I'm trying to understand cookie vs token-based authentication. I get the basics - cookies have to be stored in the server (as well as the client) to be verified each time, whereas tokens only have to be stored on the client-side. The server simply decodes any incoming tokens and verifies the request.
What I don't understand is, how does the server know whether any decoded token is valid, if a list of valid tokens isn't being stored anywhere in the server?
https://dzone.com/articles/cookies-vs-tokens-the-definitive-guide
User enters their login credentials.
Server verifies the credentials are correct and returns a signed token.
This token is stored client-side, most commonly in local storage - but can be stored in session storage or a cookie as well.
Subsequent requests to the server include this token as an additional Authorization header or through one of the other methods mentioned above.
The server decodes the JWT and if the token is valid processes the request.
Once a user logs out, the token is destroyed client-side, no interaction with the server is necessary.
To be specific - I'd like to know how #5 works. Thank you.
A full answer to your question would be very long, but here is my attempt at a brief one. The server also happens to be the entity which issues the JWT in the first place. As such, it possesses a key which it used to sign each outgoing JWT. As a result of this, when the server receives an incoming JWT, it first tries to open/unlock that JWT using its private key. If the JWT were tampered with in any way, the server might not be able to open it properly, and it would result in an exception. As an example of one sanity check the server would perform against an incoming JWT, it would observe the checksum of the JWT, would not pass in the case of tampering.
Once the server has opened the JWT and deemed it to be valid, the next thing it would likely check would be the exp claim, one of possibly several claims which are contained within the JWT. The exp, or expiry, claim, records for how long the JWT is valid. Should a user present a stale JWT, the server would immediately reject the JWT as being invalid.
So far, we have been discussing checks the server may perform using no state, that is, relying only on the state contained within the JWT itself. In reality, most of the time the server would in fact be storing some state of its own. As an example of why this might be necessary, consider the edge case of a user who logs out of the site or app. In this situation, his phone or browser would still be bearing a JWT with a valid exp expiry. In order to prevent the user from continuing to use this JWT, the server might maintain a blacklist of JWT which it will not honor, even if the exp and checksum pass inspection. So, a third step after unlocking the JWT and checking exp might be to make sure that the JWT does not appear on the blacklist.
A good JWT implementation can limit the amount of server side state to something very small, but typically the server would actually maintain some state of its own.
I know this might seem like a trivial question but I can't find the answer for it to at least put my mind at peace.
If a mobile app is communication with a server then typically they sign in and they get an access token that they can use for the remainder of the session with any future request.
WHY not just pass the user name and password over HTTPS with every request instead of the access token. An access token will need to be verified with database and so is the combination of username/password. Why go through the added effort of access token if they do the same thing? I know I am missing something but I can't figure it out
You are right in that an access token is verified pretty much the same way as a username and password. For the period when the access token is valid, it is pretty much equivalent to the username and password. In some cases (depending on your threat model) it may even be ok to send username and password with every request, maybe not from a mobile app, but for example in server to server requests, with appropriate controls.
However, you don't want to send the password in every request from a mobile app primarily because then you would have to store it.
The problem with a password (or with users, actually) is that they are reused. A password is a valuable target, because the same password is likely to be used on multiple services. So you exchange it for a shorter-lived access token, which, if stolen, presents less risk for your user. And you can't easily revoke a password - forcing users to change their passwords is a hassle. Revoking an acces stoken is easy.
Also it's very unlikely, but sometimes there are vulnerabilities in TLS - not very often, but there have been a few in the past years. Such a vulnerability might allow an attacker to decrypt some of the traffic sent over https, or for example there was a vulnerability in openssl a while ago that allowed attackers to extract parts of the server memory, potentially holding whatever the user(s) sent. It's much better if it's just an access token, and not the actual password.
Another point is sometimes (in OAuth flows) your app won't be allowed to even access the actual password. When your user logs in with a 3rd party identity provider (like for example facebook), they would not like your app to receive their facebook password. They just go to facebook, exchange their credentials for an access token, and you only see the token which you can verify with facebook if you want. But you never actually get the user's facebook password. Of course this is only valid when the identity provider is a third party.
I think the answer is all about the security and safety.
IT'S ALWAYS RECOMMENDED to use access tokens instead of username & password, because:
Access tokens (in most services) can be easily generated, blocked, monitored for their usage & statistics from your account, can be set as expirable, can have restricted permissions, and so on... Of course, you can delete it at all. The Username & Password is the master, who can control the access tokens.
Access tokens are safer as I said, and that is the most important thing. If you use Username & password in your web-application (or whatever) and that application is hacked (which happens so frequently), or someone views its source, or even some log-system saves the request parameters - then your user & password can be viewed by 3-rd parties and all your account can be hacked (and probably others too, where you have same user/pw). Access tokens eliminate all these risks.
About speed - I don't think that authorization with USER & PW has any significant advantage in speed.
Let's consider this use case:
1) I call my API login endpoint with username and password and get my Auth token that I add to every consecutive request to the header as Authorization: Bearer <token>.
2) I call /current-user endpoint with no params, only with Authorization header. Server authorizes user using token and gets user's id from that token. Then he finds the user by id in database and returns it's data.
My question is, whether this approach isn't insecure. I'm wondering, what if I was an attacker and was calling /current-user endpoint using randomly generated tokens. Once I occasionally matched real token, the server would return me other user's data.
Isn't it necessary to store user id on client along with token and call requests using both? Eg. /user?id=<stored user id> with Authorization header and get rid of /current-user endpoint? After that some kind of ACL on server would determine, whether used token has allowed access to user with passed user id.
(I also found there are JWT tokens but I see the same problem there. As an attacker I would somehow manage to guess other user's token and server would return me his data)
For security purposes, it is normally assumed that the user id is known to the attacker anyway. For example, if the attacker already has or knows a legitimate account, she might be able to guess how other user ids are assigned.
Also, if your token is long enough and totally random, it doesn't really make any difference.
Look at it this way: let's say your token has length n, your user id has length m. Without the user id, the attacker has to guess n characters, with it included she has to guess n+m characters. If your n is high enough, you don't need those extra characters. Keep in mind that the effective length of the user id might be much shorter than its apparent length if your user ids aren't completely random, so the m added might actually be really small.
Is it possible for you to use digitally signed tokens? You could basically encrypt the token with the client's (or user's) private key and then encrypt the package along with plaintext userid again with the server's public key. That way only the server can decrypt the package and once decrypted, it knows who the user is. It can then use the public key associated with that userid to decrypt to the package and get the token.
As you havent provided much info on the application and the speed of authentication required or technologies you are using, it is difficult to provide any more info.
I have built a simple test API for a CakePHP application that will let a user login from a mobile device (or any device for that matter) and get a JSON response. This API could be used for a mobile app built in PhoneGap.
The login method looks like so:
public function login()
{
if($this->request->is('post'))
{
// Use custom method in Model to find record with password params
$findUser = $this->User->findUser(
$_POST['username_or_email'],
AuthComponent::password($_POST['password'])
);
// If a user exists and matches params
if($findUser)
{
$this->User->id = $findUser['User']['id'];
$this->autoRender = false;
$this->response->type('json');
$this->response->body(json_encode(array('authenticated'=>true,'message'=>__('You have been logged in successfully'))));
}
else
{
$this->autoRender = false;
$this->response->type('json');
$this->response->body(json_encode(array('authenticated'=>false,'message'=>__('Username or password is incorrect'))));
}
}
else
{
$this->autoRender = false;
$this->response->type('json');
$this->response->body(json_encode(array('message'=>'GET request not allowed!')));
}
}
The mobile device (or any API user) can send their login details and then they get the request as JSON as true or false for authenticated. This boolean is NOT used to give the user access, it instead tells the mobile app if they can see certain screens and they ONLY get the data or can send data if the session exists!
As just stated, they are also actually logged into the API itself on the device so if they visit the website directly (from that device) they will have a session and see the same response for the JSON.
So essentially a user remains logged in for the duration of the session on the device they communicated with the server on. This is different to a token which would need to be passed for every request, where as in this example they have a session.
Now the questions...
