Configuring Neo4j Single Sign-On (SSO)

Neo4j supports SSO authentication and authorization through identity providers implementing the OpenID Connect (OIDC) standard. This page features detailed examples of how to configure Single Sign-On (SSO) for several identity providers. It also presents frequently asked questions and solutions to common problems encountered when configuring SSO.

The following configurations are crafted for a Neo4j Browser served on http://localhost:7474/browser/ (the default URL when starting the database on localhost).

Therefore, when reproducing them in the identity providers, you must modify the redirect URI to include the URI serving your Neo4j Browser application. For example:

http://localhost:7474/browser/?idp_id={provider}&auth_flow_step=redirect_uri

SSO works in the following way:

The server (Neo4j DBMS) contacts the identity provider (Okta, Entra ID, Google, etc.) and fetches the JSON Web Keys (JWKs) from the provider.
The client (e.g., Bloom, Neo4j Browser, etc.) asks the user for credentials and contacts the identity provider.
The identity provider responds with a JSON Web Token (JWT), a JSON file containing fields (claims) relative to the user (email, audience, groups, etc.).
The client provides the server with the JWT, and the server verifies its signature with the JWKs.
Introduced in 5.24 Optionally, you can control the authentication and authorization on a user level by setting dbms.security.require_local_user to true in the neo4j.conf file and the auth providers for the users, which authentication and authorization you want to control, using Cypher. This setting mandates that users with the relevant auth provider attached to them must exist in the database before they can authenticate and authorize with that auth provider. For information on how to modify or create users in this mode, see Manage users.
This mode allows you the following using Cypher:
1. Suspend SSO users.
2. Set a home database for an SSO user.
3. Set a user-friendly name for an SSO user (rather than relying on the external identifier).
4. Set a password for an SSO user.
5. Set a password change requirement for an SSO user.

For further information and examples, see Configure SSO at the user level using auth providers.

JWTs must always contain a value for sub even when using a different claim for username. It is the only claim guaranteed to be unique and stable. Other claims, such as email or preferred_username are less secure and may change over time. They should not be used for authentication. Neo4j may assign permissions to a user based on this username value in a hybrid authorization configuration. Thus, changing the username claim from sub is not recommended.

Okta

Access token

This example shows how to configure Okta for authentication and authorization using access tokens.

Configure the client with the appropriate redirect URI. You can skip the group assignments in this step:

Figure 1. Okta OIDC client creation

Figure 2. Okta OIDC client configuration
Take note of the Client ID and the Okta domain. You will need them later when configuring the Okta parameters and the Well-known OpenID Connect endpoint in the neo4j.conf file:

Figure 3. Okta OIDC client configuration
Create groups in Okta, assign users to them (the user can be added to a group either on user creation or editing the group), and map them in the neo4j.conf to native groups:

Figure 4. Okta OIDC server groups
Configure the default authorization server (the one that shows api://default as audience) to return the groups claim in access tokens:

Figure 5. Okta OIDC authorization server

Figure 6. Okta OIDC server claims

Configure Neo4j to use Okta authentication by configuring the following settings in the neo4j.conf file:

dbms.security.authentication_providers=oidc-okta
dbms.security.authorization_providers=oidc-okta
dbms.security.oidc.okta.display_name=Okta
dbms.security.oidc.okta.auth_flow=pkce
dbms.security.oidc.okta.well_known_discovery_uri=https://dev-21056049.okta.com/oauth2/default/.well-known/openid-configuration
dbms.security.oidc.okta.audience=api://default
dbms.security.oidc.okta.claims.username=sub
dbms.security.oidc.okta.claims.groups=groups
dbms.security.oidc.okta.params=client_id=0oa3oq6uw3uSOBf8y5d7;response_type=code;scope=openid profile email
dbms.security.oidc.okta.authorization.group_to_role_mapping= "engineers" = admin; \
                                                             "collaborators" = reader

The token_type_principal and the token_type_authentication are omitted, meaning access tokens are used instead.

Log in with your Okta SSO credentials using the email of an engineer role user that results in an admin role in the database:

Figure 7. Okta OIDC successful login

ID token

This example shows how to configure Okta for authentication and authorization using ID tokens.

