Sign Up for Free

RunKit +

Try any Node.js package right in your browser

This is a playground to test code. It runs a full Node.js environment and already has all of npm’s 400,000 packages pre-installed, including rabbitmq-client with all npm packages installed. Try it out:

var rabbitmqClient = require("rabbitmq-client")

This service is provided by RunKit and is not affiliated with npm, Inc or the package authors.

rabbitmq-client v2.0.0

RabbitMQ (0-9-1) client library

Look at flow-typed/rabbitmq-client_*.js for the full API.


  • does not support the "nowait" param

Notable Differences from amqplib (squaremo/amqp.node):

  • Dooes not support less than node-v10.x
  • No dependencies
  • Failed connections automatically reconnect
  • Built-in clustering/high-availability support
  • Uses native Promises inside & out
  • Uses named args instead of positional arguments
  • ch.basicPublish() returns a Promise when publisher confirms are enabled on the channel
  • "x-arguments" like "x-message-ttl" don't have camelCase aliases

Getting started

const RabbitMQConnection = require('rabbitmq-client')
const connection = new RabbitMQConnection('amqp://guest:guest@localhost:5672')
connection.on('error', (err) => {
  // connection refused, etc

connection.acquire().then(ch => {
  // This does not return a promise by default
  ch.basicPublish('my-queue', {title: 'just some object'})

  // It's your responsibility to close any acquired channels
  return ch.close()
}).then(() => {
  // Graceful shutdown once all channels are closed
  return connection.close()
}).catch(err => {

This library includes some helper functions for creating publishers/consumers that combines a few of the lower level amqp methods and can recover after connection loss.

// create a dedicated channel and create some queues
const pro = connection.createPublisher(['my-queue', 'my-other-queue'])

// publish to the new queues
// publisher-confirms are off by default so this returns undefined
pro.publish('my-queue', {title: 'Hello, AMQP!'})
pro.publish('my-other-queue', 'just some text')

// close the channel
pro.close().catch(err => { console.error(err.stack) })

// create a channel, assert a queue, set the QoS, register a consumer handler
const consumer = connection.createConsumer({
  queue: 'my-queue',
  passive: true, // don't create the channel
  prefetchCount: 1, // consume one at a time
}, async (msg) => {
  // message handler
  // do something async

// keep listening for messages until consumer.close()

The basic methods look like this (async-await syntax):

const ch = await conection.acquire()
await ch.queueDeclare({
  queue: 'my-queue',
  exclusive: true,
// enable publisher confirms manually
await ch.confirmSelect()
await ch.basicPublish('my-queue', {title: 'just some object'})
await ch.close()

If you're publishing too quickly for the socket or for the rabbitmq server to keep up then use connection.unblocked and connection.on('drain', cb) to check for backpressure. This will help you avoid filling up the TCP socket's send-buffer and crashing with an OOM error. Backpressure example:

async function publishOneMillionTimes(connection, queue, data, max=1000000) {
  let ch = await connection.acquire()
  await ch.queueDeclare(queue)
  await new Promise((resolve) => {
    let index = 0
    function write() {
      console.log('start publishing')
      while (connection.unblocked && index < max) {
        index += 1
        ch.basicPublish(queue, data)
      if (index < max) {
        console.log('pause until drained')
        connection.once('drain', write)
      } else {
RunKit is a free, in-browser JavaScript dev environment for prototyping Node.js code, with every npm package installed. Sign up to share your code.
Sign Up for Free