How does cookie-based authentication work? - authentication

What would be a step-by-step description of how cookie-based authentication work?
I've never done anything involving either authentication or cookies. What does the browser need to do? What does the server need to do? In what order? How do we keep things secure?
I've been reading about different types of authentication and about cookies, but I would like a basic description of how to use the two together. I've only read that they are often used together, but I could not find a description of how.

To expand on Conor's answer and add a little bit more to the discussion...
Can someone give me a step by step description of how cookie based authentication works? I've never done anything involving either authentication or cookies. What does the browser need to do? What does the server need to do? In what order? How do we keep things secure?
Step 1: Client > Signing up
Before anything else, the user has to sign up. The client posts a HTTP request to the server containing his/her username and password.
Step 2: Server > Handling sign up
The server receives this request and hashes the password before storing the username and password in your database. This way, if someone gains access to your database they won't see your users' actual passwords.
Step 3: Client > User login
Now your user logs in. He/she provides their username/password and again, this is posted as a HTTP request to the server.
Step 4: Server > Validating login
The server looks up the username in the database, hashes the supplied login password, and compares it to the previously hashed password in the database. If it doesn't check out, we may deny them access by sending a 401 status code and ending the request.
Step 5: Server > Generating access token
If everything checks out, we're going to create an access token, which uniquely identifies the user's session. Still in the server, we do two things with the access token:
Store it in the database associated with that user
Attach it to a response cookie to be returned to the client. Be sure to set an expiration date/time to limit the user's session
Henceforth, the cookies will be attached to every request (and response) made between the client and server.
Step 6: Client > Making page requests
Back on the client side, we are now logged in. Every time the client makes a request for a page that requires authorization (i.e. they need to be logged in), the server obtains the access token from the cookie and checks it against the one in the database associated with that user. If it checks out, access is granted.
This should get you started. Be sure to clear the cookies upon logout!

A cookie is basically just an item in a dictionary. Each item has a key and a value. For authentication, the key could be something like 'username' and the value would be the username. Each time you make a request to a website, your browser will include the cookies in the request, and the host server will check the cookies. So authentication can be done automatically like that.
To set a cookie, you just have to add it to the response the server sends back after requests. The browser will then add the cookie upon receiving the response.
There are different options you can configure for the cookie server side, like expiration times or encryption. An encrypted cookie is often referred to as a signed cookie. Basically the server encrypts the key and value in the dictionary item, so only the server can make use of the information. So then cookie would be secure.
A browser will save the cookies set by the server. In the HTTP header of every request the browser makes to that server, it will add the cookies. It will only add cookies for the domains that set them. Example.com can set a cookie and also add options in the HTTP header for the browsers to send the cookie back to subdomains, like sub.example.com. It would be unacceptable for a browser to ever sends cookies to a different domain.

Cookie-Based Authentication
Cookie-based authentication normally works in these four steps:
The user provides a username and password in the login form and the client/browser sends a login request.
After the request is made, the server validates the user on the backend by querying the database. If the request is valid, it will create a session by using the user information fetched from the database and store them. For each session a unique ID called the session ID is created. By default, the session ID will be given to the client through the browser.
The browser will submit this session ID on each subsequent request. The session ID is verified against the database. Based on this session ID, the server will identify the session belonging to which client and then give the request access.
Once a user logs out of the app, the session is destroyed both client-side and server-side.

Related

Best practice for storing custom authentication token

I'm feeling really stupid right now - I'm an experienced javascript & PHP developer but I just can't get my head round this one!
I have developed a system that allows users to create widgets and embed them on their own sites by including one dynamically-generated javascript file. I have now incorporated an authentication system so users of these widgets can register and login on my client's sites. The registration details are stored on my server.
When a user registers or logs in, I was planning on sending back some kind of encrypted token (of their unique ID) so that when they next access a widget on the same site the code will automatically log them in by sending the token to my server and decrypting it.
However, if this token is stored in the browser via localStorage or a cookie, it will be copy-and-paste-able to anyone that can get their hands on it, resulting in them being able to log in as that user. Is there any way I can achieve what I want without this risk?
Thanks in advance!
You can use JWT. Server generates JWT to frontend and when frontend receives JWT it will store it to localstorage. Localstorage is safe place to keep JWT as long as you make sure that user can't perform any XSS on your application.
You can also solve CSRF with JWT if you send it in HTTP header. That makes sure that unauthorized requests won't go through on other domains. You should check this PHP JWT library out. I have used it on my own projects.
PHP-JWT
I decided to send back a hashed token made up of the user's (encoded) id, password, and salt. I store their encoded id and token, then send both details to the server when automatically authenticating. Should the user's account be compromised through use of this token, they can change their password to invalidate it.

Separate authentication server for users and APIs

I'm working on a cloud service authentication system and I'm not entirely sure what the optimal way to handle authenticating requests is. We're planning to run our image server as a separate process from our API server so that we can scale them independently of each other. Handling request authentication with API keys is fairly simple, because we can just have the image server store its own API key and check that requests provide it in a header (over HTTPS obviously), same with the API server. For users though it gets more complex.
Right now we have it setup so that the API server will handle generating a session token and storing users in its database, however what we'd like to do is use 3 servers:
authentication server
API server
image server
and have the image and API servers authenticate requests against the authentication server. How exactly should this be done though? It seems like a bad idea performance-wise to hit the authentication server for every request that the API and image servers make. Can/should a token be verified from a different server than it was created on?
So for example: can/should I pass the token received from the authentication server to the image server, verify that the token came from "my.auth.server" and check that the user ID is the right one? Would JWTs be a good type of token for this?
Here's an alternative approach.
Your authentication issues a JWT token that is signed using a secret that is also available in your API and server images. The reason they need to be there too is that you will need to verify the tokens received to make sure you created them. The nice thing about JWTs is that their payload can hold claims as to what the user is authorised to access should different users have different access control levels.
That architecture renders authentication stateless: No need to store any tokens in a database unless you would like to handle token blacklisting (think banning users). Being stateless is crucial if you ever need to scale. That also frees up your API and image servers from having to call the authentication server at all as all the information they need for both authentication and authorisation are in the issued token.
Flow (no refresh tokens):
User authenticates with the authentication server (eg: POST /auth/login) and receives a JWT token generated and signed by the auth server.
User uses that token to talk to your API and image servers and assuming user is authorised), gets and posts the necessary resources.
There are a couple of issues here. Namely, that auth token in the wrong hands provides unlimited access to a malicious user to pretend they are the affected user and call your APIs indefinitely. To handle that, tokens have an expiry date and clients are forced to request new tokens whenever expiry happens. That expiry is part of the token's payload. But if tokens are short-lived, do we require users to authenticate with their usernames and password every time? No. We do not want to ask a user for their password every 30min to an hour, and we do not want to persist that password anywhere in the client. To get around that issue, we introduce the concept of refresh tokens. They are longer lived tokens that serve one purpose: act as a user's password, authenticate them to get a new token. Downside is that with this architecture your authentication server needs to persist these refresh token in a database to make them revokable before they expire (think "revoked" column in tokens table).
New flow (with refresh tokens):
User authenticates with the authentication server (eg: POST /auth/login) and receives a JWT token generated and signed by the auth server, alongside a long lived (eg: 6 months) refresh token that they store securely
Whenever the user needs to make an API request, the token's expiry is checked. Assuming it has not yet expired, user uses that token to talk to your API and image servers and assuming user is authorised), gets and posts the necessary resources.
If the token has indeed expired, there is a need to refresh your token, user calls authentication server (EG: POST / auth/token) and passes the securely stored refresh token. Response is a new access token issued.
Use that new token to talk to your API image servers.
OPTIONAL (banning users)
How do we ban users? Using that model there is no easy way to do so. Enhancement: Every persisted refresh token includes a blacklisted field and only issue new tokens if the refresh token isn't black listed.
Things to consider:
You may want to rotate refresh token. To do so, blacklist the refresh token each time your user needs a new access token. That way refresh tokens can only be used once. Downside you will end up with a lot more refresh tokens but that can easily be solved with a job that clears blacklisted refresh tokens (eg: once a day)
You may want to consider setting a maximum number of allowed refresh tokens issued per user (say 10 or 20) as you issue a new one every time they login (with username and password). This number depends on your flow, how many clients a user may use (web, mobile, etc) and other factors.
You can store some additional metadata (ip, geolocation, device, browser cookie, etc.) alongside refresh tokens. That way, you can be smart about when to reject malicious usages of refresh tokens in case it's compromised.
Common Q: Why store all refresh tokens, and not just revoked ones? You could do that. Ask yourself the following: Will I, at any point, need to have a functionality where I can dynamically revoke valid refresh tokens, based on some arbitrary, regulatory, legal, integrity, security etc. criteria? If so, the least you will need is a list of all issued tokens, plus any data required to implement the criterion logic. Example: Due to regulation, I need to ban all EU users equates to a delete from refresh_tokens were user_ip in <... eu logic ...>
one of the best ways to use is a JWT Token , you can generate and share it between all your servers and validate it on the server side .
https://jwt.io
also I think the best architecture to use in this case is the micro service architecture

Is it nessesarry to send credentials on every single request to MVC Web Api?

I am about to create my first restfull web service where i chose MVC WEB API to be the "provider". After reading about authentication i am a little confused.
My requirements is that on call to any url of webservice i want client to be authenticated, except sign in url.
I understand the flow this way: after client is signed, webservice returns a authenticationtoken which client have to store, and send to the server on every request in the headers. But where is this token stored on the webservice?
I am very confused at which flow of actions i have to implement if i want to avoid users to pass login parameters on every single request.
Typically how this works is that the user's authentication token will be stored in a cookie. Once you authenticate the user, you will create a 'session' for them server side. There will be a 'session token' that corresponds with this session.
When the user signs in, you will create a new session for them. This will send them a new cookie. Every future request they make will contain this cookie. You will then use this cookie to identify the user's session. From it, you can draw their username, etc.
It sounds like what you really want is the .net State management (https://msdn.microsoft.com/en-us/library/75x4ha6s(v=vs.140).aspx). You should look into using this, and see how you can apply it to your current needs.
In the long run, once you've got proper user session tokens, you will not need to send their credentials with every request. The session token will be good enough to identity the user upon every request that they make.

API Token Safety in Angular application

I'm building an Angular app with an API backend. On a combination of pieces of advice, I built the API with a flavor of token authentication. The flow is roughly as follows:
POST to login endpoint with credentials
Validate credentials and authorization, then generate a new token
Return token to client
Client uses token via HTTP Basic to access API resources
This is all working well. The problem arises in creating a session based on this token. I don't believe I should simply hold the token on the client in a cookie, but I do need a session to persist between page refreshes, etc. My Angular app is stateless and completely populated via API calls.
I'm looking for a recommendation as to hanging on to this token on the client. I feel there's danger in holding the token in a cookie because the cookie could be stolen and simply used to authenticate as someone else, but perhaps this is incorrect.
Thanks in advance for your assistance!
The only known way for me to identify a user is to use some token on the client.
HTTP is stateless and can't know which request is coming from which user (browser). You can't identify the user by his ip address (many users are behind a router and share a connection). You could try browser fingerprinting, it can work on some browsers but not on all.
I would recommend using a cookie to store this token on the client.
They are send to the server on every request and you can do some protection to keep them from getting stolen.
To protect this cookie from man in the middle attacks you need to use an encrypted connection over HTTPS to the server.
Set the following attributes on the cookie:
HTTPOnly: cookie can't be accessed by javascript (XSS protection)
Secure: cookie will only be send over https
Path: cookie will only be send on specified path e.g. /login
I would also define an expiration date on the cookie, so the cookie is invalid in like 2 days or something.
But you are right. If this token gets stolen someone else can login as this user.
Since its an Angular app, I'd assume all authenticated methods will only be served to ajax requests (you can tell your server to only respond to ajax) in which case CORS will help you.
The only way to be completely secure is HTTPS, however this method is probably more secure than you think. Read up on CORS a bit for more info, but essentially the idea is that servers will only respond to ajax requests coming from html pages that were served by that same domain.
Pre-flight OPTIONS requests are often sent to verify this. The browser sends an OPTIONS request with an Origin header (the origin of the page) before the actual request. If the origin matches the domain of the server receiving it, the subsequent request is allowed. Otherwise, it violates the Same Origin Policy and will be rejected.
This prevents someone from sniffing out the token and sending a request with the token from a page that your server didn't serve (like something running on the hackers local machine).
If you are doing credit card transactions or anything super secure, you should use HTTPS though.
http://en.wikipedia.org/wiki/Cross-origin_resource_sharing

Password protecting a REST service?

After creating a basic REST service, I've have come to the point where it would be appropriate to add some sort of password protection, as I need to verify that my users are both properly logged and have sufficient permissions to execute whatever action they are going to.
The REST service will mainly be accessed from a Javascript-heavy frontend and with that in mind, I have come up with the two following alternatives to solve this:
Make users login by first sending credentials to a /login page with POST. The page sets a session cookie wherein the user is
marked as logged in, along with the permission level. On each
following request, I verify that the user is logged in and his/her
permission level. When the session expires, automatically or
manually (logout, the user will have to re-logon).
Temporarily save the credentials hashed locally and send the users credentials along every single request made by the user to verify the credentials & permissions backend on a per-request basis.
Are there more ways to solve this and is there something else that I should be concerned with?
I'm currently developing a REST API along with a client (written in javascript), below I'll try to explain the methods used to protect the API against unauthorized access.
Make your REST API to require a Auth-Key header upon every request to the API, besides /api/authenticate.
/api/authenticate will take a username and a password (sent using POST), and return user information along side with the Auth-Key.
This Auth-Key is randomly generated after a call to /api/authenticate and stored in the backend users table with the specific user entry, a md5 hash of the remote ip + the user agent provided by the client.
On every request the value of Auth-Key, and the md5 sum mentioned, is searched for in users . If a valid user is found that has been active during the past N minutes the user will be granted access, if not: http return code 401.
In the REST client, first get the Auth-Key by posting to /api/authenticate, then store this value in a variable and send in on every future request.
If you want to stay true to the definition of a REST service then it should be stateless and not store any login (or other context specific) data on the server: http://en.wikipedia.org/wiki/Representational_state_transfer#Constraints
Your 2nd approach would fit this model
First decide what it is that you're protecting against:
Authentication? (Knowing who is requesting your service?)
Authorization? (Whether a given person can properly request a given service or not?)
I recommend that you provide hashed keys for your service. That way you can manage the key issue separately from the services. Or a client key and a secret, Amazon does this.
It is always easier for the client if you have a stateless protocol. And send everything through the parameters, cookies are a bother for the client too.
Remember that it is in your interest to make it as easy as possible for potential developers to use your service. A super secure service that no one uses is boring.
You can let clients choose the security level by giving them the choice of HTTP or SSL/HTTP endpoints to connect to. Client choice is a good thing.
Make users login by first sending credentials to a /login page with POST. The page sets a session cookie wherein the user is marked
as logged in, along with the permission level. On each following
request, I verify that the user is logged in and his/her permission
level. When the session expires, automatically or manually (logout,
the user will have to re-logon).
Temporarily save the credentials hashed locally and send the users credentials along every single request made by the user to verify the
credentials & permissions backend on a per-request basis.
Your first approach does not meat the statelessness constraint of REST. You cannot maintain client sessions on server side. This constraint makes REST highly scalable...
Your second solution is appropriate. The simplest way to use HTTP basic auth. You don't have to hash the password on client side. What you need is an encrypted connection. On server side you can have an [username, password] -> [identity, permissions] cache, so this solution is much faster and superior in every other way than having server side sessions.
By 3rd party (non-trusted) clients the authentication is more complex, I guess you don't need that part.
I'm no security-expert. I use the RESTful Play!-webframework and they do the following things to authenticate users.
The cookie is protected against manipulation. It is signed with a long secret key and is checked for each request. Just hashing it is not enough!
They recommend to set unique information the identify the user in the cookie. As the server should be the only one to manipulate the cookie, that's enough.
Don't put the password as credential in the cookie. If someone sniffs the cookie, not only the session can be hijacked, but also the complete account or even worse, other accounts with the same credentials.
If you want to protect the cookie against hijacking using https.