Follow the first two steps from the instructions for Access token.

Create the claims as indicated:

Figure 8. Okta claim creation panel

In the case of access tokens, a default sub is already provided automatically. However, for ID tokens, the name you give to your claim needs to be also indicated in the configuration dbms.security.oidc.okta.claims.username=userid.

Configure the default authorization server (the one that shows api://default as audience) as indicated:

dbms.security.authentication_providers=oidc-okta, native
dbms.security.authorization_providers=oidc-okta
dbms.security.oidc.okta.display_name=Okta
dbms.security.oidc.okta.auth_flow=pkce
dbms.security.oidc.okta.well_known_discovery_uri=https://trial-2696363.okta.com/oauth2/default/.well-known/openid-configuration
dbms.security.oidc.okta.audience=0oa42hwrygsUCFlLO697
dbms.security.oidc.okta.claims.username=userid
dbms.security.oidc.okta.claims.groups=groups
dbms.security.oidc.okta.params=client_id=0oa42hwrygsUCFlLO697;response_type=code;scope=openid profile email
dbms.security.oidc.okta.authorization.group_to_role_mapping="admin_group" = admin;
dbms.security.oidc.okta.config=token_type_principal=id_token;token_type_authentication=id_token

You should now find the audience under Okta’s sign-on tab:

Figure 9. Okta’s sign-on tab
Introduced in 5.24 (Optional) If you want control the authentication and authorization on a user level, configure dbms.security.require_local_user to true in the neo4j.conf file. This setting mandates that users with the relevant auth provider attached to them must exist in the database before they can authenticate and authorize with that auth provider. For information on how to create users in this mode, see Creating users.

For example, to create the user jake who can authenticate using native or okta, and authorize using Okta (as configured in step 3), you can use the following Cypher query:
```
CREATE USER jake
SET HOME DATABASE 'jakesHomeDb'
SET AUTH 'oidc-okta' {SET ID 'jakesUniqueOktaUserId'} // `jakesUniqueOktaUserId` must match the value of the claim that you configured via dbms.security.oidc.okta.claims.username
SET AUTH 'native' {SET PASSWORD 'changeme' SET PASSWORD CHANGE REQUIRED}
```
See Configure SSO at the user level using auth providers for further examples.

Microsoft Entra ID (formerly Azure Active Directory)

Access token

This example shows how to configure Entra ID for authentication and authorization using an access token.

After the successful creation of your SSO application in Azure, open the Token configuration tab to configure a token.
1. Click Add groups claim.
2. Select Security groups to include in your access token.
3. Save your changes.
Open the Expose an API tab and select Add a Scope.
1. If you click the Add a Scope button for the first time, you see a new pane stating that you need to add an Application ID URI before proceeding.
  
  You can find it on your app Overview page.
  
  Figure 10. The GUID is used to identify specific resources or instances within Azure. You can find it on the app registration page.
2. Click Save and continue after setting the Application ID URI.
Fill in all mandatory fields in the pane Add a scope.
1. Enter a new Scope name, Admin consent display name, and Admin consent description.
2. Make sure the Enabled scope state is selected.
3. Select the Add scope button again to create a new scope. You can add all scopes supported by your API.
  
  Once the scopes are created, make a note of them for use later.

Configure Neo4j to use Entra ID for authentication by configuring the following settings in the neo4j.conf file:

# Configure the access_token
dbms.security.oidc.azure.config=principal=unique_name;code_challenge_method=S256;token_type_principal=access_token;token_type_authentication=access_token
# Configure the OIDC token endpoint with the Directory (tenant) ID
dbms.security.oidc.azure.token_endpoint=https://login.microsoftonline.com/54e85725-ed2a-49a4-a19e-11c8d29f9a0f/oauth2/v2.0/token
# Configure the iss claim in the id token with the Directory (tenant) ID
# Make sure you add the trailing slash (`/`) at the end of the URL, or this operation might fail.
dbms.security.oidc.azure.issuer=https://sts.windows.net/54e85725-ed2a-49a4-a19e-11c8d29f9a0f/
# Provide the Entra ID parameters, such as client_id, response_type, scope, etc.
dbms.security.oidc.azure.params=client_id=4376dc8b-b5af-424f-9ada-c1c1b2d416b9;response_type=code;scope=openid profile email api://4376dc8b-b5af-424f-9ada-c1c1b2d416b9/access-token

As previously mentioned, the GUID here is also the Directory (tenant) ID. Make sure you add the trailing slash (/) at the end or this operation might fail.

