@fastify/rate-limit

A low overhead rate limiter for your routes.

Install

npm i @fastify/rate-limit

Compatibility

Plugin version	Fastify version
^10.x	^5.x
^9.x	^4.x
^8.x	^4.x
^7.x	^4.x
^6.x	^2.x and ^3.x
^5.x	^2.x and ^3.x
^4.x	^2.x and ^3.x
^3.x	^2.x and ^3.x
^2.x	^2.x
^1.x	^1.x

Please note that if a Fastify version is out of support, then so are the corresponding versions of this plugin in the table above. See Fastify's LTS policy for more details.

Usage

Register the plugin and, if required, pass some custom options.
This plugin will add an onRequest hook to check if a client (based on their IP address) has made too many requests in the given timeWindow.

1import Fastify from 'fastify'
2
3const fastify = Fastify()
4await fastify.register(import('@fastify/rate-limit'), {
5  max: 100,
6  timeWindow: '1 minute'
7})
8
9fastify.get('/', (request, reply) => {
10  reply.send({ hello: 'world' })
11})
12
13fastify.listen({ port: 3000 }, err => {
14  if (err) throw err
15  console.log('Server listening at http://localhost:3000')
16})

In case a client reaches the maximum number of allowed requests, an error will be sent to the user with the status code set to 429:

1{
2  statusCode: 429,
3  error: 'Too Many Requests',
4  message: 'Rate limit exceeded, retry in 1 minute'
5}

You can change the response by providing a callback to errorResponseBuilder or setting a custom error handler:

1fastify.setErrorHandler(function (error, request, reply) {
2  if (error.statusCode === 429) {
3    reply.code(429)
4    error.message = 'You hit the rate limit! Slow down please!'
5  }
6  reply.send(error)
7})

The response will have some additional headers:

Header	Description
x-ratelimit-limit	how many requests the client can make
x-ratelimit-remaining	how many requests remain to the client in the timewindow
x-ratelimit-reset	how many seconds must pass before the rate limit resets
retry-after	if the max has been reached, the seconds the client must wait before they can make new requests

Preventing guessing of URLS through 404s

An attacker could search for valid URLs if your 404 error handling is not rate limited. To rate limit your 404 response, you can use a custom handler:

1const fastify = Fastify()
2await fastify.register(rateLimit, { global: true, max: 2, timeWindow: 1000 })
3fastify.setNotFoundHandler({
4  preHandler: fastify.rateLimit()
5}, function (request, reply) {
6  reply.code(404).send({ hello: 'world' })
7})

Note that you can customize the behavior of the preHandler in the same way you would for specific routes:

1const fastify = Fastify()
2await fastify.register(rateLimit, { global: true, max: 2, timeWindow: 1000 })
3fastify.setNotFoundHandler({
4  preHandler: fastify.rateLimit({
5    max: 4,
6    timeWindow: 500
7  })
8}, function (request, reply) {
9  reply.code(404).send({ hello: 'world' })
10})

Options

You can pass the following options during the plugin registration:

1await fastify.register(import('@fastify/rate-limit'), {
2  global : false, // default true
3  max: 3, // default 1000
4  ban: 2, // default -1
5  timeWindow: 5000, // default 1000 * 60
6  hook: 'preHandler', // default 'onRequest'
7  cache: 10000, // default 5000
8  allowList: ['127.0.0.1'], // default []
9  redis: new Redis({ host: '127.0.0.1' }), // default null
10  nameSpace: 'teste-ratelimit-', // default is 'fastify-rate-limit-'
11  continueExceeding: true, // default false
12  skipOnError: true, // default false
13  keyGenerator: function (request) { /* ... */ }, // default (request) => request.ip
14  errorResponseBuilder: function (request, context) { /* ... */},
15  enableDraftSpec: true, // default false. Uses IEFT draft header standard
16  addHeadersOnExceeding: { // default show all the response headers when rate limit is not reached
17    'x-ratelimit-limit': true,
18    'x-ratelimit-remaining': true,
19    'x-ratelimit-reset': true
20  },
21  addHeaders: { // default show all the response headers when rate limit is reached
22    'x-ratelimit-limit': true,
23    'x-ratelimit-remaining': true,
24    'x-ratelimit-reset': true,
25    'retry-after': true
26  }
27})

