Using a custom method to authenticate users

This page describes how to support user authentication in Cloud Endpoints.

To authenticate a user, a client application must send a JSON Web Token (JWT) in the authorization header of the HTTP request to your backend API. The Extensible Service Proxy (ESP) validates the token on behalf of your API, so you don't have to add any code in your API to process the authentication. However, you do need to configure your OpenAPI document to support your chosen authentication methods.

ESP validates a JWT in a performant way by using the JWT's issuer's public keys. ESP caches the public keys for five minutes. In addition, ESP caches validated JWTs for five minutes or until JWT expiry, whichever happens first.

Before you begin

  • Add authentication code to your client application, following the authentication provider's documentation.

  • When your client application sends an HTTP request, the authorization header in the request must contain the following JWT claims:
    • iss (issuer)
    • sub (subject)
    • aud (audience)
    • iat (issued at)
    • exp (expiration time)

Configuring ESP to support client authentication

You must have a security requirement object and a security definitions object in your OpenAPI document for ESP to validate the claims in the signed JWT.

To support custom authentication:

  1. Add the following to the security definition in your OpenAPI document:

     securityDefinitions:
       your_custom_auth_id:
         authorizationUrl: ""
         flow: "implicit"
         type: "oauth2"
         # The value below should be unique
         x-google-issuer: "issuer of the token"
         x-google-jwks_uri: "url to the public key"
         # Optional. Replace YOUR-CLIENT-ID with your client ID
         x-google-audiences: "YOUR-CLIENT-ID"
    
  2. Add a security section at either the API level to apply to the entire API, or at the method level to apply to a specific method.

     security:
       - your_custom_auth_id: []
    

You can define multiple security definitions in the OpenAPI document, but each definition must have a different issuer. If you use security sections at both the API level and at the method level, the method-level settings override the API-level settings.

The x-google-audiences field isn't required. ESP accepts all JWTs with the backend service name in the form of https://SERVICE_NAME in the aud claim. To allow additional client IDs to access the backend service, you can specify the allowed client IDs in the x-google-audiences field by using comma-separated values. ESP then accepts the JWTs with any of the specified client IDs in the aud claim.

ESP supports two asymmetric public key formats defined by the x-google-jwks_uri OpenAPI extension:

  • JWK set format. For example:
    x-google-jwks_uri: "https://YOUR_ACCOUNT_NAME.YOUR_AUTH_PROVIDER_URL/.well-known/jwks.json"
    
  • X509. For example:
    x-google-jwks_uri: "https://www.googleapis.com/service_accounts/v1/metadata/x509/securetoken@system.gserviceaccount.com"
    

If you are using a symmetric key format, set x-google-jwks_uri to the URI of a file that contains the base64url-encoded key string.

If you omit x-google-jwks_uri, ESP will follow the OpenID Connect Discovery protocol to automatically discover the JWKS URI for the given OpenID provider. ESP will make a request to x-google-issuer/.well-known/openid-configuration, parse the JSON response, and read the JWKS URI from the top-level jwks_uri field.

Note that omitting x-google-jwks_uri will result in higher cold start times, as ESP has to make an extra remote call on startup. Therefore, it is only recommended to omit this field if the JWKS URI changes often. Most certified OpenID providers (such as Google, Auth0, and Okta) have stable JWKS URIs.

You may also customize JWT locations by adding x-google-extensions. For details, see openAPI extensions.

Making an authenticated call to an Endpoints API

When you send a request using an authentication token, for security reasons, we recommend that you put the token in the Authorization:Bearer header. For example:

curl -H "Authorization: Bearer ${TOKEN}" "${ENDPOINTS_HOST}/echo"

Here, ENDPOINTS_HOST and TOKEN are environment variables containing your API host name and authentication token, respectively. See Making an authenticated request to an Endpoints API. for sample code that sends a request using the Authorization:Bearer header.

If you cannot use the header when sending the request, you can put the authentication token in a query parameter called access_token. For example:

curl "${ENDPOINTS_HOST}/echo?access_token=${TOKEN}"

Receiving authenticated results in your API

ESP usually forwards all headers it receives. However, it overrides the original Authorization header when the backend address is specified by x-google-backend in OpenAPI specification or BackendRule in gRPC service configuration.

ESP will send the authentication result in the X-Endpoint-API-UserInfo to the backend API. It is recommended to use this header instead of the original Authorization header. This header is base64url encoded and contains the following JSON object:

{
  "id": "from-sub",
  "issuer": "from-iss",
  "email": "from-email",
  "audiences": ["from-aud"],
  "claims": {
     original-jwt-payload
   }
}

If you are using ESPv2, the format of the value in the header is different. See Migrate to Extensible Service Proxy V2 Beta for more information on the new format.

What's next