The audience parameter for access tokens is typically set with api:// at the front.

ID token

This example shows how to configure Entra ID for authentication and authorization using ID tokens.

Register the application

Log in to the Azure portal.
Navigate to Microsoft Entra ID > Overview.
From the Add dropdown menu, select App registration and fill in the following information to create your SSO application:

Figure 11. Entra OIDC client creation

The redirect URI http://localhost:7474/browser/?idp_id=azure&auth_flow_step=redirect_uri is the URI that will accept returned token responses after successful authentication.
Click Register.

Configure Neo4j

After the successful app creation, on the app’s Overview page, find the Application (client) ID value. Use it to configure the following properties in the neo4j.conf file.

dbms.security.oidc.azure.audience=c2830ff5-86d9-4e38-8a2b-9efad6f3d06d
dbms.security.oidc.azure.params=client_id=c2830ff5-86d9-4e38-8a2b-9efad6f3d06d;response_type=code;scope=openid profile email

Navigate to Endpoints, to find the OpenID Connect metadata document. Use it to configure the well_known_discovery_uri in the neo4j.conf file.

Figure 12. Entra OIDC client config
```
dbms.security.oidc.azure.well_known_discovery_uri=https://login.microsoftonline.com/ce976899-299d-4a01-91e5-a5fee8f98626/v2.0/.well-known/openid-configuration
```

Configure Neo4j to use Entra ID authentication by configuring the following settings in the neo4j.conf file:

dbms.security.authentication_providers=oidc-azure
dbms.security.authorization_providers=oidc-azure
dbms.security.oidc.azure.display_name=Azure
dbms.security.oidc.azure.auth_flow=pkce
dbms.security.oidc.azure.config=token_type_principal=id_token;token_type_authentication=id_token

Configure which JWT claim should be used for usernames. Possible values are sub, email, or preferred_username.

sub is the only claim guaranteed to be unique and stable. For details, see Microsoft documentation as well as the OpenId spec.
```
dbms.security.oidc.azure.claims.username=sub
```

Map Entra groups to Neo4j roles

Decide whether you want to use Entra groups directly or Entra App Roles.

Using Entra groups directly might be convenient if you already have users assigned to those groups and want to perform Group-to-Role mapping in Neo4j settings.

Entra App Roles allow a layer of separation between Neo4j roles and groups. When App Roles are used, only the roles relevant to Neo4j are sent in the JWT token. This prevents leaking permissions between applications. JWT tokens also have a limitation of 200 roles per token per user, which can be avoided by sending only the relevant App Roles.

Details about Entra ID App Roles can be found in the Microsoft documentation.

Using Entra groups directly

Configure the server to return the Group Object IDs in the JWT identity tokens. To do this, set groupMembershipClaims to SecurityGroup in the Manifest of the registered application:

Figure 13. Entra OIDC server claims
Create groups in the Entra AD console and assign users to them. Take note of the Object Id column. In the next step, you must map these to user roles in the Neo4j settings.

Figure 14. Entra OIDC server groups

Configure a mapping from Entra Group Object IDs to Neo4j roles. For details, see Map the identity provider groups to the Neo4j roles.

dbms.security.oidc.azure.authorization.group_to_role_mapping= "e8b6ddfa-688d-4ace-987d-6cc5516af188" = admin; \
                                                              "9e2a31e1-bdd1-47fe-844d-767502bd138d" = reader

Configure Neo4j to use the groups field from the JWT token.
```
dbms.security.oidc.azure.claims.groups=groups
```

Using Entra ID App roles

On the app’s home page, navigate to App roles and add the Neo4j roles to the Microsoft Entra ID.

