Basic rate-limiting middleware for Express. Use to limit repeated requests to public APIs and/or endpoints such as password reset.
Plays nice with express-slow-down.
Note: this module does not share state with other processes/servers by default. If you need a more robust solution, I recommend using an external store:
- Memory Store (default, built-in) - stores hits in-memory in the Node.js process. Does not share state with other servers or processes.
- Redis Store
- Memcached Store
- Mongo Store
This module was designed to only handle the basics and didn't even support external stores initially. These other options all are excellent pieces of software and may be more appropriate for some situations:
$ npm install --save express-rate-limit
For an API-only server where the rate-limiter should be applied to all requests:
const rateLimit = require("express-rate-limit");
// Enable if you're behind a reverse proxy (Heroku, Bluemix, AWS ELB, Nginx, etc)
// see https://expressjs.com/en/guide/behind-proxies.html
// app.set('trust proxy', 1);
const limiter = rateLimit({
windowMs: 15 * 60 * 1000, // 15 minutes
max: 100 // limit each IP to 100 requests per windowMs
});
// 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:
const rateLimit = require("express-rate-limit");
// Enable if you're behind a reverse proxy (Heroku, Bluemix, AWS ELB, Nginx, etc)
// see https://expressjs.com/en/guide/behind-proxies.html
// app.set('trust proxy', 1);
const apiLimiter = rateLimit({
windowMs: 15 * 60 * 1000, // 15 minutes
max: 100
});
// only apply to requests that begin with /api/
app.use("/api/", apiLimiter);
Create multiple instances to apply different rules to different routes:
const rateLimit = require("express-rate-limit");
// Enable if you're behind a reverse proxy (Heroku, Bluemix, AWS ELB, Nginx, etc)
// see https://expressjs.com/en/guide/behind-proxies.html
// app.set('trust proxy', 1);
const apiLimiter = rateLimit({
windowMs: 15 * 60 * 1000, // 15 minutes
max: 100
});
app.use("/api/", apiLimiter);
const createAccountLimiter = rateLimit({
windowMs: 60 * 60 * 1000, // 1 hour window
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) {
//...
});
Note: most stores will require additional configuration, such as custom prefixes, when using multiple instances. The default built-in memory store is an exception to this rule.
A req.rateLimit
property is added to all requests with the limit
, current
, and remaining
number of requests and, if the store provides it, a resetTime
Date object. These may be used in your application code to take additional actions or inform the user of their status.
Max number of connections during windowMs
milliseconds before sending a 429 response.
May be a number, or a function that returns a number or a promise. If max
is a function, it will be called with req
and res
params.
Defaults to 5
. Set to 0
to disable.
Timeframe for which requests are checked/remembered. Also used in the Retry-After header when the limit is reached.
Note: with non-default stores, you may need to configure this value twice, once here and once on the store. In some cases the units also differ (e.g. seconds vs miliseconds)
Defaults to 60000
(1 minute).
Error message sent to user when max
is exceeded.
May be a String, JSON object, or any other value that Express's res.send supports.
Defaults to 'Too many requests, please try again later.'
HTTP status code returned when max
is exceeded.
Defaults to 429
.
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.
Defaults to true
. Behavior may change in the next major release.
Enable headers conforming to the ratelimit standardization proposal: RateLimit-Limit
, RateLimit-Remaining
, and, if the store supports it, RateLimit-Reset
. May be used in conjunction with, or instead of the headers
option.
Defaults to false
. Behavior and name will likely change in future releases.
Function used to generate keys.
Defaults to req.ip:
function (req /*, res*/) {
return req.ip;
}
The function to handle requests 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.
Thereq.rateLimit
object has limit
, current
, and remaining
number of requests and, if the store provides it, a resetTime
Date object.
Defaults to:
function (req, res, /*next*/) {
res.status(options.statusCode).send(options.message);
}
Function that is called the first time a user hits the rate limit within a given window.
Thereq.rateLimit
object has limit
, current
, and remaining
number of requests and, if the store provides it, a resetTime
Date object.
Default is an empty function:
function (req, res, options) {
/* empty */
}
When set to true
, failed requests won't be counted. Request considered failed when:
- response status >= 400
- requests that were cancelled before last chunk of data was sent (response
close
event triggered) - response
error
event was triggrered by response
(Technically they are counted and then un-counted, so a large number of slow requests all at once could still trigger a rate-limit. This may be fixed in a future release.)
Defaults to false
.
When set to true
successful requests (response status < 400) won't be counted.
(Technically they are counted and then un-counted, so a large number of slow requests all at once could still trigger a rate-limit. This may be fixed in a future release.)
Defaults to false
.
Function used to skip (whitelist) requests. Returning true
from the function will skip limiting for that request.
Defaults to always false
(count all requests):
function (/*req, res*/) {
return false;
}
The storage to use when persisting rate limit attempts.
By default, the MemoryStore is used.
Available data stores are:
- MemoryStore: (default) Simple in-memory option. Does not share state when app has multiple processes or servers.
- rate-limit-redis: A Redis-backed store, more suitable for large or demanding deployments.
- rate-limit-memcached: A Memcached-backed store.
- rate-limit-mongo: A MongoDB-backed store.
You may also create your own store. 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 {Function} cb - The callback issued when the underlying
* store is finished.
*
* The callback should be called with three values:
* - error (usually null)
* - hitCount for this IP
* - resetTime - JS Date object (optional, but necessary for X-RateLimit-Reset header)
*/
this.incr = function(key, cb) {
// increment storage
cb(null, hits, resetTime);
};
/**
* 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) {
// decrement storage
};
/**
* Resets a value with the given key.
* @method function
* @param {[type]} key - The key to reset
*/
this.resetKey = function(key) {
// remove key from storage or reset it to 0
};
}
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.)
- Removed index.d.ts. (See #138)
- Express Rate Limit no longer modifies the passed-in options object, it instead makes a clone of it.
- Removed
delayAfter
anddelayMs
options; they were moved to a new module: express-slow-down. - Simplified the default
handler
function so that it no longer changes the response format. Now uses res.send. onLimitReached
now only triggers once for a given ip and window. onlyhandle
is called for every blocked request.
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.
MIT © Nathan Friedly