Updated on 2024-04-29 GMT+08:00

Process

Unified permission governance allows you to configure access permissions for the databases, tables, and fields in MRS, DLI, and GaussDB(DWS). It has the following features:

  • Centralized access control

    Permissions of different big data services, such as MRS, DLI, and GaussDB(DWS), are centrally managed. A unified portal is available for you to configure and maintain permissions easily.

  • Multi-level permission configuration model

    Permission models are clearly defined and managed by level. A permission set or role further splits the permission scope defined by the workspace permission set and associates users with permissions for permission control.

  • Refined permission management

    Role-based access control (RBAC) on the console supports refined data permission configuration and permission assignment by role, user, and user group. In addition, on-demand and efficient permission application approval is supported. Approved permissions take effect immediately.

  • Multi-dimensional permission display
    • By workspace member: You can display the data table permissions requested by each user or user group and display, configure, and revoke the permission set relationship of each user or user group.
    • By data: You can display and configure the permission relationships of data in the current permission set by database, table, or field.
    • By permission: You can display, configure, and revoke the permission policy relationships of data in the current permission set by permission policy.
  • Workspace resource management

    In addition to data permissions, workspace resources, such as data connections and agencies, can be managed.

Use Process

Figure 1 shows the process for using unified permission governance.

Figure 1 Process of using unified permission governance

Unified permission governance supports data permission management, service resource management, and Ranger permission management. Their processes are as follows:

Data permission management process
  1. Make preparations.

    Before configuring data permission management, configure permissions for agencies and check prerequisites based on Preparation 1 and Preparation 2 (DLI permission management only requires Preparation 1).

  2. Synchronize users.

    Synchronize user information from IAM to data sources so that users' access to the data sources can be managed based on user information.

  3. Configure workspace permission sets.

    As the largest permission set in a DataArts Studio workspace, the workspace permission set defines the resources that can be accessed by users in the workspace.

  4. Configure permission sets.

    A permission set associates users with permissions. You can create multiple permission sets to associate users in different scenarios with different permissions. Permissions can be managed through permission synchronization (association of permission sets with roles are more recommended in actual applications.)

  5. Configure common roles.

    Create roles in the data source to bear the association between users and permissions. In this way, you can manage permissions more intuitively.

  6. Configure managed roles.

    Manage the existing roles in the MRS data source and inherit the MRS data source permissions of the existing roles.

  7. Manage members.

    DataArts Security allows you to view the permissions of workspace members, and manage roles and permission sets.

  8. Apply for permissions.

    During access permission management, you can grant permissions to users through permission sets or roles, or apply for permissions and approve permission applications.

  9. Review permission requests.

    The approver is the administrator of the permission set or role. The requested permission takes effect immediately after being approved.

  10. Enable permission applications.

    After permission applications are configured, data sources no longer use the accounts of the data connections during script execution and job test in DataArts Factory of DataArts Studio. Instead, the current user is used for authentication. In this way, different users have different data permissions, and the permissions of roles, permission sets, and queues can be managed.

Service resource management process
  1. Configure queue permissions.
    Queue permissions can be used to allocate MRS Yarn and DLI queues to the current workspace and configure queue permission policies for user groups or users.
    • After queues are allocated to the workspace, they can be selected during the job node configuration in DataArts Factory.
    • After queue permission policies are configured for user groups or users, they have the permissions specified in the policies.
  2. Configure workspace resource permission policies.

    DataArts Security supports management of workspace resources, such as data connections and agencies. Unauthorized users cannot view or use the resources.

Ranger permission management process
  1. Configure resource permissions.

    You can create permission policies for MRS components and use the Ranger component to manage permissions.

  2. View permission reports.

    You can view resource permission policies and their details through a comprehensive permission report.

Data Permission Management

The current data permission control uses the allowlist mechanism, which adds operation conditions to the users to be authorized without affecting the permissions the users already have. If you only want to make the permissions granted by the data permission control take effect, you need to revoke the original permissions of the users to be authorized. By default, DataArts Studio users have the following permissions:
  • For DLI data sources, the DAYU Administrator or DAYU User has the DLI Service Admin permission by default. Therefore, the users to be authorized have all the data permissions of DLI database tables by default. To remove the default permissions of an authorized user, you need to delete the DLI Service Admin permission of the user.
  • For GaussDB(DWS) data sources, even if the DAYU Administrator or DAYU User has the GaussDB(DWS) Administrator permission by default, the GaussDB(DWS) database permissions are isolated from the IAM permissions on the console. Therefore, the users to be authorized do not have the data permissions of GaussDB(DWS) database tables by default. Only the data permission granted by the current data permission control takes effect.
  • For MRS data sources, the DAYU Administrator or DAYU User has the MRS Administrator permission by default. After synchronized to MRS, the user is assigned a role. For details, see Synchronizing IAM Users to MRS. Then the Ranger component enables permissions for the default policy. For details, see Configuring Component Permission Policies. Therefore, the users to be authorized have the data permissions of MRS Hive database tables by default. If you want to revoke the default permissions of the users to be authorized, perform the following operations to remove the public user group from the default system policies on the Ranger component:
    1. Log in to MRS Manager as user admin.
    2. On the Manager page, choose Cluster > Services > Ranger. On the Ranger overview page, click RangerAdmin to go to the Ranger WebUI.
      Figure 2 Accessing the Ranger WebUI

    3. Log out of the current account and use the Ranger administrator account to log in again. For a common cluster, the admin account for the Manager page can be used as the Ranger administrator account. For a security cluster, rangeradmin is the Ranger administrator account. For details about the default password of rangeradmin, see User Account List.
      Figure 3 Logging out of the current account
    4. On the home page, click the component plug-in name in the HADOOP SQL area, for example, Hive.
    5. On the Access tab page, locate the default policies whose Groups column contains public (that is, the policy whose value in the Default Policy column is True) and remove the public user group from the policies.
      Figure 4 Policy list