Syncing LDAP groups | Authentication and authorization

About configuring LDAP sync
Running LDAP sync
Running a group pruning job
Automatically syncing LDAP groups
LDAP group sync examples
LDAP sync configuration specification

As an administrator, you can use groups to manage users, change their permissions, and enhance collaboration. Your organization may have already created user groups and stored them in an LDAP server. OpenShift Container Platform can sync those LDAP records with internal OpenShift Container Platform records, enabling you to manage your groups in one place. OpenShift Container Platform currently supports group sync with LDAP servers using three common schemas for defining group membership: RFC 2307, Active Directory, and augmented Active Directory.

For more information on configuring LDAP, see Configuring an LDAP identity provider.

You must have cluster-admin privileges to sync groups.

About configuring LDAP sync

Before you can run LDAP sync, you need a sync configuration file. This file contains the following LDAP client configuration details:

Configuration for connecting to your LDAP server.
Sync configuration options that are dependent on the schema used in your LDAP server.
An administrator-defined list of name mappings that maps OpenShift Container Platform group names to groups in your LDAP server.

The format of the configuration file depends upon the schema you are using: RFC 2307, Active Directory, or augmented Active Directory.

LDAP client configuration: The LDAP client configuration section of the configuration defines the connections to your LDAP server.

The LDAP client configuration section of the configuration defines the connections to your LDAP server.

LDAP client configuration

url: ldap://10.0.0.0:389 (1)
bindDN: cn=admin,dc=example,dc=com (2)
bindPassword: password (3)
insecure: false (4)
ca: my-ldap-ca-bundle.crt (5)

1	The connection protocol, IP address of the LDAP server hosting your database, and the port to connect to, formatted as `scheme://host:port`.
2	Optional distinguished name (DN) to use as the Bind DN. OpenShift Container Platform uses this if elevated privilege is required to retrieve entries for the sync operation.
3	Optional password to use to bind. OpenShift Container Platform uses this if elevated privilege is necessary to retrieve entries for the sync operation. This value may also be provided in an environment variable, external file, or encrypted file.
4	When `false`, secure LDAP (`ldaps://`) URLs connect using TLS, and insecure LDAP (`ldap://`) URLs are upgraded to TLS. When `true`, no TLS connection is made to the server and you cannot use `ldaps://` URL schemes.
5	The certificate bundle to use for validating server certificates for the configured URL. If empty, OpenShift Container Platform uses system-trusted roots. This only applies if `insecure` is set to `false`.

LDAP query definition: Sync configurations consist of LDAP query definitions for the entries that are required for synchronization. The specific definition of an LDAP query depends on the schema used to store membership information in the LDAP server.

LDAP query definition

baseDN: ou=users,dc=example,dc=com (1)
scope: sub (2)
derefAliases: never (3)
timeout: 0 (4)
filter: (objectClass=person) (5)
pageSize: 0 (6)

1	The distinguished name (DN) of the branch of the directory where all searches will start from. It is required that you specify the top of your directory tree, but you can also specify a subtree in the directory.
2	The scope of the search. Valid values are `base`, `one`, or `sub`. If this is left undefined, then a scope of `sub` is assumed. Descriptions of the scope options can be found in the table below.
3	The behavior of the search with respect to aliases in the LDAP tree. Valid values are `never`, `search`, `base`, or `always`. If this is left undefined, then the default is to `always` dereference aliases. Descriptions of the dereferencing behaviors can be found in the table below.
4	The time limit allowed for the search by the client, in seconds. A value of `0` imposes no client-side limit.
5	A valid LDAP search filter. If this is left undefined, then the default is `(objectClass=*)`.
6	The optional maximum size of response pages from the server, measured in LDAP entries. If set to `0`, no size restrictions will be made on pages of responses. Setting paging sizes is necessary when queries return more entries than the client or server allow by default.

Table 1. LDAP search scope options
LDAP search scope	Description
`base`	Only consider the object specified by the base DN given for the query.
`one`	Consider all of the objects on the same level in the tree as the base DN for the query.
`sub`	Consider the entire subtree rooted at the base DN given for the query.

Table 2. LDAP dereferencing behaviors
Dereferencing behavior	Description
`never`	Never dereference any aliases found in the LDAP tree.
`search`	Only dereference aliases found while searching.
`base`	Only dereference aliases while finding the base object.
`always`	Always dereference all aliases found in the LDAP tree.

User-defined name mapping: A user-defined name mapping explicitly maps the names of OpenShift Container Platform groups to unique identifiers that find groups on your LDAP server. The mapping uses normal YAML syntax. A user-defined mapping can contain an entry for every group in your LDAP server or only a subset of those groups. If there are groups on the LDAP server that do not have a user-defined name mapping, the default behavior during sync is to use the attribute specified as the OpenShift Container Platform group’s name.

User-defined name mapping

groupUIDNameMapping:
  "cn=group1,ou=groups,dc=example,dc=com": firstgroup
  "cn=group2,ou=groups,dc=example,dc=com": secondgroup
  "cn=group3,ou=groups,dc=example,dc=com": thirdgroup

About the RFC 2307 configuration file

The RFC 2307 schema requires you to provide an LDAP query definition for both user and group entries, as well as the attributes with which to represent them in the internal OpenShift Container Platform records.

For clarity, the group you create in OpenShift Container Platform should use attributes other than the distinguished name whenever possible for user- or administrator-facing fields. For example, identify the users of an OpenShift Container Platform group by their e-mail, and use the name of the group as the common name. The following configuration file creates these relationships:

If using user-defined name mappings, your configuration file will differ.

LDAP sync configuration that uses RFC 2307 schema: rfc2307_config.yaml

kind: LDAPSyncConfig
apiVersion: v1
url: ldap://LDAP_SERVICE_IP:389 (1)
insecure: false (2)
rfc2307:
    groupsQuery:
        baseDN: "ou=groups,dc=example,dc=com"
        scope: sub
        derefAliases: never
        pageSize: 0
    groupUIDAttribute: dn (3)
    groupNameAttributes: [ cn ] (4)
    groupMembershipAttributes: [ member ] (5)
    usersQuery:
        baseDN: "ou=users,dc=example,dc=com"
        scope: sub
        derefAliases: never
        pageSize: 0
    userUIDAttribute: dn (6)
    userNameAttributes: [ mail ] (7)
    tolerateMemberNotFoundErrors: false
    tolerateMemberOutOfScopeErrors: false

1	The IP address and host of the LDAP server where this group’s record is stored.
2	When `false`, secure LDAP (`ldaps://`) URLs connect using TLS, and insecure LDAP (`ldap://`) URLs are upgraded to TLS. When `true`, no TLS connection is made to the server and you cannot use `ldaps://` URL schemes.
3	The attribute that uniquely identifies a group on the LDAP server. You cannot specify `groupsQuery` filters when using DN for `groupUIDAttribute`. For fine-grained filtering, use the whitelist / blacklist method.
4	The attribute to use as the name of the group.
5	The attribute on the group that stores the membership information.
6	The attribute that uniquely identifies a user on the LDAP server. You cannot specify `usersQuery` filters when using DN for userUIDAttribute. For fine-grained filtering, use the whitelist / blacklist method.
7	The attribute to use as the name of the user in the OpenShift Container Platform group record.

About the Active Directory configuration file

The Active Directory schema requires you to provide an LDAP query definition for user entries, as well as the attributes to represent them with in the internal OpenShift Container Platform group records.

For clarity, the group you create in OpenShift Container Platform should use attributes other than the distinguished name whenever possible for user- or administrator-facing fields. For example, identify the users of an OpenShift Container Platform group by their e-mail, but define the name of the group by the name of the group on the LDAP server. The following configuration file creates these relationships:

LDAP sync configuration that uses Active Directory schema: active_directory_config.yaml

kind: LDAPSyncConfig
apiVersion: v1
url: ldap://LDAP_SERVICE_IP:389
activeDirectory:
    usersQuery:
        baseDN: "ou=users,dc=example,dc=com"
        scope: sub
        derefAliases: never
        filter: (objectclass=person)
        pageSize: 0
    userNameAttributes: [ mail ] (1)
    groupMembershipAttributes: [ memberOf ] (2)

1	The attribute to use as the name of the user in the OpenShift Container Platform group record.
2	The attribute on the user that stores the membership information.

About the augmented Active Directory configuration file

The augmented Active Directory schema requires you to provide an LDAP query definition for both user entries and group entries, as well as the attributes with which to represent them in the internal OpenShift Container Platform group records.

LDAP sync configuration that uses augmented Active Directory schema: augmented_active_directory_config.yaml

kind: LDAPSyncConfig
apiVersion: v1
url: ldap://LDAP_SERVICE_IP:389
augmentedActiveDirectory:
    groupsQuery:
        baseDN: "ou=groups,dc=example,dc=com"
        scope: sub
        derefAliases: never
        pageSize: 0
    groupUIDAttribute: dn (1)
    groupNameAttributes: [ cn ] (2)
    usersQuery:
        baseDN: "ou=users,dc=example,dc=com"
        scope: sub
        derefAliases: never
        filter: (objectclass=person)
        pageSize: 0
    userNameAttributes: [ mail ] (3)
    groupMembershipAttributes: [ memberOf ] (4)

1	The attribute that uniquely identifies a group on the LDAP server. You cannot specify `groupsQuery` filters when using DN for groupUIDAttribute. For fine-grained filtering, use the whitelist / blacklist method.
2	The attribute to use as the name of the group.
3	The attribute to use as the name of the user in the OpenShift Container Platform group record.
4	The attribute on the user that stores the membership information.

Running LDAP sync

Once you have created a sync configuration file, you can begin to sync. OpenShift Container Platform allows administrators to perform a number of different sync types with the same server.

Syncing the LDAP server with OpenShift Container Platform

You can sync all groups from the LDAP server with OpenShift Container Platform.

Prerequisites

Create a sync configuration file.

Procedure

To sync all groups from the LDAP server with OpenShift Container Platform:

$ oc adm groups sync --sync-config=config.yaml --confirm

By default, all group synchronization operations are dry-run, so you must set the --confirm flag on the oc adm groups sync command to make changes to OpenShift Container Platform group records.

Syncing OpenShift Container Platform groups with the LDAP server

You can sync all groups already in OpenShift Container Platform that correspond to groups in the LDAP server specified in the configuration file.

Prerequisites

Create a sync configuration file.

Procedure

To sync OpenShift Container Platform groups with the LDAP server:

$ oc adm groups sync --type=openshift --sync-config=config.yaml --confirm

By default, all group synchronization operations are dry-run, so you must set the --confirm flag on the oc adm groups sync command to make changes to OpenShift Container Platform group records.

Syncing subgroups from the LDAP server with OpenShift Container Platform

You can sync a subset of LDAP groups with OpenShift Container Platform using whitelist files, blacklist files, or both.

You can use any combination of blacklist files, whitelist files, or whitelist literals. Whitelist and blacklist files must contain one unique group identifier per line, and you can include whitelist literals directly in the command itself. These guidelines apply to groups found on LDAP servers as well as groups already present in OpenShift Container Platform.

Prerequisites

Create a sync configuration file.

Procedure

To sync a subset of LDAP groups with OpenShift Container Platform, use any the following commands:

$ oc adm groups sync --whitelist=<whitelist_file> \
                   --sync-config=config.yaml      \
                   --confirm

$ oc adm groups sync --blacklist=<blacklist_file> \
                   --sync-config=config.yaml      \
                   --confirm

$ oc adm groups sync <group_unique_identifier>    \
                   --sync-config=config.yaml      \
                   --confirm

$ oc adm groups sync <group_unique_identifier>  \
                   --whitelist=<whitelist_file> \
                   --blacklist=<blacklist_file> \
                   --sync-config=config.yaml    \
                   --confirm

$ oc adm groups sync --type=openshift           \
                   --whitelist=<whitelist_file> \
                   --sync-config=config.yaml    \
                   --confirm

By default, all group synchronization operations are dry-run, so you must set the --confirm flag on the oc adm groups sync command to make changes to OpenShift Container Platform group records.

Running a group pruning job

An administrator can also choose to remove groups from OpenShift Container Platform records if the records on the LDAP server that created them are no longer present. The prune job will accept the same sync configuration file and whitelists or blacklists as used for the sync job.

For example:

$ oc adm prune groups --sync-config=/path/to/ldap-sync-config.yaml --confirm

$ oc adm prune groups --whitelist=/path/to/whitelist.txt --sync-config=/path/to/ldap-sync-config.yaml --confirm

$ oc adm prune groups --blacklist=/path/to/blacklist.txt --sync-config=/path/to/ldap-sync-config.yaml --confirm

Automatically syncing LDAP groups

You can automatically sync LDAP groups on a periodic basis by configuring a cron job.

Prerequisites

You have access to the cluster as a user with the cluster-admin role.
You have configured an LDAP identity provider (IDP).

This procedure assumes that you created an LDAP secret named ldap-secret and a config map named ca-config-map.

Procedure

Create a project where the cron job will run:
```
$ oc new-project ldap-sync (1)
```
1 This procedure uses a project called ldap-sync.
Locate the secret and config map that you created when configuring the LDAP identity provider and copy them to this new project.

The secret and config map exist in the openshift-config project and must be copied to the new ldap-sync project.

Define a service account:

Example ldap-sync-service-account.yaml

kind: ServiceAccount
apiVersion: v1
metadata:
  name: ldap-group-syncer
  namespace: ldap-sync

Create the service account:

$ oc create -f ldap-sync-service-account.yaml

Define a cluster role:

Example ldap-sync-cluster-role.yaml

apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: ldap-group-syncer
rules:
  - apiGroups:
      - ''
      - user.openshift.io
    resources:
      - groups
    verbs:
      - get
      - list
      - create
      - update

Create the cluster role:

$ oc create -f ldap-sync-cluster-role.yaml

Define a cluster role binding to bind the cluster role to the service account:

Example ldap-sync-cluster-role-binding.yaml

kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: ldap-group-syncer
subjects:
  - kind: ServiceAccount
    name: ldap-group-syncer              (1)
    namespace: ldap-sync
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: ldap-group-syncer                (2)

1	Reference to the service account created earlier in this procedure.
2	Reference to the cluster role created earlier in this procedure.

Create the cluster role binding:

$ oc create -f ldap-sync-cluster-role-binding.yaml

Define a config map that specifies the sync configuration file:

Example ldap-sync-config-map.yaml

kind: ConfigMap
apiVersion: v1
metadata:
  name: ldap-group-syncer
  namespace: ldap-sync
data:
  sync.yaml: |                                 (1)
    kind: LDAPSyncConfig
    apiVersion: v1
    url: ldaps://10.0.0.0:389                  (2)
    insecure: false
    bindDN: cn=admin,dc=example,dc=com         (3)
    bindPassword:
      file: "/etc/secrets/bindPassword"
    ca: /etc/ldap-ca/ca.crt
    rfc2307:                                   (4)
      groupsQuery:
        baseDN: "ou=groups,dc=example,dc=com"  (5)
        scope: sub
        filter: "(objectClass=groupOfMembers)"
        derefAliases: never
        pageSize: 0
      groupUIDAttribute: dn
      groupNameAttributes: [ cn ]
      groupMembershipAttributes: [ member ]
      usersQuery:
        baseDN: "ou=users,dc=example,dc=com"   (6)
        scope: sub
        derefAliases: never
        pageSize: 0
      userUIDAttribute: dn
      userNameAttributes: [ uid ]
      tolerateMemberNotFoundErrors: false
      tolerateMemberOutOfScopeErrors: false

1	Define the sync configuration file.
2	Specify the URL.
3	Specify the `bindDN`.
4	This example uses the RFC2307 schema; adjust values as necessary. You can also use a different schema.
5	Specify the `baseDN` for `groupsQuery`.
6	Specify the `baseDN` for `usersQuery`.

Create the config map:

$ oc create -f ldap-sync-config-map.yaml

Define a cron job:

Example ldap-sync-cron-job.yaml

kind: CronJob
apiVersion: batch/v1
metadata:
  name: ldap-group-syncer
  namespace: ldap-sync
spec:                                                                                (1)
  schedule: "*/30 * * * *"                                                           (2)
  concurrencyPolicy: Forbid
  jobTemplate:
    spec:
      backoffLimit: 0
      template:
        spec:
          containers:
            - name: ldap-group-sync
              image: "registry.redhat.io/openshift4/ose-cli:latest"
              command:
                - "/bin/bash"
                - "-c"
                - "oc adm groups sync --sync-config=/etc/config/sync.yaml --confirm" (3)
              volumeMounts:
                - mountPath: "/etc/config"
                  name: "ldap-sync-volume"
                - mountPath: "/etc/secrets"
                  name: "ldap-bind-password"
                - mountPath: "/etc/ldap-ca"
                  name: "ldap-ca"
          volumes:
            - name: "ldap-sync-volume"
              configMap:
                name: "ldap-group-syncer"
            - name: "ldap-bind-password"
              secret:
                secretName: "ldap-secret"                                            (4)
            - name: "ldap-ca"
              configMap:
                name: "ca-config-map"                                                (5)
          restartPolicy: "Never"
          terminationGracePeriodSeconds: 30
          activeDeadlineSeconds: 500
          dnsPolicy: "ClusterFirst"
          serviceAccountName: "ldap-group-syncer"

1	Configure the settings for the cron job. See "Creating cron jobs" for more information on cron job settings.
2	The schedule for the job specified in cron format. This example cron job runs every 30 minutes. Adjust the frequency as necessary, making sure to take into account how long the sync takes to run.
3	The LDAP sync command for the cron job to run. Passes in the sync configuration file that was defined in the config map.
4	This secret was created when the LDAP IDP was configured.
5	This config map was created when the LDAP IDP was configured.

Create the cron job:
```
$ oc create -f ldap-sync-cron-job.yaml
```

Additional resources

LDAP group sync examples

This section contains examples for the RFC 2307, Active Directory, and augmented Active Directory schemas.

These examples assume that all users are direct members of their respective groups. Specifically, no groups have other groups as members. See the Nested Membership Sync Example for information on how to sync nested groups.

Syncing groups using the RFC 2307 schema

For the RFC 2307 schema, the following examples synchronize a group named admins that has two members: Jane and Jim. The examples explain:

How the group and users are added to the LDAP server.
What the resulting group record in OpenShift Container Platform will be after synchronization.

In the RFC 2307 schema, both users (Jane and Jim) and groups exist on the LDAP server as first-class entries, and group membership is stored in attributes on the group. The following snippet of ldif defines the users and group for this schema:

LDAP entries that use RFC 2307 schema: rfc2307.ldif

  dn: ou=users,dc=example,dc=com
  objectClass: organizationalUnit
  ou: users
  dn: cn=Jane,ou=users,dc=example,dc=com
  objectClass: person
  objectClass: organizationalPerson
  objectClass: inetOrgPerson
  cn: Jane
  sn: Smith
  displayName: Jane Smith
  mail: jane.smith@example.com
  dn: cn=Jim,ou=users,dc=example,dc=com
  objectClass: person
  objectClass: organizationalPerson
  objectClass: inetOrgPerson
  cn: Jim
  sn: Adams
  displayName: Jim Adams
  mail: jim.adams@example.com
  dn: ou=groups,dc=example,dc=com
  objectClass: organizationalUnit
  ou: groups
  dn: cn=admins,ou=groups,dc=example,dc=com (1)
  objectClass: groupOfNames
  cn: admins
  owner: cn=admin,dc=example,dc=com
  description: System Administrators
  member: cn=Jane,ou=users,dc=example,dc=com (2)
  member: cn=Jim,ou=users,dc=example,dc=com

1	The group is a first-class entry in the LDAP server.
2	Members of a group are listed with an identifying reference as attributes on the group.

Prerequisites

Create the configuration file.

Procedure

Run the sync with the rfc2307_config.yaml file:

$ oc adm groups sync --sync-config=rfc2307_config.yaml --confirm

OpenShift Container Platform creates the following group record as a result of the above sync operation:

OpenShift Container Platform group created by using the rfc2307_config.yaml file

apiVersion: user.openshift.io/v1
kind: Group
metadata:
  annotations:
    openshift.io/ldap.sync-time: 2015-10-13T10:08:38-0400 (1)
    openshift.io/ldap.uid: cn=admins,ou=groups,dc=example,dc=com (2)
    openshift.io/ldap.url: LDAP_SERVER_IP:389 (3)
  creationTimestamp:
  name: admins (4)
users: (5)
- jane.smith@example.com
- jim.adams@example.com

1	The last time this OpenShift Container Platform group was synchronized with the LDAP server, in ISO 6801 format.
2	The unique identifier for the group on the LDAP server.
3	The IP address and host of the LDAP server where this group’s record is stored.
4	The name of the group as specified by the sync file.
5	The users that are members of the group, named as specified by the sync file.

Syncing groups using the RFC2307 schema with user-defined name mappings

When syncing groups with user-defined name mappings, the configuration file changes to contain these mappings as shown below.

LDAP sync configuration that uses RFC 2307 schema with user-defined name mappings: rfc2307_config_user_defined.yaml

kind: LDAPSyncConfig
apiVersion: v1
groupUIDNameMapping:
  "cn=admins,ou=groups,dc=example,dc=com": Administrators (1)
rfc2307:
    groupsQuery:
        baseDN: "ou=groups,dc=example,dc=com"
        scope: sub
        derefAliases: never
        pageSize: 0
    groupUIDAttribute: dn (2)
    groupNameAttributes: [ cn ] (3)
    groupMembershipAttributes: [ member ]
    usersQuery:
        baseDN: "ou=users,dc=example,dc=com"
        scope: sub
        derefAliases: never
        pageSize: 0
    userUIDAttribute: dn (4)
    userNameAttributes: [ mail ]
    tolerateMemberNotFoundErrors: false
    tolerateMemberOutOfScopeErrors: false

1	The user-defined name mapping.
2	The unique identifier attribute that is used for the keys in the user-defined name mapping. You cannot specify `groupsQuery` filters when using DN for groupUIDAttribute. For fine-grained filtering, use the whitelist / blacklist method.
3	The attribute to name OpenShift Container Platform groups with if their unique identifier is not in the user-defined name mapping.
4	The attribute that uniquely identifies a user on the LDAP server. You cannot specify `usersQuery` filters when using DN for userUIDAttribute. For fine-grained filtering, use the whitelist / blacklist method.

Prerequisites

Create the configuration file.

Procedure

Run the sync with the rfc2307_config_user_defined.yaml file:

$ oc adm groups sync --sync-config=rfc2307_config_user_defined.yaml --confirm

OpenShift Container Platform creates the following group record as a result of the above sync operation:

OpenShift Container Platform group created by using the rfc2307_config_user_defined.yaml file

apiVersion: user.openshift.io/v1
kind: Group
metadata:
  annotations:
    openshift.io/ldap.sync-time: 2015-10-13T10:08:38-0400
    openshift.io/ldap.uid: cn=admins,ou=groups,dc=example,dc=com
    openshift.io/ldap.url: LDAP_SERVER_IP:389
  creationTimestamp:
  name: Administrators (1)
users:
- jane.smith@example.com
- jim.adams@example.com

1	The name of the group as specified by the user-defined name mapping.

Syncing groups using RFC 2307 with user-defined error tolerances

By default, if the groups being synced contain members whose entries are outside of the scope defined in the member query, the group sync fails with an error:

Error determining LDAP group membership for "<group>": membership lookup for user "<user>" in group "<group>" failed because of "search for entry with dn="<user-dn>" would search outside of the base dn specified (dn="<base-dn>")".

This often indicates a misconfigured baseDN in the usersQuery field. However, in cases where the baseDN intentionally does not contain some of the members of the group, setting tolerateMemberOutOfScopeErrors: true allows the group sync to continue. Out of scope members will be ignored.

Similarly, when the group sync process fails to locate a member for a group, it fails outright with errors:

Error determining LDAP group membership for "<group>": membership lookup for user "<user>" in group "<group>" failed because of "search for entry with base dn="<user-dn>" refers to a non-existent entry".
Error determining LDAP group membership for "<group>": membership lookup for user "<user>" in group "<group>" failed because of "search for entry with base dn="<user-dn>" and filter "<filter>" did not return any results".

This often indicates a