As I know, JWT auth is definitely stateless, as a web server does not save any status of an user. However, Oauth2 is different. It stores each key of each user in an auth server, which means "stateful". No?
As I know, JWT auth is definitely stateless
No, JWT is not definitely stateless. It is just a format for signed/encrypted data and you still have to manage key materials to compute or use such token.
OAuth2 is different
OAuth2 is a framework protocol. There is no notion of server "state" in the specification. But to ensure a high security level, the servers that implement that protocol should have to manage clients, initial access tokens, access tokens or refresh tokens, authorization codes...
Related
I'm developing rest api in spring for my company. I need to secure it using google OAtuh2. Firstly I used jwt token validation. Frontend app obtains token_id andd pass it to backend resource server. I've read that I sholud use access tokens instead of id_tokens (is that right?). For using opaque tokens introspection I need to provide introspection uri. I can't find it. Does google auth server support token introspection?
Yes you should use access-tokens only to secure requests to resource-server(s). ID tokens are for clients to get user-info.
Using access-token introspection has serious performance impact compared to JWT decoding: resource-server must issue a query to authorization-server for each and every request it processes (when JWT validation can be done locally with authorization-server public key retrieved only once). I recomand you don't do that.
If you cannot configure authorization-server to add the claims you need to JWT access-tokens, you should consider using an authorization-server capable of identity federation in front of it. It is very common to do so when several etherogenous identity sources are used (Google + Facebook + Github + LinkedIn + corporate LDAP for instance), but it would work with a single one too. Keycloak for instance does that and allows you to put whatever you need in access-tokens.
I'm new to security and I'm trying to understand how to implement proper security without any overkill.
Below are my questions.
I don't want to allow 3rd party clients to use my API and hence I don't see any importance of OAuth 2.0. Hence I'm looking to use the sessions generated by keycloak (or Ory Kratos) in my Spring Boot Security. Any guidance on how to do that.
I have come across an application https://opstra.definedge.com/ which security is implemented using keycloak (can see the URL pattern). But in the requests, I can't see any JWT token in the chrome DevTools Network Tab while performing any network requests. I think they are implementing it the way I wanted. Any overview on how it is implemented.
I'm not architect at definedge, but I'm pretty sure they do not use Keycloak sessions in Opstra (they would have to run Opstra inside Keycloak servlet for that). It more looks like they use OAuth2 to authenticate users from a Java client and that this client has sessions of its own enabled (JSESSIONID cookie for opstra.definedge.com VS sso.definedge.com). It is quite possible that this java client uses access-tokens to authorize requests to resource-server(s), we just can't see it from the browser.
Restricting the clients allowed to consume your API has little to do with authorization method:
with basic authorization header, any client with login and password can access
with Bearer authorization header, any client with a valid token can access (which you already had anticipated)
even JSESSIONID cookie can be set for any origin (I believe), in which condition any request from the same browser would be authorized, whatever the web client.
Restricting your API clients is more about CORS configuration, which aims at just that: filtering which origins (host + port) can access which resource location (URL)
There is a notion of confidential client in Keycloak where the client must provide a password in addition to client-id to exchange authorization codes for access-tokens, but this does not apply to rich clients (clients running on devices you cannot trust): Angular, Vue, React, native mobile apps, etc. code can be reversed enginereed to read that password. But it is possible to configure a Java client of your own as "confidential" and as so, allow this client only to get access tokens to query resource-server (API).
OAuth2 comes with much more than just easing multi-client scenarios and JWTs with session-less java applications greatly ease horizontal scalability and fault tolerance. You should read this article for a refresher on
OAuth2 (and its value)
resource-server security configuration (with CORS)
I'm having a hard time figuring a good way to implement Oauth 2.0 and OpenID Connect authentication alongside an existing internal email+password authentication for a B2B Web app's API using Spring security.
We have a backend REST API that is a Spring Boot servlet application which currently authenticates users with OAuth 1.0 and the password grant. The front-end is an Angular single-page app through which users must log in with their username and password. The API's /oauth/token endpoint then delivers an opaque access token to be used for fetching secured resources that are then displayed within the app.
We'd like to add the possibility to log in using external authentication with OpenID connect, which is a perfect opportunity for switching to OAuth 2.0 and JWT tokens. Our API would then accept JWT tokens it delivered as well as external JWT tokens emitted by accepted issuers.
