user, a automation controller system administrator may specify the user as being either a System Administrator (also referred to as the Superuser) or a System Auditor.
System administrators implicitly inherit all capabilities for all objects (read/write/execute) within the automation controller environment.
System Auditors implicitly inherit the read-only capability for all objects within the automation controller environment.
When editing an organization, system administrators may specify the following roles:
One or more users as organization administrators
One or more users as organization auditors
And one or more users (or teams) as organization members
Users/teams that are members of an organization can view their organization administrator.
Users who are organization administrators implicitly inherit all capabilities for all objects within that automation controller organization.
Users who are organization auditors implicitly inherit the read-only capability for all objects within that automation controller organization.
When editing a project in an organization for which they are the administrator, system administrators and organization administrators may specify:
One or more users/teams that are project administrators
One or more users/teams that are project members
And one or more users/teams that may update the project from SCM, from among the users/teams that are members of that organization.
Users who are members of a project can view their project administrators.
Project administrators implicitly inherit the capability to update the project from SCM.
Administrators can also specify one or more users/teams (from those that are members of that project) that can use that project in a job template.
All access that is granted to use, read, or write credentials is handled through roles, which use automation controller's RBAC system to grant ownership, auditor, or usage roles.
System administrators and organization administrators may create inventories and credentials within organizations under their administrative capabilities.
Whether editing an inventory or a credential, System administrators and organization administrators may specify one or more users/teams (from those that are members of that organization) to be granted the usage capability for that inventory or credential.
System administrators and organization administrators may specify one or more users/teams (from those that are members of that organization) that have the capabilities to update (dynamic or manually) an inventory. Administrators can also execute ad hoc commands for an inventory.
System administrators, organization administrators, and project administrators, within a project under their administrative capabilities, may create and modify new job templates for that project.
When editing a job template, administrators (automation controller, organization, and project) can select among the inventory and credentials in the organization for which they have usage capabilities or they may leave those fields blank so that they will be selected at runtime.
Additionally, they may specify one or more users/teams (from those that are members of that project) that have execution capabilities for that job template. The execution capability is valid regardless of any explicit capabilities the user/team may have been granted against the inventory or credential specified in the job template.
A user can:
See any organization or project for which they are a member
Create their own credential objects which only belong to them
See and execute any job template for which they have been granted execution capabilities
If a job template that a user has been granted execution capabilities on does not specify an inventory or credential, the user will be prompted at run-time to select among the inventory and credentials in the organization they own or have been granted usage capabilities.
Users that are job template administrators can make changes to job templates; however, to change to the inventory, project, playbook, credentials, or instance groups used in the job template, the user must also have the "Use" role for the project and inventory currently being used or being set.
All access that is granted to use, read, or write credentials is handled through roles, and roles are defined for a resource.
The following table lists the RBAC system roles and a brief description of the how that role is defined with regard to privileges in automation controller.
What it can do
System Administrator - System wide singleton
Manages all aspects of the system
System Auditor - System wide singleton
Views all aspects of the system
Ad Hoc Role - Inventory
Runs ad hoc commands on an Inventory
Admin Role - Organizations, Teams, Inventory, Projects, Job Templates
Manages all aspects of a defined Organization, Team, Inventory, Project, or Job Template
Auditor Role - All
Views all aspects of a defined Organization, Project, Inventory, or Job Template
Execute Role - Job Templates
Runs assigned Job Template
Member Role - Organization, Team
Manages all of the settings associated with that Organization or Team
Read Role - Organizations, Teams, Inventory, Projects, Job Templates
Views all aspects of a defined Organization, Team, Inventory, Project, or Job Template
Update Role - Project
Updates the Project from the configured source control management system
Update Role - Inventory
Updates the Inventory using the cloud source update system
Owner Role - Credential
Owns and manages all aspects of this Credential
Use Role - Credential, Inventory, Project, IGs, CGs
Uses the Credential, Inventory, Project, IGs, or CGs in a Job Template
A Singleton Role is a special role that grants system-wide permissions. automation controller currently provides two built-in Singleton Roles but the ability to create or customize a Singleton Role is not supported at this time.
