group-sync-operator

README

Group Sync Operator

Synchronizes groups from external providers into OpenShift

Overview

The OpenShift Container Platform contains functionality to synchronize groups found in external identity providers into the platform. Currently, the functionality that is included in OpenShift is limited to synchronizing LDAP only. This operator is designed to integrate with external providers in order to provide new solutions.

Group Synchronization is facilitated by creating a GroupSync resource. The following describes the high level schema for this resource:

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: example-groupsync
spec:
  providers:
    - <One or more providers to synchronize>

Deploying the Operator

Use the following steps to deploy the operator to an OpenShift cluster

1. Clone the project locally and changed into the project . .

git clone https://github.com/redhat-cop/group-sync-operator.git
cd group-sync-operator

1. Deploy the Operator

make deploy IMG=quay.io/redhat-cop/group-sync-operator:latest

Note: The make deploy command will execute the manifests target that will require additional build tools to be made available. This target can be skipped by including the -o manifests in the command above.

Authentication

In most cases, authentication details must be provided in order to communicate with providers. Authentication details are provider specific with regards to the required values. In supported providers, the secret can be referenced in the credentialsSecret by name and namespace where it has been created as shown below:

credentialsSecret:
  name: <secret_name>
  namespace: <secret_namespace>

Providers

Integration with external systems is made possible through a set of pluggable external providers. The following providers are currently supported:

• Azure
• GitHub
• GitLab
• LDAP
• Keycloak/Red Hat Single Sign On
• Okta

The following sections describe the configuration options available for each provider

Azure

Groups contained within Azure Active Directory can be synchronized into OpenShift. The following table describes the set of configuration options for the Azure provider:

Name	Description	Defaults	Required
authorityHost	Azure Active Directory Endpoint	https://login.microsoftonline.com	No
baseGroups	List of groups to start searching from instead of listing all groups in the directory		No
credentialsSecret	Name of the secret containing authentication details (See below)		Yes
filter	Graph API filter		No
groups	List of groups to filter against		No
userNameAttributes	Fields on a user record to use as the User Name	userPrincipalName	No
prune	Prune Whether to prune groups that are no longer in Azure	false	No

The following is an example of a minimal configuration that can be applied to integrate with a Azure provider:

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: azure-groupsync
spec:
  providers:
  - name: azure
    azure:
      credentialsSecret:
        name: azure-group-sync
        namespace: group-sync-operator

Authenticating to Azure

Authentication to Azure can be performed using Application Registration with access to query group information in Azure Active Directory.

The App Registration must be granted access to the following Microsoft Graph API's:

• Group.Read.All
• GroupMember.Read.All
• User.Read.All

A secret must be created in the same namespace that contains the GroupSync resource:

The following keys must be defined in the secret

• AZURE_TENANT_ID - Tenant ID
• AZURE_CLIENT_ID - Client ID
• AZURE_CLIENT_SECRET - Client Secret

The secret can be created by executing the following command:

oc create secret generic azure-group-sync --from-literal=AZURE_TENANT_ID=<AZURE_TENANT_ID> --from-literal=AZURE_CLIENT_ID=<AZURE_CLIENT_ID> --from-literal=AZURE_CLIENT_SECRET=<AZURE_CLIENT_SECRET>

GitHub

Teams stored within a GitHub organization can be synchronized into OpenShift. The following table describes the set of configuration options for the GitHub provider:

Name	Description	Defaults	Required
ca	Reference to a resource containing a SSL certificate to use for communication (See below)		No
caSecret	DEPRECATED Reference to a secret containing a SSL certificate to use for communication (See below)		No
credentialsSecret	Reference to a secret containing authentication details (See below)		Yes
insecure	Ignore SSL verification	false	No
organization	Organization to synchronize against		Yes
teams	List of teams to filter against		No
url	Base URL for the GitHub or GitHub Enterprise host (Must contain a trailing slash)		No
prune	Prune Whether to prune groups that are no longer in GitHub	false	No

The following is an example of a minimal configuration that can be applied to integrate with a GitHub provider:

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: github-groupsync
spec:
  providers:
  - name: github
    github:
      organization: ocp
      credentialsSecret:
        name: github-group-sync
        namespace: group-sync-operator

Authenticating to GitHub

Authentication to GitHub can be performed using an OAuth Personal Access Token or as an GitHub App, using a secret key and appId. The OAuth Personal Access Token needs to scope of admin:org/read:org. A secret must be created in the same namespace that contains the GroupSync resource:

OAuth

When using an OAuth token, the following key is required:

• token - OAuth token

The secret can be created by executing the following command:

oc create secret generic github-group-sync --from-literal=token=<token>

As a GitHub app

When authenticating as a Github App, the following keys are required:

• privateKey and appId

First create a GitHub app

In GitHub, go to developer-settings -> github apps.

• Create a new app, it does not need webhook callbacks.
• Generate a private-key and download it
• Under "permissions and events", the app will need read-only access to the "Members" permission in the "Organization" section. NOTE: If you enable mapByScimId, this permissions needs to be Read & Write, though the operator only does read-only operations. The reason for this is the use of the v4 graphql api-endpoint.
• Take note of the "App ID" as you need it for later.
• Install the app to your organization.

Create the secret

The secret can be created by executing the following command:

oc create secret generic github-group-sync --from-literal=appId=<theAppId> --from-file=privateKey=</path/to/thefile>

GitLab

Groups stored within a GitLab can be synchronized into OpenShift. The following table describes the set of configuration options for the GitLab provider:

Name	Description	Defaults	Required
ca	Reference to a resource containing a SSL certificate to use for communication (See below)		No
caSecret	DEPRECATED Reference to a secret containing a SSL certificate to use for communication (See below)		No
credentialsSecret	Reference to a secret containing authentication details (See below)		Yes
insecure	Ignore SSL verification	false	No
groups	List of groups to filter against		No
prune	Prune Whether to prune groups that are no longer in GitLab	false	No
scope	Scope for group synchronization. Options are one for one level or sub to include subgroups	sub	No
url	Base URL for the GitLab instance	https://gitlab.com	No

The following is an example of a minimal configuration that can be applied to integrate with a GitHub provider:

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: gitlab-groupsync
spec:
  providers:
  - name: gitlab
    gitlab:
      credentialsSecret:
        name: gitlab-group-sync
        namespace: group-sync-operator

Authenticating to GitLab

Authentication to GitLab can be performed using a Token or a Username and Password (Note: 2FA not supported). A secret must be created in the same namespace that contains the GroupSync resource:

When using an OAuth token, the following token types are supported:

• Personal Access Token
• OAuth Token
• Job Token

the following key is required:

• token - OAuth token

Optionally, the tokenType key can be specified to indicate the type of token being provided from the following values:

• OAuth - oauth
• Personal Access Token - personal
• Job Token - job

If no tokenType is provided, oauth is used by default

The secret can be created by executing the following command:

oc create secret generic gitlab-group-sync --from-literal=token=<token>

To specify a token type, such as a Personal Access Token, the following command can be executed:

oc create secret generic gitlab-group-sync --from-literal=token=<token> --from-literal=tokenType=personal

The following keys are required for username and password:

• username - Username for authenticating with GitLab
• password - Password for authenticating with GitLab

The secret can be created by executing the following command:

oc create secret generic gitlab-group-sync --from-literal=username=<username> --from-literal=password=<password>

LDAP

Groups stored within an LDAP server can be synchronized into OpenShift. The LDAP provider implements the included features of the Syncing LDAP groups feature and makes use of the libraries from the OpenShift Command Line tool to streamline the migration to this operator based implementation.

The configurations of the three primary schemas (rfc2307, activeDirectory and augmentedActiveDirectory) can be directly migrated as is without any modification.

Name	Description	Defaults	Required
ca	Reference to a resource containing a SSL certificate to use for communication (See below)		No
caSecret	DEPRECATED Reference to a secret containing a SSL certificate to use for communication (See below)		No
credentialsSecret	Reference to a secret containing authentication details (See below)		No
insecure	Ignore SSL verification	false	No
groupUIDNameMapping	User defined name mapping		No
rfc2307	Configuration using the rfc2307 schema		No
activeDirectory	Configuration using the activeDirectory schema		No
augmentedActiveDirectory	Configuration using the activeDirectory schema		No
url	Connection URL for the LDAP server	ldap://ldapserver:389	No
whitelist	Explicit list of groups to synchronize		No
blacklist	Explicit list of groups to not synchronize		No
prune	Prune Whether to prune groups that are no longer in LDAP	false	No

The following is an example using the rfc2307 schema:

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: ldap-groupsync
spec:
  providers:
  - ldap:
      credentialsSecret:
        name: ldap-group-sync
        namespace: group-sync-operator
      insecure: true
      rfc2307:
        groupMembershipAttributes:
        - member
        groupNameAttributes:
        - cn
        groupUIDAttribute: dn
        groupsQuery:
          baseDN: ou=Groups,dc=example,dc=com
          derefAliases: never
          filter: (objectClass=groupofnames)
          scope: sub
        tolerateMemberNotFoundErrors: true
        tolerateMemberOutOfScopeErrors: true
        userNameAttributes:
        - cn
        userUIDAttribute: dn
        usersQuery:
          baseDN: ou=Users,dc=example,dc=com
          derefAliases: never
          scope: sub
      url: ldap://ldapserver:389
    name: ldap

The examples provided in the OpenShift documented referenced previously can be used to construct the schemas for the other LDAP synchronization types.

Authenticating to LDAP

If authentication is required in order to communicate with the LDAP server, a secret should be created in the same namespace that contains the GroupSync resource. The following keys can be defined:

• username - Username (Bind DN) for authenticating with the LDAP server
• password - Password for authenticating with the LDAP server

The secret can be created by executing the following command:

oc create secret generic ldap-group-sync --from-literal=username=<username> --from-literal=password=<password>

Whitelists and Blacklists

Groups can be explicitly whitelisted or blacklisted in order to control the groups that are eligible to be synchronized into OpenShift. When running LDAP group synchronization using the command line, this configuration is referenced via separate files, but these are instead specified in the blacklist and whitelist properties as shown below:

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: ldap-groupsync
spec:
  providers:
  - ldap:
...
      whitelist:
      - cn=Online Corporate Banking,ou=Groups,dc=example,dc=com
...
    name: ldap

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: ldap-groupsync
spec:
  providers:
  - ldap:
...
      blacklist:
      - cn=Finance,ou=Groups,dc=example,dc=com
...
    name: ldap

Keycloak

Groups stored within Keycloak can be synchronized into OpenShift. The following table describes the set of configuration options for the Keycloak provider:

Name	Description	Defaults	Required
ca	Reference to a resource containing a SSL certificate to use for communication (See below)		No
caSecret	DEPRECATED Reference to a secret containing a SSL certificate to use for communication (See below)		No
credentialsSecret	Reference to a secret containing authentication details (See below)		Yes
groups	List of groups to filter against		No
insecure	Ignore SSL verification	false	No
loginRealm	Realm to authenticate against	master	No
realm	Realm to synchronize		Yes
scope	Scope for group synchronization. Options are one for one level or sub to include subgroups	sub	No
url	Base URL for the Keycloak server. Older versions (<17.0.0) including Red Hat SSO should include the context path /auth appended to the hostname		Yes
prune	Prune Whether to prune groups that are no longer in Keycloak	false	No

The following is an example of a minimal configuration that can be applied to integrate with a Keycloak provider:

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: keycloak-groupsync
spec:
  providers:
  - name: keycloak
    keycloak:
      realm: ocp
      credentialsSecret:
        name: keycloak-group-sync
        namespace: group-sync-operator
      url: https://keycloak-keycloak-operator.apps.openshift.com/auth

Authenticating to Keycloak

A user with rights to query for Keycloak groups must be available. The following permissions must be associated to the user:

• Password must be set (Temporary option unselected) on the Credentials tab
• On the Role Mappings tab, select master-realm or realm-management next to the Client Roles dropdown and then select query-groups, query-users, and view-users.

A secret must be created in the same namespace that contains the GroupSync resource. It must contain the following keys for the user previously created:

• username - Username for authenticating with Keycloak
• password - Password for authenticating with Keycloak

The secret can be created by executing the following command:

oc create secret generic keycloak-group-sync --from-literal=username=<username> --from-literal=password=<password>

Okta

Okta Groups assigned to Okta Applications can be synchronized into OpenShift. The developer docs for the Okta API that the Okta Syncer uses can be found here. The following table describes the set of configuration options for the Okta provider:

Name	Description	Defaults	Required
credentialsSecret	Reference to a secret containing authentication details (See below)	''	Yes
groups	List of groups to filter against	nil	No
url	Okta URL which can be found under the "Okta Domain" in your application settings (must contain the scheme and a trailing slash)	''	Yes
appId	Okta Application (Client) ID that is attached to the application groups you wish to sync	''	Yes
extractLoginUsername	Bool to determine if you should extract username from okta login	false	No
profileKey	Attribute field on Okta User Profile you would like to use as identity	'login'	No
groupLimit	Integer to set the maximum number of groups to retrieve from OKTA per request.	1000	No
prune	Prune Whether to prune groups that are no longer in OKTA	false	No

The following is an example of a minimal configuration that can be applied to integrate with an Okta provider:

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: okta-sync
spec:
  providers:
    - name: okta
      okta:
        credentialsSecret:
          name: okta-api-token
          namespace: group-sync-operator
        url: "https://example.okta.com/"
        appId: xxxxxxxxxxxxxxxxxxxx

Authenticating to Okta

A secret must be created in the same namespace as the group-sync-operator pod. It must contain the following key:

• okta-api-token - Okta API Token for interacting with Okta

The secret can be created by executing the following command:

oc create secret generic okta-api-token --from-literal=okta-api-token=<OKTA_API_TOKEN> -n group-sync-operator

Support for Additional Metadata (Beta)

Additional metadata based on Keycloak group are also added to the OpenShift groups as Annotations including:

• Parent/child relationship between groups and their subgroups
• Group attributes

CA Certificates

Several providers allow for certificates to be provided in either a ConfigMap or Secret to communicate securely to the target host through the use of a property called ca.

The certificate can be added to a Secret called keycloak-certs using the key ca.crt representing the certificate using the following command.

oc create secret generic keycloak-certs --from-file=ca.crt=<file>

An example of how the CA certificate can be added to the Keycloak provider is shown below:

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: keycloak-groupsync
spec:
  providers:
  - name: keycloak
    keycloak:
      realm: ocp
      credentialsSecret:
        name: keycloak-group-sync
        namespace: group-sync-operator
      ca:
        kind: Secret
        name: keycloak-certs
        namespace: group-sync-operator
        key: ca.crt
      url: https://keycloak-keycloak-operator.apps.openshift.com

Alteratively, a ConfigMap can be used instead instead of a Secret. This is useful when using the Certificate injection using Operators feature.

The following command can be used to create a ConfigMap containing the certificate:

oc create configmap keycloak-certs --from-file=ca.crt=<file>

An example of how the CA certificate can be added to the Keycloak provider is shown below:

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: keycloak-groupsync
spec:
  providers:
  - name: keycloak
    keycloak:
      realm: ocp
      credentialsSecret:
        name: keycloak-group-sync
        namespace: group-sync-operator
      ca:
        kind: ConfigMap
        name: keycloak-certs
        namespace: group-sync-operator
        key: ca.crt
      url: https://keycloak-keycloak-operator.apps.openshift.com

Scheduled Execution

A cron style expression can be specified for which a synchronization event will occur. The following specifies that a synchronization should occur nightly at 3AM

apiVersion: redhatcop.redhat.io/v1alpha1
kind: GroupSync
metadata:
  name: keycloak-groupsync
spec:
  schedule: "0 3 * * *"
  providers:
  - ...

If a schedule is not provided, synchronization will occur only when the object is reconciled by the platform.

Accessing Secrets and ConfigMaps in Other Namespaces

By default, the operator monitors resources in the namespace that it has been deployed within. This is defined by setting the WATCH_NAMESPACE environment variable. Support is available for accessing ConfigMaps and Secrets in other namespaces so that existing resources may be utilized as desired.

To enable the operator to access resources across multiple, set the environment variable with a comma separate list of namespaces that include the namespace the operator is deployed within and any additional namespaces that are desired.

To make use of this feature when deploying through the Operator Lifecycle Manager, set the following configuration on the Subscription resource:

apiVersion: operators.coreos.com/v1alpha1
kind: Subscription
metadata:
  name: group-sync-operator
  namespace: group-sync-operator
spec:
  channel: alpha
  installPlanApproval: Automatic
  name: group-sync-operator
  source: community-operators
  sourceNamespace: openshift-marketplace
  config:
    env:
      - name: WATCH_NAMESPACE
        value: "<comma separated list of namespaces>"

Deploying the Operator

This is a namespace level operator that you can deploy in any namespace. However, group-sync-operator is recommended.

It is recommended to deploy this operator via OperatorHub, but you can also deploy it using Helm.

Deploying from OperatorHub

If you want to utilize the Operator Lifecycle Manager (OLM) to install this operator, you can do so in two ways: from the UI or the CLI.

Multiarch Support

Arch	Support
amd64	✅
arm64	✅
ppc64le	✅
s390x	✅

Deploying from OperatorHub UI

• If you would like to launch this operator from the UI, you'll need to navigate to the OperatorHub tab in the console.
• Search for this operator by name: group sync operator. This will then return an item for our operator and you can select it to get started. Once you've arrived here, you'll be presented with an option to install, which will begin the process.
• After clicking the install button, you are presented with the namespace that the operator will be installed in. A suggested name of group-sync-operator is presented and can be created automatically at installation time.
• Select the installation strategy you would like to proceed with (Automatic or Manual).
• Once you've made your selection, you can select Subscribe and the installation will begin. After a few moments you can go ahead and check your namespace and you should see the operator running.

Deploying from OperatorHub using CLI

If you'd like to launch this operator from the command line, you can use the manifests contained in this repository by running the following:

oc new-project group-sync-operator
oc apply -f config/operatorhub -n group-sync-operator

This will create the appropriate OperatorGroup and Subscription and will trigger OLM to launch the operator in the specified namespace.

Deploying with Helm

Here are the instructions to install the latest release with Helm.

oc new-project group-sync-operator
helm repo add group-sync-operator https://redhat-cop.github.io/group-sync-operator
helm repo update
helm install group-sync-operator group-sync-operator/group-sync-operator

This can later be updated with the following commands:

helm repo update
helm upgrade group-sync-operator group-sync-operator/group-sync-operator

Metrics

Prometheus compatible metrics are exposed by the Operator and can be integrated into OpenShift's default cluster monitoring. To enable OpenShift cluster monitoring, label the namespace the operator is deployed in with the label openshift.io/cluster-monitoring="true".

oc label namespace <namespace> openshift.io/cluster-monitoring="true"

Test metrics

export operatorNamespace=group-sync-operator-local # or group-sync-operator
oc label namespace ${operatorNamespace} openshift.io/cluster-monitoring="true"
oc rsh -n openshift-monitoring -c prometheus prometheus-k8s-0 /bin/bash
export operatorNamespace=group-sync-operator-local # or group-sync-operator
curl -v -s -k -H "Authorization: Bearer $(cat /var/run/secrets/kubernetes.io/serviceaccount/token)" https://group-sync-operator-controller-manager-metrics-service.${operatorNamespace}.svc.cluster.local:8443/metrics
exit

Development

Running the operator locally

make install
export repo=redhatcopuser #replace with yours
docker login quay.io/$repo/group-sync-operator
make docker-build IMG=quay.io/$repo/group-sync-operator:latest
make docker-push IMG=quay.io/$repo/group-sync-operator:latest
oc new-project group-sync-operator-local
kustomize build ./config/local-development | oc apply -f - -n group-sync-operator-local
export token=$(oc serviceaccounts get-token 'group-sync-operator-controller-manager' -n group-sync-operator-local)
oc login --token ${token}
make run ENABLE_WEBHOOKS=false

Test helm chart locally

Define an image and tag. For example...

export imageRepository="quay.io/redhat-cop/group-sync-operator"
export imageTag="$(git -c 'versionsort.suffix=-' ls-remote --exit-code --refs --sort='version:refname' --tags https://github.com/redhat-cop/group-sync-operator.git '*.*.*' | tail --lines=1 | cut --delimiter='/' --fields=3)"

Deploy chart...

make helmchart IMG=${imageRepository} VERSION=${imageTag}
helm upgrade -i group-sync-operator-local charts/group-sync-operator -n group-sync-operator-local --create-namespace

Delete...

helm delete group-sync-operator-local -n group-sync-operator-local
kubectl delete -f charts/group-sync-operator/crds/crds.yaml

Building/Pushing the operator image

export repo=redhatcopuser #replace with yours
docker login quay.io/$repo/group-sync-operator
make docker-build IMG=quay.io/$repo/group-sync-operator:latest
make docker-push IMG=quay.io/$repo/group-sync-operator:latest

Deploy to OLM via bundle

make manifests
make bundle IMG=quay.io/$repo/group-sync-operator:latest
operator-sdk bundle validate ./bundle --select-optional name=operatorhub
make bundle-build BUNDLE_IMG=quay.io/$repo/group-sync-operator-bundle:latest
docker login quay.io/$repo/group-sync-operator-bundle
docker push quay.io/$repo/group-sync-operator-bundle:latest
operator-sdk bundle validate quay.io/$repo/group-sync-operator-bundle:latest --select-optional name=operatorhub
oc new-project group-sync-operator
oc label namespace group-sync-operator openshift.io/cluster-monitoring="true"
operator-sdk cleanup group-sync-operator -n group-sync-operator
operator-sdk run bundle -n group-sync-operator quay.io/$repo/group-sync-operator-bundle:latest

Releasing

git tag -a "<tagname>" -m "<commit message>"
git push upstream <tagname>

If you need to remove a release:

git tag -d <tagname>
git push upstream --delete <tagname>

If you need to "move" a release to the current master

git tag -f <tagname>
git push upstream -f <tagname>

Cleaning up

operator-sdk cleanup group-sync-operator -n group-sync-operator
oc delete operatorgroup operator-sdk-og
oc delete catalogsource group-sync-operator-catalog