Everything I'm reading shows that in order for an application to use onedrive, it has to do the oauth2 thing to get credentials. But what if you're a batch process and don't have a web interface for your users.
Google's API has a special type of account called a service account where once you set it up, you can control access to everything from that one account, no need to interact with users. Does such a thing exist for onedrive?
App-only authentication doesn't require the user be prompted for credentials but it also isn't supported in 100% of scenarios. For example, the APIs need a user principle for creating special folders and resolving a user's personal site. Also, it is only supported for OneDrive for Business, not Consumer. Consumer always requires the user be prompted for initial authentication.
Another option would be to spin up a web service of some sort that handles initial user authentication, ie. a sign up page. With that, you can retrieve a refresh token for offline authentication and store it for the user. Every authentication from then on can be done using the refresh tokens, which doesn't require a user prompt.
I finally found this. It's the same basic idea as google's service account, but I think it's harder to use. But at least the concept is supported.
http://blogs.msdn.com/b/exchangedev/archive/2015/01/21/building-demon-or-service-apps-with-office-365-mail-calendar-and-contacts-apis-oauth2-client-credential-flow.aspx
Related
I was wondering if it was possible to login to different salesforce environments (Sandboxes, scratch orgs, production env, etc) using either Apex/LWC/Aura (or anything that I can make a quick action to). For example, I have a list of credential records, with the username and password, and I would like to have a login button that creates a separate tab that can automatically redirect to that specific instance and log in.
Currently, if a user wants to login to a particular instance, they have to either go to test.salesforce.com or login.salesforce.com (depending on if it's a sandbox or production) manually, then copy the password and username in. The ideal situation is to have a login button that can do this automatically from the record page where the username and password is located.
I think previously this could have been accomplished through the URL, but salesforce has recently patched this out due to security concerns. Is there another good way to do this?
It sounds like you're trying to solve two specific challenges:
Your users need to be able to manage very high volume of credentials.
You need authentication to survive password resets.
The clear solution, in my mind, is to use the OAuth Web Server flow to execute initial authentication and then store the refresh token that results from this flow. This token survives password resets, and may be used more or less indefinitely to create new access tokens - which users can then use to log in via a frontdoor link.
There's an out-of-the-box tool that does this already: the Salesforce CLI. You can authenticate orgs to its toolchain, name them, and subsequently access them with a single command (sfdx force:org:open). Users that prefer a GUI can access the exact same functions in Visual Studio Code.
If you're hellbent on doing custom development to handle this use case, you can, but you need to be very careful of the security implications. As one example, you could implement an LWC + Apex solution that executed the relevant OAuth flows against orgs and stored the resulting data in an sObject, then allowing users to click a button to generate a new access token and do a one-click login.
But... if you do this, you're storing highly sensitive credentials in an sObject, which can be accessed by your system administrators and potentially other users who have relevant permissions. That data could be exfiltrated from your Salesforce instance by an attacker and misused. There's all kinds of risks involved in storing that kind of credential, especially if any of them unlock orgs that contain PII or customer data.
One of the two best answers for that (the other one being 'pure Apex' and relatively more complex) is using Flow.
"You can use a login flow to customize the login experience and integrate business processes with Salesforce authentication. Common use cases include collecting and updating user data at login, configuring multi-factor authentication, or integrating third-party strong authentication methods.enter image description here"
"You can use login flows to interact with external third-party authentication providers by using an API.
For example, Yubico offers strong authentication using a physical security key called a YubiKey. Yubico also provides an example Apex library and login flow on GitHub. The library supplies Apex classes for validating YubiKey one-time passwords (OTPs). The classes allow Salesforce users to use a YubiKey as a second authentication factor at login. For more information, see yubikey-salesforce-client.
You can also implement a third-party SMS or voice delivery service, like Twilio or TeleSign, to implement an SMS-based multi-factor authentication and identity verification flow. For more information, see Deploy Third-Party SMS-Based Multi-Factor Authentication."
learn more here: enter link description here
My team and I have been working on a web application for our clients that uses JSON web tokens for authentication and authorization. Using Azure AD as our identity provider, we verify a user's identity and generate a signed JWT with the user's permissions in it. The JWT then gets included in the authorization header of all subsequent requests to the APIs. Pretty standard stuff as far as JWTs go.
We're now being asked to provide the capability to link directly into our system from another third-party web application without forcing the user to reauthenticate. I'm trying to figure out if there's a way to do so without creating a massive security loophole.
The way I picture this working would be to implement an endpoint for programmatic authentication in our system that accepts a cryptographically signed payload with an API key and the user's ID or email address. The third-party system would have a private key with which to sign the payload, and we'd have a public one to verify the signature. If the request is legitimate, we'd issue a token for the specified user, and they could use that to link to whatever they like.
I'm already getting yelled at by at least one person that this is a complete joke from a security standpoint because, among other things, it completely bypasses AAD authentication. I believe the third-party system in question does use AAD for authentication, but that's not really relevant either way because we're trusting them implicitly whether they've authenticated their users or not. Either way I take his point.
I'm not a security expert and I don't claim to know whether there even is a proper way to do this kind of thing, but from my vantage it doesn't really seem all that much less secure than any other mechanism of authentication and authorization using JWTs. Is that true? Are we nuts for even trying? Is there a way to do it that's more secure? What should I know about this that I demonstrably don't already?
Thanks in advance for the help. At the very least I hope this spurs some helpful conversation.
Single Sign-On (SSO) enables users to enter their credentials once to sign in and establish a session which can be reused across multiple applications without requiring to authenticate again. This provides a seamless experience to the user and reduces the repeated prompts for credentials.
Azure AD provides SSO capabilities to applications by setting a session cookie when the user authenticates the first time. The MSAL.js library allows applications to leverage this in a few ways.
MSAL relies on the session cookie to provide SSO for the user between different applications.
Read more in this documentation.
I am very new to the concepts of SSO and Keycloak. I am trying to read the official documentation of Keycloak. In the "Supported Protocols" part (https://www.keycloak.org/docs/latest/securing_apps/index.html), the documentation talks about the two use cases of using OIDC like this:
"The first is an application that asks the Keycloak server to authenticate a user for them. After a successful login, the application will receive an identity token and an access token. The identity token contains information about the user such as username, email, and other profile information. The access token is digitally signed by the realm and contains access information (like user role mappings) that the application can use to determine what resources the user is allowed to access on the application.
The second type of use cases is that of a client that wants to gain access to remote services. In this case, the client asks Keycloak to obtain an access token it can use to invoke on other remote services on behalf of the user. Keycloak authenticates the user then asks the user for consent to grant access to the client requesting it. The client then receives the access token. This access token is digitally signed by the realm. The client can make REST invocations on remote services using this access token. The REST service extracts the access token, verifies the signature of the token, then decides based on access information within the token whether or not to process the request."
What I do not understand is this: In the first paragraph it talks about an application making a request and in the second one it talks about a client. But aren't applications counted as clients? Why the specific differentiation? And can anyone given an example of the remote services that is talked about in the second part?
Thank you.
But aren't applications counted as clients? Why the specific differentiation? And can anyone given an example of the remote services that is talked about in the second part?
Yes exactly it. The reason for the differentiation is because there could be many applications more than just this one client. And the client, that the user is authed against may want to access all those other applications' data.
For example take the google ecosystem. Does google email have access to drive, and photos, etc... While it could out the box, it doesn't. You need to explicitly allow email "offline access" to those other applications, even though they are all part of the same platform.
Keycloak understands this and provides that terminology. But it is a bit confusing because this isn't the best way to think about it. Instead a better explanation is that there is just the user and service clients. The service clients all talk to each other and ask for a user's data. While a user may want their data by going straight to one application, other applications may want that user's data too.
Assuming you want to actually allow one service to ask for user data from another service, you want to be using something that supports authorization as a service and not just authentication. There are some examples of this, such as PolicyServer and Authress.
I'm developing an application where Google Drive will be used to manage some documents. The idea is to create a document with some initial template data and provide the users access by adding them as collaborators of the document.
I'm familiar with the OAuth authentication process, I used it in another part of the system to manage the users Calendar...
But in this case these documents will be stored in a generic account of the company, so I can't have the approval prompt for authentication, since users won't have the password of the account.
I'd like to directly authenticate in this account, could be with the username and password hardcoded in the Java code.
Problem that this method of authentication was depreacated and I didn't found a relpacement.
Any ideas?
Thanks in advance,
Phillip
There are 2 ways that comes to mind:
Service accounts: best suited for server side OAuth with traditional backend
Regular Account owned by the application : similar to the process already in place for client side Oauth that you are already familiar with; Auth, store the refresh, ask new token if the AuthCode is expired, and so on.
I personally use and prefer the second solution more as I feel is more flexible to adapt in the future for Oauth Client Side get the tokens and use them server side.
I am exploring the possibilities of a banking mobile HTML5 application. It will be contacting with the main server via RESTful API. Very often I hear that people are using OAuth in their mobile apps to access APIs. For example, SpringSource's html5expense demo app.
So I don't fully understand why bother? Couldn't the user just login in a standard way, receive a cookie with session id (or in case of Play framework, session data), that will be used to identify user when the app makes requests to REST?
Oauth is usually a lot more secure than most BASIC AUTH, or "logging in in a standard way" approaches (and OAuth is becoming more and more of a standard).
When you login, through most "standard" ways, the user enters his username & password, into the application, and username/password are then often either stored locally, or transferred to the application, to then potentially be relayed to a "main server" that for example provides the API. So the user will have to enter his very secret login information (e.g. for banking?), into a client, app or system he doesn't know or trust...
With OAuth, the user is directed to a login page of the owner of that API .. e.g. his bank for example, where he logs into the secure login page that he knows and is asked for his consent that the application "xyz" would like to access his data.... The application that has requested that access, is then given a token with which it can access the API without needing to know the username and password. That way the username/password is only entered once, at a location the user trusts.
Furthermore, the user could later log into and admit page .. (the bank app? or and admin frontend), and delete the given access right to the API, and so stop an application accessing his information, without having to change his password.
Beyond the effect of being actually safe, using something like OAuth, for a banking app also makes sense as it will give people more confidence if modern security techniques are applied. It makes it also feel safer.
If you are not going to publish your API to third party developers; there really is no reason to bother with OAuth.
The biggest reason OAuth exists is to enable integrations with your API without your users having to give out their username and password to a third party. Other reasons is that it makes it possible to put a time frame on third party access to resources, or to scope access.