Figure 15. Entra OIDC app roles config
The Value column in the App roles config must either correspond to Neo4j roles or be mapped in the neo4j.conf file. For details, see Map the identity provider groups to the Neo4j roles.
```
dbms.security.oidc.azure.authorization.group_to_role_mapping= "managers" = admin; \
                                                              "engineers" = reader
```
Configure Neo4j to use the roles field from the JWT token.
```
dbms.security.oidc.azure.claims.groups=roles
```
Introduced in 5.24 (Optional) If you want control the authentication and authorization on a user level, configure dbms.security.require_local_user to true in the neo4j.conf file. This setting mandates that users with the relevant auth provider attached to them must exist in the database before they can authenticate and authorize with that auth provider. For information on how to create users in this mode, see Creating users.

For example, to create a user jake who can authenticate and authorize using Azure, you can use the following Cypher query:
```
CREATE USER jake
SET HOME DATABASE 'jakesHomeDb'
SET AUTH 'oidc-azure' {SET ID 'jakesUniqueAzureUserId'} // `jakesUniqueAzureUserId` must match the value of the claim that you configured via dbms.security.oidc.azure.claims.username
```
See Configure SSO at the user level using auth providers for further examples.

Google

ID token

This example shows how to use Google OpenID Connect for authentication using ID tokens in conjunction with native authorization.

Configure the client and the redirect URI:

Figure 16. Google OIDC client creation

Figure 17. Google OIDC client configuration

SSO authorization does not work with Google, as the JWT returned by Google does not contain information about the groups that a user belongs to, and cannot be configured to. Therefore, it is recommended to use native (or another flavor) authorization by creating a native version of the user in Neo4j.

Configure Neo4j to use Google authentication by setting the following configurations in the neo4j.conf file:

dbms.security.authentication_providers=oidc-google
dbms.security.authorization_providers=native
dbms.security.oidc.google.display_name=Google
dbms.security.oidc.google.auth_flow=pkce
dbms.security.oidc.google.well_known_discovery_uri=https://accounts.google.com/.well-known/openid-configuration
dbms.security.oidc.google.audience=345461137297-v9brpjmgbvbm3d5s9fq65tktevosd3rn.apps.googleusercontent.com
dbms.security.oidc.google.claims.username=email
dbms.security.oidc.google.params=client_id=345461137297-v9brpjmgbvbm3d5s9fq65tktevosd3rn.apps.googleusercontent.com;response_type=code;scope=openid profile email
dbms.security.oidc.google.token_params=client_secret=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
dbms.security.oidc.google.config=token_type_principal=id_token;token_type_authentication=id_token

Using one of the following options, create a user in the database who can authenticate and authorize natively to be able to give the users roles from native authorization.
Introduced in 5.24

This approach relies on the existence of an admin user who can authenticate natively and then create less privileged users via auth providers, who can authenticate only using oidc-google, but will receive the roles granted to them using native authorization.

An admin user with the name neo4j is created by default when the database is first started.

In the neo4j.conf file, temporarily enable native authentication for the admin user only and enable the user-level control of authentication and authorization:

dbms.security.authentication_providers=oidc-google, native dbms.security.require_local_user=true

This will switch to user auth providers mode whereby users can only authenticate and authorize if they have a corresponding auth provider in the database.

Create a user who can authenticate and authorize only using oidc-google:

CREATE USER jake SET HOME DATABASE 'jakesHomeDb' SET AUTH 'oidc-google' {SET ID 'jakesUniqueGoogleUserId'} (1)

1 jakesUniqueGoogleUserId must match the value of the claim that you configured via dbms.security.oidc.google.claims.username.

Grant the user jake roles, for example, reader:

GRANT ROLE reader TO jake

The user implicitly receives native authorization because native is in the list of authorization providers and you have explicitly granted the user a role.

Once you have set up your users in this way, you can disable native authentication for the database completely. This will prevent all users, including the admin, from logging in with a username and password:

dbms.security.authentication_providers=oidc-google
Alternatively, if you do not use auth providers, you can temporarily enable native authentication to create an SSO-authenticated admin user alice who can then create other users who can only authenticate using SSO.

Temporarily enable native authentication:

dbms.security.authentication_providers=oidc-google, native

Create an SSO-authenticated admin user (in this example an equivalent of alice@neo4j-test.com must be set up in the Google SSO provider and their credentials must be known):

CREATE USER `alice@neo4j-test.com` SET PASSWORD 'secretpassword'; GRANT ROLE admin to `alice@neo4j-test.com`;

Disable native authentication for the database to prevent users logging in with username and password:

dbms.security.authentication_providers=oidc-google

Log in via Google SSO as alice@neo4j-test.com, the admin user.

Create other users who can authenticate only using oidc-google and will receive the roles granted to them using native authorization.

CREATE USER jakesUniqueGoogleUserId (1) SET HOME DATABASE 'jakesHomeDb' SET PASSWORD 'secretpassword' SET PASSWORD CHANGE NOT REQUIRED

1 jakesUniqueGoogleUserId must match the value of the claim that you configured via dbms.security.oidc.google.claims.username.

Grant the user roles using native authorization:

GRANT ROLE reader TO jakesUniqueGoogleUserId

FAQ

When should `pkce` be used as auth flow?

Assuming the client (Neo4j Browser or Bloom) can be accessed through the public internet, always use pkce auth-flow rather than implicit because the latter requires the client’s secret to be available to the public client. In general, if both flows are available, it is recommended to opt for pkce because it is more secure than implicit.

Is Google authentication secure if it has a client secret listed in the config?

Yes. Google uses the pkce flow, but identity providers sometimes also use a client secret to ensure the client asking for a token is the one using it (pkce does not guarantee that). The client secret does not add any additional security as it is public but the pkce flow provides sufficient security.

Could not parse JWT of type "access_token"

When getting the message Failed to get credentials: Could not parse JWT of type "access_token" on Browser, it probably means the provider only accepts ID tokens.

Figure 18. Failed to parse JWT of type access_token

Change to ID tokens in your neo4j.conf:

dbms.security.oidc.{{provider}}.config=token_type_principal=id_token;token_type_authentication=id_token

When should identity tokens vs. access tokens be used?

It is generally safer to use access tokens when possible due to being shorter-lived. If authorization permissions change on the identity provider, Neo4j will fail authorization. Neo4j Browser will try to reconnect and reflect the changed permissions faster than if ID tokens were used.

Debug logging of JWT claims

While setting up an OIDC integration, it is sometimes necessary to perform troubleshooting. In these cases, it can be useful to view the claims contained in the JWT supplied by the identity provider. To enable the logging of these claims at DEBUG level in the security log, set dbms.security.logs.oidc.jwt_claims_at_debug_level_enabled to true and the security log level to DEBUG.

Make sure to set dbms.security.logs.oidc.jwt_claims_at_debug_level_enabled back to false for production environments to avoid unwanted logging of potentially sensitive information. Also, bear in mind that the set of claims provided by an identity provider in the JWT can change over time.

How to debug further problems with the configuration

Apart from the logs available in logs/debug.log and logs/security.log in the Neo4j path, you can also use the web-development console in your web browser when doing the SSO authentication flow with Bloom or Neo4j Browser. This could reveal potential problems, such as the one presented below with an example identity provider and the Cross-Origin Request policy:

Figure 19. CORS error

The solution involves adding the redirect domain to the list of allowed domains in the provider (in this case, localhost:8080):

Figure 20. CORS error solution allowing the redirect domain on the provider

Configuring Neo4j Single Sign-On (SSO)

Okta

Access token

ID token

Microsoft Entra ID (formerly Azure Active Directory)

Access token

ID token

Register the application

Configure Neo4j

Map Entra groups to Neo4j roles

Using Entra groups directly

Using Entra ID App roles

Google

ID token

FAQ

When should pkce be used as auth flow?

Is Google authentication secure if it has a client secret listed in the config?

Could not parse JWT of type "access_token"

When should identity tokens vs. access tokens be used?

Debug logging of JWT claims

How to debug further problems with the configuration

When should `pkce` be used as auth flow?