I've been reading through a bunch of documentation for using OAuth with Azure AD, but am still completely confused about how to properly implement things for my situation. Hopefully someone can steer me in the right direction.
I have created an ASP.NET Web API application that uses the EWS Managed API to access Exchange on behalf of different users. My application exposes endpoints such as /Mailbox/Messages and /Appointments with the intent that some front end web application will eventually use them to retrieve a user's emails and appointments. Currently the endpoints are working using basic http authentication, but I'd like to update them to use OAuth. The application has been registered in my Azure AD instance and I've configured it to require the "Access mailboxes as the signed-in user via Exchange Web Services" API permission.
Since the front end hasn't been implemented yet, I've been trying to test by manually calling the authentication endpoint. This prompts me to log in and provide consent. If I consent, I'm redirected to the callback URL that I provided when I registered the app with the authorization code contained in the query parameters. I'm still not quite sure how I'm supposed to be using this callback, but for the sake of testing I currently have the callback redeem the authorization code for an access token. This is done by calling the AcquireTokenByAuthorizationCode method on an instance of the AuthenticationContext class and providing my application's id and secret. Again, just for the sake of testing I return the access token to the browser. I can then call my aforementioned endpoints (after some modifications) with this access token and get the emails for the user. I'm guessing much of this is not the correct way to be doing things.
Some of my points of confusion:
What should the callback that I registered in Azure AD actually be doing when it gets the authorization code? Is this intended for a different type of application? Perhaps one that isn't just playing the role of a middle man.
I'm trying to make my application somewhat RESTful, so I don't want to have to maintain the access tokens on my end between requests. As such, does it make sense for my endpoints to expect that the access token be provided in the authentication header for each request? If so, does that mean the front end application should be responsible acquiring the access token and passing it to me?
Being completely new to OAuth and Azure, I'm not sure if any other details are pertinent, but I can provide more information as needed.
What you are implementing is this scenario: https://learn.microsoft.com/en-us/azure/active-directory/active-directory-authentication-scenarios#daemon-or-server-application-to-web-api
Here's how it works:
Your client app redirects the user to sign in at the authorization endpoint
Your client app gets back an authorization code (if using the auth code grant flow, there are others)
The client app exchanges the code for an access token for your API app
It will need to provide its client id and secret along with the code and the API's resource URI to get it
The client app calls to your API app, passing the access token in the Authorization header
Your API app then validates the access token, and requests for another access token from Azure AD for the Exchange API
It will pass the access token sent by the client app, along with its client id and secret and the Exchange API's resource URI to Azure AD
Your API app receives an access token so you can call to the Exchange API as the user
And to answer your two questions:
Authorization code flow is not used with APIs, only with apps that have a user signing in, thus the redirect URL is basically never used
Your API can and must expect and authenticate the access token for it to be in every request. But the access token it uses to call the Exchange API can and should be cached on the API's side. This is provided out-of-the-box with ADAL, though the tokens are only in memory.
Related
We have planned to implement authentication in our API using OAUTH. For this purpose I read so many articles on web to explore it. After read these articles what I am understanding is
Send credentials to authorization server and after successful
authentication it will send you the access token.
Use this access token for further calling of your api methods.
To authenticate our api user needs to pass the following parameters.
Authorization Token
Employee ID
What I am thinking is to pass these values via request headers. Problem is that these request headers can easily be viewed in browser console and someone can misused it easily. Please suggest Is this the right way to authenticate api or we used something else for this purpose?
I'm following the guide and example provided by Microsoft here and I'm able to get the demo working, with the authentication happening in a console app, then making a request to a Web API with the correct token.
I'm looking to use this but the code in the console app would need to move to a Web App. Essentially: external server tries to access secure Web API, providing Azure AD username/password in the Authentication header of a HTTPS request. I pick up these credentials in the first insecure Web API, and attempt to authenticate the credentials against AD, obtaining the token. From here, I would then call the [Authorize]-protected Web API by making a request with the AD token.
At this point I'm using the same code from the example linked above, simply moving the code in the Console app up into the first insecure Web API controller, but I'm having no luck. I read on CloudIdentity that "You can only use those flows from a native client. A confidential client, such as a web site, cannot use direct user credentials.". Is this true? If so, is there another way to achieve my aim? I need to use the credentials as it may be likely that more services would use the API in the future, so each of these would need their own credentials to use that could be managed within Azure.
EDIT: In reading more around this, should I actually be aiming to use Client authentication, creating an "Application" within the Azure AD, and providing the client ID to each external service looking to call the API, to then authenticate with that, rather than credentials?
Yes, your edit is correct. The Resource Owner Password Credentials grant is meant to authenticate users, not applications. Typical use would be from an application that prompts you for username and password and then retrieves a token from Azure AD.
You can use the Client Credentials grant to get a token from Azure AD from a confidential client to call an API without user context. This flow requires that you register the application in Azure AD and generate a key (which will be used as the client secret). You can then use the ADAL library to ge a token from AAD as shown here.
I have a question regarding the best standard architecture of Authorization in web application that is written in Asp.Net Web Api on the backend and and has an angularjs client side.