Reading and validating JWT tokens won't be a problem using Spring security's OAuth Resource Server. However things get complicated with how to make the classic username+password login flow work with JWT tokens.
We thought about delivering JWT access tokens the same way we used to with our old OAuth 1.0 tokens. The thing is, newer OAuth specifications advise against using the password grant, plus it simply isn't supported in the Spring authorization server project we're planning to use. The authorization-code flow w/ PKCE seems like too much for this case as we do not want the back-end API to render a login form but use credentials entered in the existing login form that is part of the single-page app.
I've read a lot about not using OAuth for 1st party login since its primary use is for external authentication, but again, that doesn't apply since we also want 3rd party authentication.
What would be a secure way to implement a Spring boot authorization server that can deliver JWT access tokens to a 1st party client in exchange for a user's credentials, all this using the existing log in page, and given the password grant type no longer being supported?
I thought about implementing a preliminary step that would be a POST request with Basic authentication to a /login endpoint that just returns a 200 HTTP status, then proceeding to the /oauth2/authorize request that would deliver the authorization code immediately without redirecting since my session is authenticated.
I'll be happy to provide more details if needed. Here are the resources I'm using for this project.
What about setting up an authorization-server capable of identity federation?
In such configuration, the authorization-server is responsible for providing with identities, proxying one or more sources (your existing user database for instance, plus maybe Google, Facebook, Github, etc.)
Keycloak does it for instance.
From the client point of view (your Angular app), the authorization-server is used to get:
access-token: put in Authorization header of requests to secured resource-server(s) (can be a JWT or an opaque string, doesn't matter, clients should not try to extract data from access-tokens)
refresh-token: send to authorization-server to renew access-token before it expires
id-token: get user-profile data (email, username, profile picture, etc.)
You should have a look at https://github.com/damienbod/angular-auth-oidc-client for connecting an Angular app to an OIDC authorization-server.
From resource-server point of view, access-tokens are the source ofr setting-up security-context, either by decoding and validating a JWT locally or with token introspection on authorization-server.
Have a look at those tutorials for such resource-servers configuration.
Spring Security OAuth2 Client uses HttpSession to store the access token if we use oauth2Login() to obtain the access token from Okta (Authorization server). That will force us to have only a single instance of OAuth2 client and won't be able to horizontally scale in an environment like Kubernetes.
To solve this issue I was able to find different approaches on the internet.
Use implicit flow where the access token is obtained and maintained in the browser client which is less secure to the above approach.
Once you obtained the access token in the backend, produce a new custom JWT token with all the claims copied and store it in the user's browser. So every subsequent request can be authenticated using that JWT token thereafter.
Use the Spring-Session library to share the HttpSession between multiple OAuth2 client instances and stick to plain old HTTP sessions to store access token.
What would be a better approach?
I'm considering using authenticated encrypted JWT tokens to authenticate / authorized access to an ASP.NET Web API application.
Based on what I've read so far, it seems to me like it is an option to generate JWT tokens from a token service and pass them to Web API via the http authorization header.
I have found some good code examples on implementing the JWT creation and consumption (Pro ASP.NET Web API Security by Badrinarayanan Lakshmiraghavan).
I'm trying to understand if I need a full OAuth implementation to support this, or if I can simply pass the tokens along in the auth header.
Assuming the tokens are properly encrypted and signed, is there any inherent security flaw in keeping things simple without having to use OAuth?
Trying to keep things as simple as possible for my needs without compromising security.
It is not that you must always OAuth when you use tokens. But given the fact that your application is a JavaScript app, you would be better off implementing a 3-legged authentication. Thinktecture identity server does support implicit grant. But if the client application getting access to the user credential is not a problem for you, your JavaScript app can get the user ID and password from the user and make a token request from a token issuer ensuring the user ID and password are not stored any where in JavaScript app (including DOM). This request for token can be a simple HTTP POST as well and it does not need to be anything related to OAuth. If your end user will not enter the credentials in the client application, OAuth implicit grant is the way. BTW, you don't need to encrypt JWT. TIS issues signed JWT and that will ensure token integrity. But if you are worried about the confidentiality, you can use HTTPS to both obtain the token as well as present the token.
It looks like you don't really need auth delegation as the one provided by OAuth. Isn't HMAC authentication enough for your scenario ?. With HMAC, you will not have to deal with JWT at all. This is an implementation I made for HMAC authentication for .NET
https://github.com/pcibraro/hawknet
Pablo.