• global : indicates if the plugin should apply rate limiting to all routes within the encapsulation scope.
• max: maximum number of requests a single client can perform inside a timeWindow. It can be an async function with the signature async (request, key) => {} where request is the Fastify request object and key is the value generated by the keyGenerator. The function must return a number.
• ban: maximum number of 429 responses to return to a client before returning 403 responses. When the ban limit is exceeded, the context argument that is passed to errorResponseBuilder will have its ban property set to true. Note: 0 can also be passed to directly return 403 responses when a client exceeds the max limit.
• timeWindow: the duration of the time window. It can be expressed in milliseconds, as a string (in the ms format), or as an async function with the signature async (request, key) => {} where request is the Fastify request object and key is the value generated by the keyGenerator. The function must return a number.
• cache: this plugin internally uses an LRU cache to handle the clients, you can change the size of the cache with this option
• allowList: array of string of IPs to exclude from rate limiting. It can be a sync or async function with the signature (request, key) => {} where request is the Fastify request object and key is the value generated by the keyGenerator. If the function return a truthy value, the request will be excluded from the rate limit.
• redis: by default, this plugin uses an in-memory store, but if an application runs on multiple servers, an external store will be needed. This plugin requires the use of ioredis.
  Note: the default settings of an ioredis instance are not optimal for rate limiting. We recommend customizing the connectTimeout and maxRetriesPerRequest parameters as shown in the example.
• nameSpace: choose which prefix to use in the redis, default is 'fastify-rate-limit-'
• continueExceeding: Renew user limitation when user sends a request to the server when still limited. This will take priority over exponentialBackoff
• store: a custom store to track requests and rates which allows you to use your own storage mechanism (using an RDBMS, MongoDB, etc.) as well as further customizing the logic used in calculating the rate limits. A simple example is provided below as well as a more detailed example using Knex.js can be found in the example/ folder
• skipOnError: if true it will skip errors generated by the storage (e.g. redis not reachable).
• keyGenerator: a sync or async function to generate a unique identifier for each incoming request. Defaults to (request) => request.ip, the IP is resolved by fastify using request.connection.remoteAddress or request.headers['x-forwarded-for'] if trustProxy option is enabled. Use it if you want to override this behavior
• groupId: a string to group multiple routes together introducing separate per-group rate limit. This will be added on top of the result of keyGenerator.
• errorResponseBuilder: a function to generate a custom response object. Defaults to (request, context) => ({statusCode: 429, error: 'Too Many Requests', message: ``Rate limit exceeded, retry in ${context.after}``})
• addHeadersOnExceeding: define which headers should be added in the response when the limit is not reached. Defaults all the headers will be shown
• addHeaders: define which headers should be added in the response when the limit is reached. Defaults all the headers will be shown
• enableDraftSpec: if true it will change the HTTP rate limit headers following the IEFT draft document. More information at draft-ietf-httpapi-ratelimit-headers.md.
• onExceeding: callback that will be executed before request limit has been reached.
• onExceeded: callback that will be executed after request limit has been reached.
• onBanReach: callback that will be executed when the ban limit has been reached.
• exponentialBackoff: Renew user limitation exponentially when user sends a request to the server when still limited.

keyGenerator example usage:

1await fastify.register(import('@fastify/rate-limit'), {
2  /* ... */
3  keyGenerator: function (request) {
4    return request.headers['x-real-ip'] // nginx
5    || request.headers['x-client-ip'] // apache
6    || request.headers['x-forwarded-for'] // use this only if you trust the header
7    || request.session.username // you can limit based on any session value
8    || request.ip // fallback to default
9  }
10})

Variable max example usage:

1// In the same timeWindow, the max value can change based on request and/or key like this
2fastify.register(rateLimit, {
3  /* ... */
4  keyGenerator (request) { return request.headers['service-key'] },
5  max: async (request, key) => { return key === 'pro' ? 3 : 2 },
6  timeWindow: 1000
7})

errorResponseBuilder example usage:

1await fastify.register(import('@fastify/rate-limit'), {
2  /* ... */
3  errorResponseBuilder: function (request, context) {
4    return {
5      statusCode: 429,
6      error: 'Too Many Requests',
7      message: `I only allow ${context.max} requests per ${context.after} to this Website. Try again soon.`,
8      date: Date.now(),
9      expiresIn: context.ttl // milliseconds
10    }
11  }
12})

Dynamic allowList example usage:

1await fastify.register(import('@fastify/rate-limit'), {
2  /* ... */
3  allowList: function (request, key) {
4    return request.headers['x-app-client-id'] === 'internal-usage'
5  }
6})

Custom hook example usage (after authentication):

1await fastify.register(import('@fastify/rate-limit'), {
2  hook: 'preHandler',
3  keyGenerator: function (request) {
4    return request.userId || request.ip
5  }
6})
7
8fastify.decorateRequest('userId', '')
9fastify.addHook('preHandler', async function (request) {
10  const { userId } = request.query
11  if (userId) {
12    request.userId = userId
13  }
14})

Custom store example usage:

NOTE: The timeWindow will always be passed as the numeric value in milliseconds into the store's constructor.

1function CustomStore (options) {
2  this.options = options
3  this.current = 0
4}
5
6CustomStore.prototype.incr = function (key, cb) {
7  const timeWindow = this.options.timeWindow
8  this.current++
9  cb(null, { current: this.current, ttl: timeWindow - (this.current * 1000) })
10}
11
12CustomStore.prototype.child = function (routeOptions) {
13  // We create a merged copy of the current parent parameters with the specific
14  // route parameters and pass them into the child store.
15  const childParams = Object.assign(this.options, routeOptions)
16  const store = new CustomStore(childParams)
17  // Here is where you may want to do some custom calls on the store with the information
18  // in routeOptions first...
19  // store.setSubKey(routeOptions.method + routeOptions.url)
20  return store
21}
22
23await fastify.register(import('@fastify/rate-limit'), {
24  /* ... */
25  store: CustomStore
26})

The routeOptions object passed to the child method of the store will contain the same options that are detailed above for plugin registration with any specific overrides provided on the route. In addition, the following parameter is provided:

• routeInfo: The configuration of the route including method, url, path, and the full route config

Custom onExceeding example usage:

1await fastify.register(import('@fastify/rate-limit'), {
2  /* */
3  onExceeding: function (req, key) {
4    console.log('callback on exceeding ... executed before response to client')
5  }
6})

Custom onExceeded example usage:

1await fastify.register(import('@fastify/rate-limit'), {
2  /* */
3  onExceeded: function (req, key) {
4    console.log('callback on exceeded ... executed before response to client')
5  }
6})

Custom onBanReach example usage:

1await fastify.register(import('@fastify/rate-limit'), {
2  /* */
3  ban: 10,
4  onBanReach: function (req, key) {
5    console.log('callback on exceeded ban limit')
6  }
7})

Options on the endpoint itself

Rate limiting can also be configured at the route level, applying the configuration independently.

For example the allowList if configured:

• on plugin registration will affect all endpoints within the encapsulation scope
• on route declaration will affect only the targeted endpoint

The global allowlist is configured when registering it with fastify.register(...).

The endpoint allowlist is set on the endpoint directly with the { config : { rateLimit : { allowList : [] } } } object.

ACL checking is performed based on the value of the key from the keyGenerator.

In this example, we are checking the IP address, but it could be an allowlist of specific user identifiers (like JWT or tokens):

1import Fastify from 'fastify'
2
3const fastify = Fastify()
4await fastify.register(import('@fastify/rate-limit'),
5  {
6    global : false, // don't apply these settings to all the routes of the context
7    max: 3000, // default global max rate limit
8    allowList: ['192.168.0.10'], // global allowlist access.
9    redis: redis, // custom connection to redis
10  })
11
12// add a limited route with this configuration plus the global one
13fastify.get('/', {
14  config: {
15    rateLimit: {
16      max: 3,
17      timeWindow: '1 minute'
18    }
19  }
20}, (request, reply) => {
21  reply.send({ hello: 'from ... root' })
22})
23
24// add a limited route with this configuration plus the global one
25fastify.get('/private', {
26  config: {
27    rateLimit: {
28      max: 3,
29      timeWindow: '1 minute'
30    }
31  }
32}, (request, reply) => {
33  reply.send({ hello: 'from ... private' })
34})
35
36// this route doesn't have any rate limit
37fastify.get('/public', (request, reply) => {
38  reply.send({ hello: 'from ... public' })
39})
40
41// add a limited route with this configuration plus the global one
42fastify.get('/public/sub-rated-1', {
43  config: {
44    rateLimit: {
45      timeWindow: '1 minute',
46      allowList: ['127.0.0.1'],
47      onExceeding: function (request, key) {
48        console.log('callback on exceeding ... executed before response to client')
49      },
50      onExceeded: function (request, key) {
51        console.log('callback on exceeded ... to black ip in security group for example, request is give as argument')
52      }
53    }
54  }
55}, (request, reply) => {
56  reply.send({ hello: 'from sub-rated-1 ... using default max value ... ' })
57})
58
59// group routes and add a rate limit
60fastify.get('/otp/send', {
61  config: {
62    rateLimit: {
63      max: 3,
64      timeWindow: '1 minute',
65      groupId:"OTP"
66    }
67  }
68}, (request, reply) => {
69  reply.send({ hello: 'from ... grouped rate limit' })
70})
71
72fastify.get('/otp/resend', {
73  config: {
74    rateLimit: {
75      max: 3,
76      timeWindow: '1 minute',
77      groupId:"OTP"
78    }
79  }
80}, (request, reply) => {
81  reply.send({ hello: 'from ... grouped rate limit' })
82})

In the route creation you can override the same settings of the plugin registration plus the following additional options:

• onExceeding : callback that will be executed each time a request is made to a route that is rate-limited
• onExceeded : callback that will be executed when a user reaches the maximum number of tries. Can be useful to blacklist clients

You may also want to set a global rate limiter and then disable it on some routes:

1import Fastify from 'fastify'
2
3const fastify = Fastify()
4await fastify.register(import('@fastify/rate-limit'), {
5  max: 100,
6  timeWindow: '1 minute'
7})
8
9// add a limited route with global config
10fastify.get('/', (request, reply) => {
11  reply.send({ hello: 'from ... rate limited root' })
12})
13
14// this route doesn't have any rate limit
15fastify.get('/public', {
16  config: {
17    rateLimit: false
18  }
19}, (request, reply) => {
20  reply.send({ hello: 'from ... public' })
21})
22
23// add a limited route with global config and different max
24fastify.get('/private', {
25  config: {
26    rateLimit: {
27      max: 9
28    }
29  }
30}, (request, reply) => {
31  reply.send({ hello: 'from ... private and more limited' })
32})

Examples of Custom Store

These examples show an overview of the store feature and you should take inspiration from it and tweak as you need:

• Knex-SQLite
• Knex-MySQL
• Sequelize-PostgreSQL

IETF Draft Spec Headers

The response will have the following headers if enableDraftSpec is true:

Header	Description
ratelimit-limit	how many requests the client can make
ratelimit-remaining	how many requests remain to the client in the timewindow
ratelimit-reset	how many seconds must pass before the rate limit resets
retry-after	contains the same value in time as ratelimit-reset

Contribute

To run tests locally, you need a Redis instance that you can launch with this command:

npm run redis

License

Licensed under MIT.