JWT Services
Last updated
Last updated
CBSecurity also provides you with a JWT (Json Web Tokens) authentication and authorization system.
JSON Web Token (JWT) is an open standard (RFC 7519) that defines a compact and self-contained way for securely transmitting information between parties as a JSON object. This information can be verified and trusted because it is digitally signed. JWTs can be signed using a secret (with the HMAC algorithm) or a public/private key pair using RSA or ECDSA.
Signed tokens can verify the integrity of the claims contained within it, while encrypted tokens hide those claims from other parties. When tokens are signed using public/private key pairs, the signature also certifies that only the party holding the private key is the one that signed it.
You can find much more information about JWT at jwt.io.
JSON Web Tokens have become the standard for authenticating and authorizing API requests. They can be used on their own or with an oauth/single sign-on server as well.
Authorization: This is the most common scenario for using JWT. Once the user is logged in, each subsequent request will include the JWT, allowing the user to access routes, services, and resources that are permitted with that token.
Information Exchange: JSON Web Tokens are a good way of securely transmitting information between parties. Because JWTs can be signed—for example, using public/private key pairs—you can be sure the senders are who they say they are. Additionally, as the signature is calculated using the header and the payload, you can also verify that the content hasn't been tampered with.
The ColdBox Security module will assist you with all the generation, decoding, encoding and security aspects of JWT. All you need to do is, configure it, create a few standard files and off you go.
The tokens created by the JWT services will have the mandatory headers, but also will have a standardizes payload structure. This payload structure can also be customized as you see fit.
A JSON Web Token encodes a series of claims in a JSON object. Some of these claims have specific meaning, while others are left to be interpreted by the users. You can consider claims to be the keys of the payload structure and it can contain, well, pretty much anything you like.
Here are the base claims that the ColdBox Security JWT token creates for you automatically:
Issuer (iss
) - The issuer of the token (defaults to the application's base URL)
Issued At (iat
) - When the token was issued (unix timestamp)
Subject (sub
) - This holds the identifier for the token (defaults to user id)
Expiration time (exp
) - The token expiry date (unix timestamp)
Unique ID (jti
) - A unique identifier for the token (md5 of the sub and iat claims)
Scopes (scope)
- A space delimited string of scopes attached to the token
Refresh Token (cbsecurity_refresh
) - If you are using refresh tokens, this custom claim will be added to the payload.
You can add much more to this payload via the JWT service methods or via the User that models the token.
The service can be found here cbsecurity.models.JWTService
and can be retrieved by either injecting the service (JwtService@cbsecurity
) or using our helper method (jwtAuth()
).
In order to begin exploring the JWT capabilities, let's explore how to configure it first.
Our JWT services have several configuration settings, let's explore them:
The WireBox Id of the service to provide our authentication. cbauth is our default provider, but you can use any authentication service that adheres to our interface.
The WireBox Id of the service to provide our user retrieval and validation functions. You can use any service that adheres to our interface.
The default variable name in the prc
scope that will be used to store an authenticated user object if the JWT request is valid. The default is prc.oCurrentUser
The issuer authority for the tokens, placed in the iss
claim of the token. If empty, we will use the event.buildLink()
to create the issuer. By default, our validators also check that tokens are created by the same issuer.
The secret key is used to sign the JWT tokens. By default it will try to load an environment variable called JWT_SECRET
, if that setting is also empty, then we will auto-generate a secret token that will last as long as the ColdFusion application scope lasts. So technically, your secret will rotate only if a secret is not specified.
Also, this key is ignored in modules. To specify a fixed key to be used in your modules, you will have to configure it by adding a cbsecurity key settings in the moduleSettings structure within the config/Coldbox.cfc.
Your secret key will auto-rotate every application scope rotation. Please note that all tokens used after that scope rotation will automatically become invalid.
Please note that we use the jwt-cfml library for encoding/decoding tokens. Please refer to it's documentation in order to leverage RS and ES algorithms with certificates.
By default, our jwt services will look into the authorization
header for a bearer token. However, it can also look in a custom header by this name, which defaults to x-auth-token
. Finally, if not found, it will also look into the rc
scope for a rc[ 'x-auth-token' ]
as well.
The default expiration in minutes for the JWT tokens. Defaults to 60 minutes
The encryption algorithm to use for the tokens. The default is HS512, but the available ones for are:
HS256
HS384
HS512
RS256
RS384
RS512
ES256
ES384
ES512
In the case of the RS
and ES
algorithms, asymmetric keys are expected to be provided in unencrypted PEM or JWK format (in the latter case first deserialize the JWK to a CFML struct). When using PEM, private keys need to be encoded in PKCS#8 format.
If your private key is not currently in this format, conversion should be straightforward:
When decoding tokens, either a public key or certificate can be provided. (If a certificate is provided, the public key will be extracted from it.)
This is an array of claim names that each token MUST have in order to be authenticated. If a token comes in but does not have these claims in the payload structure, it will be deemed invalid.