Is it bad practice for the user to be 'actually' logged into the API
with a session like shown above? It seems like the most secure way to handle authentication for a device as it's using the same logic as the direct web root.
I've seen some APIs use access tokens instead which I've also
implemented (user gets their token returned instead of the boolean
and no session is created). But from what I can tell, this seems
like more work as then I need to check for the access token against
a user record every time a request is made.
edit
For the sake of clarity, I am not a supporter of REST, I AM a supporter of RESTful/RESTlike services. If you look at all of the API's on the internet, very few actually stick to one standard. Whatever scheme you choose will depend on your specific problem-space. Just try to be secure and use intuitive design choices (ie dont name a service "cats" if it returns info about "dogs")
end edit
It is good practice in RESTful API's to manage some form of session/tokenizing scheme. Really the ideal (at least in my opinion, there are many schools of thought on this problem) setup involves rolling tokens.
If you are at all concerned with the security of your API, then permissions should be managed out of your database layer. Yes, this creates a bottleneck, BUT THAT IS ACTUALLY A GOOD THING. Needing to hit the database every single time to validate a client's token adds an extra step in the entire process. This slows down the API, which is actually desireable in a secure system. You don't want a malicious individual to be able to hit your API 3000 times a second, you want their requests to hang for a (somewhat) sizeable fraction of a second.
This is similar to MD5 hashing algorithms. Many of them recalculate the hash a few hundred times, with random pauses in between. This helps to keep a malicious client from attempting to brute force a password (by making it take more time to test each variation of the password string). The same applies to your API.
The other benefit, is that if you DO have a malicious user trying to log in over and over again, if you are managing them from the database layer, then you can red flag their IP Address/username/what-have-you and just drop their requests at step 1.
Anyway, for a suggested process (with rolling tokens, you can cut out parts of this if it seems overkill, but this is hella secure):
User hits a 'login' service, this requires a username/password, and returns two tokens, a Private Access Token and a Public Request Token (the server stores these tokens in the db).
The client stores these Tokens in a secure place
User accesses another endpoint to push/pull some data
Request includes a timestamp
Request includes the Public Request Token
Request includes an Access Token=> This token should be a MD5 hash of the string resulting from concatenating the timestamp string to the end of the Private Access Token string
The server takes the Public Request Token, uses that to lookup the Private Access Token that was stored
The server takes that Private Access Token, and concatenates on the Timestamp String, it then takes the MD5 of this string
If the new Access Token matches the one that the client sent the server, HURRAY, this client is validated, so push/pull the data
(Optional) The server generates new tokens on every request, and returns them to the client. This way every transaction invalidates the old tokens, and if there was some kind of man-in-the-middle attack occurring, if the VALID user has already completed their request, the malicious user now has invalid tokens and can't start messing with your API. This scheme tries to ensure that a malicious user can not expect to intercept a single communication between the server and the client, and still gain access to the system. If they do, then the REAL user should immediately get invalidated tokens. Which should then trigger their API client to hit the 'login' service AGAIN, getting new valid tokens. This once again kicks the malicious user out of the system.
This scheme is not 100% secure, no user access system ever will be. It can be made more secure by adding expiration dates on tokens. This scheme also has the added benefit that you can assign specific permissions to users/tokens (ie Read-Only access, only certain End-Points can be seen, etc)
This is not the only way you can do things, I would look up other Authentication Schemes and take what you want from each of them (OAUTH is a good place to start, then I'd look at Facebook/Twitter/Instagram)
Make your app login everytime, but not with login-pass pair as Swayok lastly suggested. When you login, server generates a token and returns it back to the client. Client then uses this token whenever it makes a request. On each request, server checks whether the token is valid and if so, executes the request.
This is very similar to how sessions work in that, server side frameworks manage it internally and these tokens expire from time to time. However, as Swayok rightuflly pointed out, you don't want session mainly because you're RESTful API should have no state. You get the same utility without storing any user specific data regarding user and logging user in with every request.
Here's a good article on this, or you can try the Facebook Graph API explorer to see it in action
Restful API restricts using sessions and saving system state at all. Each request must log-in user.
Access tokes are great but also require additional handling.
The easiest way is to send authorisation data via HTTP Basic Auth ("Authorization" HTTP header)
http://www.httpwatch.com/httpgallery/authentication/
Mobile Applications can easily do that and it is easy to add this header for each request to API.
On server side:
$username = env('PHP_AUTH_USER');
$password = env('PHP_AUTH_PW');
And process user log-in with this data in ApiAppController->beforeFilter()
To answer your questions
Its not a bad practice as long as you close their session on app close and recreate it when needed. it is same as if they were logged in on a browser they would know and have facility to log out however the same should be available on the app as well otherwise they might have closed the app but not actually ended their session. You can handle this in many ways by asking them to log out automatic checking when they close app
Tokens are an enhanced way of doing the above however you have to consider how secure the token is when transmitted and server need to verify the token on each request. You have said that it seems like more work so yes its more work and if you have time or money constrains and looking for an answer to say if the session style would harm your application in future it wont as long as you are in control of session and not leaving user without ending the session. If you have time then implement tokens and you would like that.
What happens when you log into a website?
I know cookies are stored and some info (what info?) gets sent to the server...but maybe some more detail?
Once upon a time, somewhere on the Internet....
Browser: "hey, can I see this web page?, Trouble is, I don't remember speaking to you before"
Website: "sure, fill in the form, I need your username and your password"
Browser: "Here ya go"
Website: "Great!, Welcome back koldfyre! Here's the page. Look, if you want more pages, take this token and use it when you ask for another one"
Browser: Cool. That site gave me a token. I'll memorize it!
A few minutes later
Browser: "Ooh! Can I see this other web page? Here's my token"
Website: "That token looks good, hello again koldfyre, here's your web page"
That, in essence, is it. In order to remember a user has logged in, it gives the user a token which it must present with its next request. This is normally achieved by the server telling the browser to store this token in a cookie.
Delving deeper - transport layer authentication
The way the credentials are passed to the server, and the nature of the token it returns, vary depending on the authentication method employed.
The very simplest, HTTP Basic Authentication, is provided by most web server implementations. It causes your browser to pop open the familiar login dialog box. The "token" is simply your plaintext username and password base64 encoded. Not particularly secure.
A server might also provide Digest Authentication, which avoids transmission of the actual credentials - the client instead generates a hash of their credentials with a server generated salt. It is designed to prevent password sniffing and replay attacks.
Deeper still - application layer authentication
For maximum flexibility and control, most sites opt to implement the authorization in the application layer rather than the HTTP transport layer. This gives a wider variety of security choices. Any site which asks for credentials in a web page (rather than the browser's login dialog box) is using a custom authorization method.
Custom methods will vary wildly in their initial interaction, but they almost always result in the user being given a session cookie containing a randomly generated identifier. The browser then automatically presents the cookie with each subsequent request. The web application will inspect the cookie value to ensure it is still valid.
It's also possible to hand off the authorization to a trusted third party, generally to provide some sort of single-signon service. In cases like that, when you notice a user isn't authenticated, you bounce them off to the authentication provider. They authenticate them, and they will be returned to you with some kind of token you verify with the provider. Shibboleth is one example of this. You also logged into this very site using a similar method employed by OpenID
Further reading
Here's some nice answers from a similar question
PART I: How To Log In
PART II: How To Remain Logged In - The Infamous "Remember Me" Checkbox
PART III: Using Secret Questions
PART IV: Forgotten Password Functionality
PART V: Checking Password Strength
PART VI: Much More - Or: Preventing Rapid-Fire Login Attempts
PART VII: Distributed Brute Force Attacks
Other answers in that question provide even more links to round out your education!
That's a pretty general question. What you're doing, over all, is establishing some kind of credentials with the site itself. If we take the simple version, you enter a user name and a password; that means you identify yourself to the website, and then show it a secret you and the website share that no one else knows (we hope). That establishes you as authentically the person with that user name, and so we say you have authenticated yourself.
Once you've done so, there are some design decisions the website designer has to make. most people don't want to log in for every page, so the web site wants to store a little information, a credential, on your end. This means that it can tell it's still you. Often, as you say, that's a "cookie", which is nothing more that a tiny text file named with the web site's URL. This file is stored by the browser.