Automation controller support personnel typically works on ensuring that the controller is available and manages it a way to balance supportability and ease-of-use for users. Often, automation controller support will assign “Organization Owner/Admin” to users in order to allow them to create a new Organization and add members from their team the respective access needed. This minimizes supporting individuals and focuses more on maintaining uptime of the service and assisting users who are using automation controller.
Below are some common roles managed by the automation controller Organization:
Team Lead -
This user has the ability to control access for other users in their organization.
They can add/remove and grant users specific access to projects, inventories, and job templates.
This user also has the ability to create/remove/modify any aspect of an organization’s projects,
templates, inventories, teams, and credentials.
Security Engineer -
This account can view all aspects of the organization in read-only mode.
This may be good for a user who checks in and maintains compliance.
This might also be a good role for a service account who manages or
ships job data from automation controller to some other data collector.
All other users
These users by default as an organization member do not receive any access to any aspect
of the organization. In order to grant them access the respective organization owner needs
to add them to their respective team and grant them Admin, Execute, Use, Update, Ad-hoc
permissions to each component of the organization’s projects, inventories, and job templates.
Power users -
Organization Owners can provide “admin” through the team interface, over any component
of their organization including projects, inventories, and job templates. These users are able
to modify and utilize the respective component given access.
This will be the most common and allows the organization member the ability to execute
job templates and read permission to the specific components. This is permission applies to templates.
This permission applies to an organization’s credentials, inventories, and projects.
This permission allows the ability for a user to use the respective component within their job template.
This permission applies to projects. Allows the user to be able to run an SCM update on a project.
Organization “resource roles” functionality are specific to a certain resource type - such as workflows. Being a member of such a role usually provides two types of permissions, in the case of workflows, where a user is given a "workflow admin role" for the organization "Default":
this user can create new workflows in the organization "Default"
user can edit all workflows in the "Default" organization
One exception is job templates, where having the role is irrelevant of creation permission (more details on its own section).
Resource-specific organization roles are independent of the organization roles of admin and member. Having the "workflow admin role" for the "Default" organization will not allow a user to view all users in the organization, but having a "member" role in the "Default" organization will. The two types of roles are delegated independently of each other.
Users can edit fields not impacting job runs (non-sensitive fields) with a Job Template admin role alone. However, to edit fields that impact job runs in a job template, a user needs the following:
admin role to the job template and container groups
use role to related project
use role to related inventory
use role to related instance groups
An "organization job template admin" role was introduced, but having this role isn't sufficient by itself to edit a job template within the organization if the user does not have use role to the project / inventory / instance group or an admin role to the container group that a job template uses.
In order to delegate full job template control (within an organization) to a user or team, you will need grant the team or user all 3 organization-level roles:
job template admin
This will ensure that the user (or all users who are members of the team with these roles) have full access to modify job templates in the organization. If a job template uses an inventory or project from another organization, the user with these organization roles may still not have permission to modify that job template. For clarity of managing permissions, it is best-practice to not mix projects / inventories from different organizations.
Each role should have a content object, for instance, the org admin role has a content object of the org. To delegate a role, you need admin permission to the content object, with some exceptions that would result in you being able to reset a user's password.
Parent is the organization.
Allow is what this new permission will explicitly allow.
Scope is the parent resource that this new role will be created on. Example:
An assumption is being made that the creator of the resource should be given the admin role for that resource. If there are any instances where resource creation does not also imply resource administration, they will be explicitly called out.
Here are the rules associated with each admin type:
Allow: Create, read, update, delete any project
User Interface: Project Add Screen - Organizations
Parent: Org admin
Allow: Create, read, update, delete any inventory
User Interface: Inventory Add Screen - Organizations
As it is with the Use role, if you give a user Project Admin and Inventory Admin, it allows them to create Job Templates (not workflows) for your organization.
Parent: Org admin
Allow: Create, read, update, delete shared credentials
User Interface: Credential Add Screen - Organizations
Parent: Org admin
Allow: Assignment of notifications
Parent: Org admin
Allow: Create a workflow
Parent: Org admin
Allow: Executing JTs and WFJTs
The following is a sample scenario showing an organization with its roles and which resource(s) each have access to: