Basic rate-limiting middleware for Express. Use to limit repeated requests to public APIs and/or endpoints such as password reset.

A brute-force protection middleware for express routes that rate-limits incoming requests, increasing the delay with each request in a fibonacci-like sequence.

Here is the full control functions;

  • store An instance of ExpressBrute.MemoryStore or some other ExpressBrute store (see a list of known stores below).
  • options
    • freeRetries The number of retires the user has before they need to start waiting (default: 2)
    • minWait The initial wait time (in milliseconds) after the user runs out of retries (default: 500 milliseconds)
    • maxWait The maximum amount of time (in milliseconds) between requests the user needs to wait (default: 15 minutes). The wait for a given request is determined by adding the time the user needed to wait for the previous two requests.
    • lifetime The length of time (in seconds since the last request) to remember the number of requests that have been made by an IP. By default it will be set to maxWait * the number of attempts before you hit maxWait to discourage simply waiting for the lifetime to expire before resuming an attack. With default values this is about 6 hours.
    • failCallback Gets called with (req, resp, next, nextValidRequestDate) when a request is rejected (default: ExpressBrute.FailForbidden)
    • attachResetToRequest Specify whether or not a simplified reset method should be attached at req.brute.reset. The simplified method takes only a callback, and resets all ExpressBrute middleware that was called on the current request. If multiple instances of ExpressBrute have middleware on the same request, only those with attachResetToRequest set to true will be reset (default: true)
    • refreshTimeoutOnRequest Defines whether the lifetime counts from the time of the last request that ExpressBrute didn’t prevent for a given IP (true) or from of that IP’s first request (false). Useful for allowing limits over fixed periods of time, for example: a limited number of requests per day. (Default: true). More info
    • handleStoreError Gets called whenever an error occurs with the persistent store from which ExpressBrute cannot recover. It is passed an object containing the properties message (a description of the message), parent (the error raised by the session store), and [key, ip] or [req, res, next] depending on whether or the error occurs during reset or in the middleware itself.

Install

$ npm install –save express-rate-limit

Usage

For an API-only server where the rate-limiter should be applied to all requests:

var RateLimit = require(‘express-rate-limit‘);
 
app.enable(‘trust proxy‘); // only if you’re behind a reverse proxy (Heroku, Bluemix, AWS if you use an ELB, custom Nginx setup, etc)
 
var limiter = new RateLimit({
  windowMs: 15*60*1000, // 15 minutes
  max: 100, // limit each IP to 100 requests per windowMs
  delayMs: 0 // disable delaying – full speed until the max limit is reached
});
 
//  apply to all requests
app.use(limiter);

For a “regular” web server (e.g. anything that uses express.static()), where the rate-limiter should only apply to certain requests:

var RateLimit = require(‘express-rate-limit‘);
 
app.enable(‘trust proxy‘); // only if you’re behind a reverse proxy (Heroku, Bluemix, AWS if you use an ELB, custom Nginx setup, etc)
 
var apiLimiter = new RateLimit({
  windowMs: 15*60*1000, // 15 minutes
  max: 100,
  delayMs: 0 // disabled
});
 
// only apply to requests that begin with /api/
app.use(‘/api/‘, apiLimiter);
 

Create multiple instances to apply different rules to different routes:

var RateLimit = require(‘express-rate-limit‘);
 
app.enable(‘trust proxy‘); // only if you’re behind a reverse proxy (Heroku, Bluemix, AWS if you use an ELB, custom Nginx setup, etc)
 
var apiLimiter = new RateLimit({
  windowMs: 15*60*1000, // 15 minutes
  max: 100,
  delayMs: 0 // disabled
});
app.use(‘/api/‘, apiLimiter);
 
var createAccountLimiter = new RateLimit({
  windowMs: 60*60*1000, // 1 hour window
  delayAfter: 1, // begin slowing down responses after the first request
  delayMs: 3*1000, // slow down subsequent responses by 3 seconds per request
  max: 5, // start blocking after 5 requests
  message: “Too many accounts created from this IP, please try again after an hour“
});
app.post(‘/create-account‘, createAccountLimiter, function(req, res) {
 //…
});

A req.rateLimit property is added to all requests with the limit, current, and remainingnumber of requests for usage in your application code.

Configuration

  • windowMs: milliseconds – how long to keep records of requests in memory. Defaults to 60000 (1 minute).
  • delayAfter: max number of connections during windowMs before starting to delay responses. Defaults to 1. Set to 0 to disable delaying.
  • delayMs: milliseconds – how long to delay the response, multiplied by (number of recent hits – delayAfter). Defaults to 1000 (1 second). Set to 0 to disable delaying.
  • max: max number of connections during windowMs milliseconds before sending a 429 response. Defaults to 5. Set to 0 to disable.
  • message: Error message returned when max is exceeded. Defaults to 'Too many requests, please try again later.'
  • statusCode: HTTP status code returned when max is exceeded. Defaults to 429.
  • headers: Enable headers for request limit (X-RateLimit-Limit) and current usage (X-RateLimit-Remaining) on all responses and time to wait before retrying (Retry-After) when max is exceeded.
  • skipFailedRequests: when true failed requests (response status >= 400) won’t be counted. Defaults to false.
  • keyGenerator: Function used to generate keys. By default user IP address (req.ip) is used. Defaults:
function (req /*, res*/) {
    return req.ip;
}
  • skip: Function used to skip requests. Returning true from the function will skip limiting for that request. Defaults:
function (/*req, res*/) {
    return false;
}
  • handler: The function to execute once the max limit is exceeded. It receives the request and the response objects. The “next” param is available if you need to pass to the next middleware. Defaults:
function (req, res, /*next*/) {
  if (options.headers) {
    res.setHeader(‘Retry-After‘, Math.ceil(options.windowMs / 1000));
  }
  res.format({
    html: function(){
      res.status(options.statusCode).end(options.message);
    },
    json: function(){
      res.status(options.statusCode).json({ message: options.message });
    }
  });
}
  • onLimitReached: Function to listen each time the limit is reached. You can use it to debug/log. Defaults:
function (req, res, options) {
  /* empty */
}
  • store: The storage to use when persisting rate limit attempts. By default, the MemoryStore is used. It must implement the following in order to function:
function SomeStore() {
    /**
      * Increments the value in the underlying store for the given key.
      * @method function
      * @param {string} key – The key to use as the unique identifier passed
      *                     down from RateLimit.
      * @param {Store~incrCallback} cb – The callback issued when the underlying
      *                                store is finished.
      */
    this.incr = function(key, cb) {
      // …
    };
 
    /**
      * Decrements the value in the underlying store for the given key. Used only when skipFailedRequests is true
      * @method function
      * @param {string} key – The key to use as the unique identifier passed
      *                     down from RateLimit.
      */
    this.decrement = function(key) {
      // …
    };
 
    /**
     * This callback is called by the underlying store when an answer to the
     * increment is available.
     * @callback Store~incrCallback
     * @param {?object} err – The error from the underlying store, or null if no
     *                      error occurred.
     * @param {number} value – The current value of the counter
     */
 
    /**
     * Resets a value with the given key.
     * @method function
     * @param  {[type]} key – The key to reset
     */
    this.resetKey = function(key) {
      // …
    };
};

Avaliable data stores are:

  • MemoryStore: (default)Simple in-memory option. Does not share state when app has multiple processes or servers.

The delayAfter and delayMs options were written for human-facing pages such as login and password reset forms. For public APIs, setting these to 0 (disabled) and relying on only windowMs and max for rate-limiting usually makes the most sense.

Instance API

  • resetKey(key): Resets the rate limiting for a given key. (Allow users to complete a captcha or whatever to reset their rate limit, then call this method.)

v2 changes

v2 uses a less precise but less resource intensive method of tracking hits from a given IP. v2 also adds the limiter.resetKey() API and removes the global: true option.

License