By default, our JWT services will store tokens in CacheBox for you in order to be able to invalidate them. We ship with two providers for token storage: db and cachebox.
By default the token storage is enabled.
The key prefix to use when storing the keys in the permanent storage. Defaults to cbjwt_
The driver to use. Can be either db or cachebox or your own WireBox Id for using a custom storage.
A struct of properties to configure each storage with.
Refresh tokens have several configuration items, check them out in our refresh token configuration section.
The next step is to make sure that our JWT services can handle the construction of the JWT tokens as per YOUR requirements. So your User
object must implement our JWTSubject
interface with the following functions:
Basically, it's two functions:
getJwtCustomClaims( payload )
- This is a struct of custom claims to incorporate into the token payload at construction time. This can be ANYTHING you like.
getJwtScopes()
- We will also call this at construction time in order to incorporate the right permission scopes into the token according to your user. This must be an array of scopes/permissions.
Since also the authentication services will be used with JWT, your user object might end up looking like this:
Please note that the JWT validators must talk to the authentication and user services. Please refer to the Authentication Services page to configure and create them.
Ok, now we can focus on all the wonderful methods the JWT service offers:
attempt( username, password, [ customClaims:struct ] ):token
- Attempt to authenticate a user with the authentication service and if successful, return the token using the identifier and custom claims. Exception if invalid authentication
fromUser( user, [ customClaims:struct ] ):token
- Generate a token according to the passed user object and custom claims.
encode( struct payload ):token
- Generate a raw jwt token from a native payload struct.
verify( required token ):boolean
- Verify a token string or throws exception
decode( required token ):struct
- Decode and retrieve the passed in token to CFML struct
parseToken( token, storeInContext, authenticate ):struct
- Get the decoded token using the headers strategy and store it in the prc.jwt_token
and the decoded data as prc.jwt_payload
if it verifies correctly. Throws: TokenExpiredException
if the token is expired, TokenInvalidException
if the token doesn't verify decoding, TokenNotFoundException
if not found
getToken():string
- Get the stored token from prc.jwt_token
, if it doesn't exist, it tries to parse it via parseToken()
, if not token is set this will be an empty string.
getPayload():struct
- Get the stored token from prc.jwt_payload
, if it doesn't exist, it tries to parse it via parseToken()
, if not token is set this will be an empty struct.
setToken( token ):JWTService
- Store the token in prc.jwt_token
, and store the decoded version in prc.jwt_payload
authenticate( [payload] ):User
- Authenticates a passed or detected token payload and return the user it represents
getuser()
- Get the authenticated user according to the access token detected
logout()
- Logout a user and invalidate their token
invalidateAll( async:false )
- Invalidate all access and refresh tokens in permanent storage
invalidate( token )
- Invalidates the incoming token by removing it from the permanent storage.
isTokenInStorage( token )
- Checks if the passed token exists in permanent storage.
getTokenStorage( force:false )
- Get the current token storage implementation. You can also force create it again if needed.
attempt( username, password, [ customClaims:struct ] ):struct
- Attempt to authenticate a user with the authentication service and if successful, return a struct containing an access and refresh token.
fromUser( user, [ customClaims:struct ] ):struct
- Generate a struct of refresh and access token according to the passed user object and custom claims.
That's it, we are ready to put it all together. Now cbsecurity knows about your authentication/user services, can talk to your user to create tokens and can guard the incoming requests via the JWT Validator. Here is a sample controller for login, logout and user registration:
Let's configure some routes first:
Then build out the Auth
controller
Make sure you add validation!
That's it, we now can login a user, give them a token, register a new user and give them their token, and also log them out. The next step is for you to build your rules and/or security annotations and make sure the JWT validator is configured for your global app or module.
In order to implement JWT authentication in your application, you may need to modify some web server settings. Most web servers have default content length restrictions on the size of an individual header. If your web server platform has such default enabled, you will need to increase the buffer size to accommodate the presence of JTW tokens in both the request and response headers. The size of a JWT token header, encrypted via the default cbSecurity HMAC512 algorithm, is around 44 kilobytes. As such you will need to allow for at least that size. Below are some examples for common web server configurations
The following configuration may be applied to the main NGINX http
configuration block to allow for the presence of tokens in both the request and response headers:
You will need to modify two registry keys:
HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\HTTP\Parameters\MaxFieldLength
- Sets an upper limit, in bytes, for each header. The default value is 65534 bytes and the maximum value is 65534 bytes ( 64kb )
HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\HTTP\Parameters\MaxRequestBytes
- Sets the upper limit for the request line and the headers, combined. As such 128K should allow for both long URLs, as well as JWT tokens in the headers. The default value is 16384 bytes and the maximum value is 16777216 bytes ( 16 MB )
You will need to add a LimitRequestFieldSize
setting in each <VirtualHost...>
entry in order increase the default header size from the default 8 kilobytes. Example, with a setting of 128 kilobytes: