Celery is an asynchronous task/job queue based on distributed message passing. node-celery allows to queue tasks from Node.js. If you are new to Celery check out http://celeryproject.org/
Simple example, included as examples/hello-world.js:
var celery = require('node-celery'),
client = celery.createClient({
CELERY_BROKER_URL: 'amqp://guest:guest@localhost:5672//',
CELERY_RESULT_BACKEND: 'amqp://'
});
client.on('error', function(err) {
console.log(err);
});
client.on('connect', function() {
client.call('tasks.echo', ['Hello World!'], function(result) {
console.log(result);
client.end();
});
});
Note: When using AMQP as result backend with celery prior to version 3.1.7 the result queue needs to be non durable or it will fail with a: Queue.declare: (406) PRECONDITION_FAILED.
var celery = require('node-celery'),
client = celery.createClient({
CELERY_TASK_RESULT_DURABLE: false
});
For RabbitMQ backends, the entire broker options can be passed as an object that is handed off to AMQP. This allows you to specify parameters such as SSL keyfiles, vhost, and connection timeout among others.
var celery = require('node-celery'),
client = celery.createClient({
CELERY_BROKER_OPTIONS: {
host: 'localhost',
port: '5672',
login: 'guest',
password: 'guest',
authMechanism: 'AMQPLAIN',
vhost: '/',
ssl: {
enabled: true,
keyFile: '/path/to/keyFile.pem',
certFile: '/path/to/certFile.pem',
caFile: '/path/to/caFile.pem'
}
},
CELERY_RESULT_BACKEND: 'amqp'
});
The ETA (estimated time of arrival) lets you set a specific date and time that is the earliest time at which your task will be executed:
var celery = require('node-celery'),
client = celery.createClient({
CELERY_BROKER_URL: 'amqp://guest:guest@localhost:5672//',
});
client.on('connect', function() {
client.call('send-email', {
to: '[email protected]',
title: 'sample email'
}, {
eta: new Date(Date.now() 60 * 60 * 1000) // an hour later
});
});
The expires argument defines an optional expiry time, a specific date and time using Date:
var celery = require('node-celery'),
client = celery.createClient({
CELERY_BROKER_URL: 'amqp://guest:guest@localhost:5672//',
});
client.on('connect', function() {
client.call('tasks.sleep', [2 * 60 * 60], null, {
expires: new Date(Date.now() 60 * 60 * 1000) // expires in an hour
});
});
The backend is used to store task results. Currently AMQP (RabbitMQ) and Redis backends are supported.
var celery = require('node-celery'),
client = celery.createClient({
CELERY_BROKER_URL: 'amqp://guest:guest@localhost:5672//',
CELERY_RESULT_BACKEND: 'redis://localhost/0'
});
client.on('connect', function() {
var result = client.call('tasks.add', [1, 2]);
setTimeout(function() {
result.get(function(data) {
console.log(data); // data will be null if the task is not finished
});
}, 2000);
});
AMQP backend allows to subscribe to the task result and get it immediately, without polling:
var celery = require('node-celery'),
client = celery.createClient({
CELERY_BROKER_URL: 'amqp://guest:guest@localhost:5672//',
CELERY_RESULT_BACKEND: 'amqp'
});
client.on('connect', function() {
var result = client.call('tasks.add', [1, 2]);
result.on('ready', function(data) {
console.log(data);
});
});
The simplest way to route tasks to different queues is using CELERY_ROUTES configuration option:
var celery = require('node-celery'),
client = celery.createClient({
CELERY_BROKER_URL: 'amqp://guest:guest@localhost:5672//',
CELERY_ROUTES: {
'tasks.send_mail': {
queue: 'mail'
}
}
}),
send_mail = client.createTask('tasks.send_mail'),
calculate_rating = client.createTask('tasks.calculate_rating');
client.on('error', function(err) {
console.log(err);
});
client.on('connect', function() {
send_mail.call([], {
to: '[email protected]',
title: 'hi'
}); // sends a task to the mail queue
calculate_rating.call([], {
item: 1345
}); // sends a task to the default queue
});