According to what I had seen before, the "Resource Owner Credentials" flow is what would be used in such cases, where the webapp would send the user's credentials to the server and obtain access token (and refresh token) and then using an interceptor, every call to the backend apis would contain the access token in the header.
However, I have recently seen arguments about it being a bad idea, as it gives the user's credentials to the client app.
What is the best flow for a scenario when you have javascript client directly calling you WebApis? What is the best way to secure it using Identity Server?
You could also consider implicit flow or hybrid flow, when the client app (angular) is redirecting the user to login on the openid identity provider (Identity Server) an upon a successful authentication, this returns an access/identity token which can be used in subsequent calls to the Apis.
In this case the client app is never touching client credentials, it always has to manage tokens.
see also https://www.scottbrady91.com/OpenID-Connect/OpenID-Connect-Flows
I have a RESTful API written in ASP.Net that implements OAuth 2 for authentication, and it's currently accessed through a web application. I've also got a legacy desktop client that accesses the same resources directly (not through the RESTful API and without OAuth, but using the same login credentials and hitting the same database). The requirement I'm trying to meet right now is to allow a user to click a link in the desktop application in order to open the web app to a specific screen, and when they do, to have the web app authenticate automatically so that they don't have to manually log into it (since they've already logged into the desktop app).
I'm trying to work out how I can handle this within the constraints of the framework. I'm not too familiar with OAuth 2 in general, but from what I understand I shouldn't share tokens between clients and there are no flows specifically for this kind of hand-off (unless I'm missing something). Worst case scenario, I could generate a temporary token outside of OAuth that's used by the web client to authenticate rather than a username and password, but I'm hoping to avoid stepping outside of what's already in the framework to do what I need to do.
So the question is this: is there some decent way built into the OAuth 2.0 framework to handle this sort of "handshake" between two applications, or should I just build my own method of dealing with it?
Using temporary one-time tokens is actually part of OAuth spec (authorization_code grant type). In this case this short-lived code can be exchanged for access_token (and refresh_token). You will have to implemenent generating and validating of this authorization_code.
If you are using OWIN OAuth middleware:
You can generate the code at separate API endpoint accessed by your desktop client app.
After receiving token, pass it to your browser and direct it to auth endpoint with grant_type=authorization_code over secure connection. Example: call Process.Start("https://example.com/ExternalLogin/authorization_code_goes_here"). At the webpage redirect user to your OAuth Token endpoint with grant_type=authorization_code.
AuthenticationTokenProvider.Receive will be called, in which you will validate your token. (Example code here).
After successful validation OAuthAuthorizationServerProvider.GrantAuthorizationCode will be called, in which you will process the authenticated user in the same way you process it with grant_type=password.
Remember that your token validation logic should ensure that your tokens are short-lived, usable only once and transmitted over secure connection.
This is sometimes called "single sign-on" if you want to research this topic further.
I have an admin-console for an existing service that I want to provide access to by adding login, using our company's OAuth 2.0 service. I want to use Apigee here, so that the web-app with the admin-console does not have to implement the login-logic.
My idea was to use AuthorizationCode flow and let Apigee manage the tokens and I looked into https://github.com/apigee/api-platform-samples/tree/master/sample-proxies/oauth-login-app, but I really can't see how our existing OAuth service fits in.
Is there a sample like that? Perhaps using Google's or Facebook's OAuth service to authenticate the user?
First, Apigee needs to be a proxy into the admin-console. This means that all traffic to the admin-console has to go through Apigee. Otherwise, you won't be able to enforce authentication.
Second, there are a couple different options for integrating with the external oauth 2.0 service. Apigee has the ability to store an external access token and use it as its own, or Apigee can generate a token and store the external access token as a custom attribute.
High level thoughts on how the Apigee proxy could look like:
ProxyEndpoint - endpoint exposed to clients connecting to admin console
TargetEndpoint (not shown in that oauth login-app example) - endpoint for the actual admin console
The flows that execute in the Apigee proxy before sending the request to admin-console will need to implement logic that checks an authentication token. If it's valid, let the request pass onto the TargetEndpoint (admin-console). If the request isn't valid, step through logic that goes calls the external oauth 2.0 server's auth code flow. This will require the following:
Apigee needs to be registered with external oauth 2.0 server.
Logic needs to be built in this proxy to support the redirection based flow of authorization code grant_type (obtaining auth code, receiving the auth code, obtaining token --> all while being redirection based and transparent to user).
In addition to #2, Apigee will need to store the external token as custom attribute and expose the apigee token, or store the external token for verification purposes later on. http://apigee.com/docs/api-services/content/authorize-requests-using-oauth-20 (see Delegating token management). After the token is stored, you'd need to respond with another 302 redirect to the initial uri + token so the request can pass through to admin-console as an authenticated request.
#2 isn't exactly straight-forward and there won't be an example proxy that shows this implementation. If the oauth 2.0 service supported a password grant, it may simplify the implementation, but allows the credentials to pass through apigee and not directly with the authorization server.