On many web sites, like for banking, you also want to guarantee that the data being exchanged can't be intercepted by a third party. If so, you establish a secure connection using a protocol known as SSL or TLS. What this adds to the basic connection is an exchange of information that establishes a session key. This session key is then used to encrypt the communications. This usually happens before you exchange the user name and password, so that your password is never visible to a malicious third party either.
Under the covers, when you establish a secure connection, the web site sends your browser a block of formatted data called an x509 certificate. This is another form of authentication; the certificate will have been signed by an issuer (the certificate authority or "CA") and the browser can use stored data about the CA's to ensure that the certificate is authentic.
This completely depends on the implementation of the website. Even the usage of cookies is not mandatory, but very common.
In most cases however, something like this happens:
You send in your username and password using an HTML form.
The server looks up the relevant user (using a database)
The server checks if the password matches the password that is stored in the database alongside the user.
If the password is correct, the server will store what user currently is active in the session. The identifier of this session is stored in a cookie, the actual data of this session (the current user) is stored on the server under this identifier.
Now you are logged in. You will remain logged in during the remainder of the session:
When you request another page from the server, you will send the cookie with the sesison identifier.
The server loads the session using this identifier. In this session, the current user is stored, so the server knows what user is logged in.
When you log into a web site, first your credential are authenticated. If your credentials match, then something is put into the session (on the server) to keep track of who you are so you can access data that is yours without having to re-log-in. This is obviously useless on the web server unless the client can provide information about who it is on each request. Note that the "Session" is usually maintained entirely on the web server, with the client having only a key that allows access to the session.
Remember that HTTP itself is a stateless protocol. The HTTP standard contains no method for HTTP requests to keep or persist any state between individual HTTP requests. Thus, the state is usually kept entirely on the server and you just need a method for the client to identify which session the current HTTP request belongs to.
The two common ways this is done are:
Use a cookie (for example, Apache Tomcat uses the JSESSIONID cookie) to store some hashed authentication token that will successfully look up the web session, or
rewrite the URL so that every request has the session ID added to the end of the request. Still using Apache Tomcat as the example, if cookies are disabled then the URL will be rewritten to end with a string like ";jsessionid=....". Thus, every request, every HTTP GET and POST (and the rest) will end with this string.
Thus, on each request the client makes, the session ID is provided to the web server, allowing the persisted state for this client to be quickly looked up, allowing HTTP to act like a stateful protocol.
What information is sent to the server when you log in? Whatever information you provided on the login form. Some web servers also track the TCP/IP address the request came from to avoid session hijacking attacks. This is usually all the information that is needed by the server.
If you don't allow your browser to save cookies, then you will have to log in to the web server each time you open your browser and initially open the server's web page. However, if you allow your browser to save cookies, then many servers allow you the option of saving the cookie (that is, not just using a session cookie) so that each time you go to a web page of the server, the persisted cookie will identify you so you don't need to re-login. Here, the cookie will save enough information -- often in an encrypted form that only the server can understand -- to identify you. In this case, the Cookie is not a simple session ID.
Very simply explained, what happens is mentioned below:
What goes in?
Username
Password
What happens inside?
Password is converted to its hash
Hash(password) is compared with the DB table or a Directory Service
(unless someone is down-rightly foolish, the site won't save your password in clear text)
If Authenticated, A status-token is stored in Session and/or cookie.
This token can just contain a status, Login Timestamps, your userId, userType(if any), et al.
This token is read and verified on every page you access if that page requires you to be logged with as a certain type of user.
If authentication fails, you are redirected to a page displaying error asking you to re-login.
What comes out
You are redirected your personal profile page/the page you were accesing to which verifies you with the help of the token.
Additionally, a Digital Certificate may come in picture if you are accessing a banking site or other critically secure site
There are two main ways of performing authentication on the web, and a few less popular ways that are also worth knowing about.
The first is HTTP authentication, as defined by RFC 2617. When you request a protected page, the server responds with a 401 status code, signalling that you aren't permitted to access the resource. In addition to this, it also sends a WWW-Authenticate header, which instructs the browser on how it wants you to authorise yourself. The browser sees this status code and the header, and prompts you for your authentication details. When you enter them, your browser prepares them according to the specific authentication scheme the server specified, and requests the page again, including an Authorization header with the prepared details. The server checks these details against its user database, and either responds with another 401 (wrong details), or the protected page with an accompanying 200 status code to indicate success.
HTTP authentication is one of those ancient features that browsers didn't implement well to begin with and have never really been improved. Because of this, it has become much more popular for web developers to implement authentication themselves using cookies to persist state. In this case, the user is presented with a standard HTML form. When the user enters their credentials into the fields and submits the form, the browser encodes it and sends it to the server in the same way it encodes any normal HTML form. The server checks the credentials, and if they are legitimate, sets a cookie with a randomly-generated ID number, along with a corresponding database/filesystem entry that recognises that ID number as belonging to a particular user.
From this point on, every request the browser makes to the server includes this ID number cookie as an HTTP header. The server recognises the cookie, looks up the ID number, and knows which user you are. When you choose to log out, the server sends a response asking your browser to forget the ID number, at which point you are just another anonymous user.
A less commonly-used option is the use of SSL client certificates. Many people are familiar with the idea of using SSL to identify a server. A cryptographic keypair is generated, signed by a trusted authority, and used to prove that the data being sent originated with the owner of the keypair. What many people aren't aware of though, is that the same can be used by a client to prove its identity to a server. This is less convenient, however, as you need to carry your certificate around with you if you want to use it on more than one machine.
There are variations and lesser-known options available of course, but these are the most prominent ones.
As others have mentioned, login procedures vary depending on implementation, but the basic case (simple web app authentication) uses something like the following pseudocode:
function login(username, password) {
user = db->get_user(username)
if (user == false) {
report_error("Unknown username")
exit
}
if (user->password != hash(password)) {
report_error("Incorrect password")
exit
}
// User authenticated, set session cookie
session->set_data('current_user', user->username)
}
Of course, in most cases, it gets a little more involved than that, but every login function starts its life looking essentially like the above. Now, if we add autologin ("remember me") to the mix, we get something like this:
function login(username, password, remember_me) {
user = db->get_user(username)
if (user == false) {
report_error("Unknown username")
exit
}
if (user->password != hash(password)) {
report_error("Incorrect password")
exit
}
// User authenticated, set session cookie
session->set_data('current_user', user->username)
if (remember_me == true) {
cookie_token = random_string(50)
set_cookie('autologin_cookie', cookie_token, ONE_MONTH)
// Finally, save a hash of the random token in the user table
db->update_user(user, 'autologin_token', hash(cookie_token))
}
}
Plus the function to perform the automatic login if there is a cookie present:
function cookie_login() {
cookie = get_cookie('autologin_cookie')
if (cookie == false) {
return false
}
// Only for demonstration; cookie should always include username as well
user = db->get_user_by_cookie(cookie)
if (user == false) {
// Corrupt cookie data or deleted user
return false
}
// User authenticated, set session cookie
session->set_data('current_user', user->username)
return true
}
NOTE: The above isn't a 'best practices' approach, and it's not very secure. In production code, you would always include a user identifier in the cookie data, use several levels of throttling, store data on failed and successful logins, etc. All of this has been stripped away to make the basic structure of authentication simple to follow.
Anyway, I hope this is what you were looking for, koldfyre. I don't know your background, but if you're unsure of how sessions and cookies work, you should read up on them separately, and if you need more elaborate details, just ask.
P.S.: You may also want to check the question "The Definitive Guide To Website Authentication" for best practice approaches
Look, it's a little hard to give you a lot more information that you already have here; I'm not sure why you want to set a bounty on it. A cookie is just a little bit of named information, and you can put anything you like in it. For a session, you'd want some kind of session ID. There are conventions for that, or you can do it yourself. Whatever you do, when you set the cookie, you leave a little data lying about on the person's browser that is more or less like this:
mydomain.com:
mystuff: this is my stuff, by golly.
When you come back, you retrieve the cookie and get that back.
If you want to see all the details of that protocol, have a look at the Wikipedia article.