Secure your clusters with Active Directory

These steps show how you can secure your Elasticsearch clusters and Kibana instances with the Lightweight Directory Access Protocol (LDAP) using an Active Directory.

To learn more about how securing Elasticsearch clusters with Active Directory works, check Active Directory user authentication.

You can configure the deployment to authenticate users by communicating with an Active Directory Domain Controller. To integrate with Active Directory, you need to configure an active_directory realm and map Active Directory groups to user roles in Elasticsearch.

Contrary to the ldap realm, the active_directory realm only supports a user search mode, but you can choose whether to use a bind user.

The Active Directory realm authenticates users using an LDAP bind request. By default, all LDAP operations run as the authenticated user if you don’t specify a bind_dn. Alternatively, you can choose to configure your realm with a bind user.

1. Add your user settings for the active_directory realm as follows:

   xpack:
     security:
       authc:
         realms:
           active_directory:
             my_ad:
               order: 2 1
               domain_name: ad.example.com 2
               url: ldap://ad.example.com:389 3
   

   1. The order in which the active_directory realm is consulted during an authentication attempt.
   2. The primary domain in Active Directory. Binding to Active Directory fails if the domain name is not mapped in DNS.
   3. The LDAP URL pointing to the Active Directory Domain Controller that should handle authentication. If your Domain Controller is configured to use LDAP over TLS and it uses a self-signed certificate or a certificate that is signed by your organization’s CA, refer to using self-signed certificates.

You can choose to configure an Active Directory realm using a bind user. When you specify a bind_dn, this specific user is used to search for the Distinguished Name (DN) of the authenticating user based on the provided username and an LDAP attribute. If found, this user is authenticated by attempting to bind to the LDAP server using the found DN and the provided password.

1. Add your user settings for the active_directory realm as follows:

   Important

   You must apply the user settings to each deployment template.

   xpack:
     security:
       authc:
         realms:
           active_directory:
             my_ad:
               order: 2 1
               domain_name: ad.example.com 2
               url: ldap://ad.example.com:389 3
               bind_dn: es_svc_user@ad.example.com 4
   

   1. The order in which the active_directory realm is consulted during an authentication attempt.
   2. The primary domain in Active Directory. Binding to Active Directory fails if the domain name is not mapped in DNS.
   3. The LDAP URL pointing to the Active Directory Domain Controller that should handle authentication. If your Domain Controller is configured to use LDAP over TLS and it uses a self-signed certificate or a certificate that is signed by your organization’s CA, refer to using self-signed certificates.
   4. The user to run as for all Active Directory search requests.

2. Configure the password for the bind_dn user by adding the appropriate secure_bind_password setting to the Elasticsearch keystore.

   1. From the Deployments page, select your deployment.

      Narrow the list by name, ID, or choose from several other filters. To further define the list, use a combination of filters.

   2. From your deployment menu, select Security.

   3. Under the Elasticsearch keystore section, select Add settings.

   4. On the Create setting window, select the secret Type to be Secret String.

   5. Set the Setting name to xpack.security.authc.realms.active_directory.<my_ad>.secure_bind_password and add the password for the bind_dn user in the secret field.

      Warning

      After you configure secure_bind_password, any attempt to restart the deployment will fail until you complete the rest of the configuration steps. If you wish to rollback the Active Directory realm related configuration effort, you need to remove the xpack.security.authc.realms.active_directory.my_ad.secure_bind_password that was just added by clicking Remove by the setting name under Existing Keystores.

If your LDAP server uses a self-signed certificate or a certificate that is signed by your organization’s CA, you need to enable the deployment to trust this certificate. These steps are required only if TLS is enabled and the Active Directory controller is using self-signed certificates.

You’ll prepare a custom bundle that contains your certificate in the same way that you would on Elasticsearch Service. Custom bundles are extracted in the path /app/config/BUNDLE_DIRECTORY_STRUCTURE, where BUNDLE_DIRECTORY_STRUCTURE is the directory structure within the bundle ZIP file itself. For example:

$ tree .
.
└── adcert
      └── ca.crt

In the following example, the keystore file would be extracted to /app/config/adcert/ca.crt, where ca.crt is the name of the certificate.

xpack.security.authc.realms.active_directory.my_ad.ssl.truststore.path:
"/app/config/truststore/ca.p12"

1. Create a ZIP file that contains your CA certificate file, such as adcert.zip.

2. Update your plan in the advanced configuration editor so that it uses the bundle you prepared in the previous step. You need to modify the user_bundles JSON attribute similar to the following example:

   Note

   You must specify the user_bundles attribute for each deployment template. You can alter 7.* to 8.* when needed.

   {
   "cluster_name": "REPLACE_WITH_YOUR_CLUSTER_NAME",
   "plan": {
   
       ...
   
       "elasticsearch": {
         "version": "7.*",
         "user_bundles": [
           {
             "name": "adcert",
             "url": "https://www.myurl.com/adcert.zip", 1
             "elasticsearch_version": "7.*" 2
           }
         ]
       }
     }
   

   1. The URL that points to the adcert.zip file must be accessible to the cluster. Uploaded files are stored using Amazon’s highly-available S3 service.
   2. This bundle is compatible with any Elasticsearch 7.* version.::::{tip}
      Using a wildcard for the minor version ensures that the bundle is compatible with the specified Elasticsearch major version, and eliminates the need to upload a new bundle when upgrading to a new minor version.
      ::::

3. Update your user settings for the active_directory realm as follows:

   xpack:
     security:
       authc:
         realms:
           active_directory:
             my_ad:
               order: 2
               domain_name: ad.example.com
               url: /app/config/adcert/ca.crt 1
               bind_dn: es_svc_user@ad.example.com
               ssl:
                 certificate_authorities: ["/app/config/cacerts/ca.crt"]
   

   1. The ldaps URL pointing to the Active Directory Domain Controller.

   The ssl.verification_mode setting (not shown) indicates the type of verification to use when using ldaps to protect against man-in-the-middle attacks and certificate forgery. The value for this property defaults to full. When you configure Elasticsearch to connect to a Domain Controller using TLS, it attempts to verify the hostname or IP address specified by the url attribute in the realm configuration with the Subject Alternative Names (SAN) in the certificate. If the SAN values in the certificate and realm configuration don’t match, Elasticsearch does not allow a connection to the Domain Controller. You can disable this behavior by setting the ssl. verification_mode property to certificate.

You have two ways of mapping Active Directory groups to roles for your users. The preferred one is to use the role mapping API. If for some reason this is not possible, you can use a role mapping file to specify the mappings instead.

Let’s assume that you want all your users that authenticate through AD to have read-only access to a certain index my-index and the AD users that are members of the cn=administrators, dc=example, dc=com group in LDAP, to become superusers in your deployment:

1. Create the read-only role

   POST /_security/role/read-only-my-index 1
   {
   "indices": [
   {
   "names": [ "my-index" ],
   "privileges": [ "read" ]
   }
     ]
   }
   

   1. The name of the role.

2. Create the relevant role mapping rule for read only users

   POST /_security/role_mapping/ad-read-only 1
   {
   "enabled": true,
   "roles": [ "read-only-my-index" ], 2
   "rules": {
   "field": { "realm.name": "my_ad" } 3
   },
   "metadata": { "version": 1 }
   }
   

   1. The name of the role mapping.
   2. The name of the role we created earlier.
   3. The name of our Active Directory realm.

3. Create the relevant role mapping rule for superusers

   POST /_security/role_mapping/ldap-superuser 1
   {
   "enabled": true,
   "roles": [ "superuser" ], 2
   "rules": {
   "all" : [
   { "field": { "realm.name": "my_ad" } },<3>
   { "field": { "groups": "cn=administrators, dc=example, dc=com" } }<4>
       ]
   },
   "metadata": { "version": 1 }
   }
   

   1. The name of the role mapping.
   2. The name of the role we want to assign, in this case superuser.
   3. The name of our active_directory realm.
   4. The DN of the AD group whose members should get the superuser role in the deployment.

Let’s assume that you want all your users that authenticate through AD and are members of the cn=my-users,dc=example, dc=com group in AD to have read-only access to a certain index my-index and only the users cn=Senior Manager, cn=management, dc=example, dc=com and cn=Senior Admin, cn=management, dc=example, dc=com to become superusers in your deployment:

1. Create a file named role-mappings.yml with the following contents:

   superuser:
   - cn=Senior Manager, cn=management, dc=example, dc=com
   - cn=Senior Admin, cn=management, dc=example, dc=com
   read-only-user:
   - cn=my-users, dc=example, dc=com
   

2. Prepare the custom bundle ZIP file mappings.zip, that contains the role-mappings.yml file in the same way that you would on Elastic Cloud.

3. Custom bundles get unzipped under the path /app/config/BUNDLE_DIRECTORY_STRUCTURE, where BUNDLE_DIRECTORY_STRUCTURE is the directory structure within the bundle ZIP file itself. For example:

   $ tree .
   .
   └── mappings
         └── role-mappings.yml
   

   In our example, the role mappings file is extracted to /app/config/mappings/role-mappings.yml

4. Update your plan in the advanced configuration editor so that it uses the bundle you prepared in the previous step. Modify the user_bundles JSON attribute as shown in the following example:

   Note

   You must specify the user_bundles attribute for each deployment template. You can alter 7.* to 8.* when needed.

   {
   "cluster_name": "REPLACE_WITH_YOUR_CLUSTER_NAME",
   "plan": {
   
       ...
   
       "elasticsearch": {
         "version": "7.*",
         "user_bundles": [
           {
             "name": "role-mappings",
             "url": "https://www.myurl.com/mappings.zip", 1
             "elasticsearch_version": "7.*" 2
           }
         ]
       }
     }
   

   1. The URL that points to mappings.zip must be accessible to the cluster.
   2. The bundle is compatible with any Elasticsearch 7.* version.
   Tip

   Using a wildcard for the minor version ensures bundles are compatible with the stated Elasticsearch major version to avoid the need to re-upload a new bundle with minor versions upgrades.

5. Update your user settings for the ldap realm as follows:

   xpack:
     security:
       authc:
         realms:
           active_directory:
             my_ad:
               order: 2
               domain_name: ad.example.com
               url: ldaps://ad.example.com:636 1
               bind_dn: es_svc_user@ad.example.com
               ssl:
                 certificate_authorities: ["/app/config/cacerts/ca.crt"]
                 verification_mode: certificate
               files:
                 role_mapping: "/app/config/mappings/role-mappings.yml" 1
   

   1. The path where our role mappings file is unzipped.