1.0.2 • Published 8 years ago

eko-queue v1.0.2

Weekly downloads
59
License
ISC
Repository
-
Last release
8 years ago

EkoQueue

Build Status Code Climate Test Coverage

EkoQueue is a queue manager that allows you to easily dispatch and process jobs with an easy to use interface. EkoQueue wraps the excellent kue.js and also comes with a UI to monitor and manage the queue jobs.

EkoQueue is meant to be fully scalable with no persistance on disk. The only dependency is a Redis server.

Requirements

  • Node.js 6.*
  • Redis Server

Installation

$ npm install eko-queue --save

Configuration

The queue module exposes a constructor function that accepts a job provider and a hash of options to connect to redis.

  • jobProvider (an object that exposes instances of jobs and optionally a job path for job classes)
  • options (object/hash of configuration options for redis) host (host for redis server) port (port for redis server) * prefix (prefix to use for queue's jobs)

Job provider

The job provider is a simple object that allows you to easily compose your queue jobs with dependency injection as well as provide a path for job classes that don't rely on dependency injection.

The job path must use the key jobPath and the keys of the initialized instances must match the name of the job in the dispatcher and processor.

Look inside the examples directory for more examples.

Example

const path = require('path');
const jobPath = path.join(path.dirname(__filename), 'jobs');

const Email = require('./services/email');
const EmailJob = require('./jobs/email-service-job');

/**
 * Wrap email job in function to inject dependencies.
 * These functions themselves cannot have any dependencies.
 * @return {EmailJob}
 */
const emailJob = () => {
  /**
   * Simulating instance of email client that sends email
   * @type {Object}
   */
  const emailClient = {
    send: (address, subject, body, callback) => {
      callback(null, `Email sent to ${address}`);
    },
  };

  /**
   * Email service
   * @type {Email}
   */
  const emailService = new Email(emailClient);

  /**
   * Finally instantiate and return email job with email service injected
   */
  return new EmailJob(emailService);
};

/**
 * Export the jobPath for job classes that require no dependency injection
 * and then export the jobs that have been initialized by their job name
 * @type {Object}
 */
module.exports = {
  jobPath,
  'email-service': emailJob(),
};

Dispatching jobs

The queue dispatcher has a single method to dispatch jobs. The dispatch method takes the job name as the first paramater and following optional parameters act as the parameters for the handle method of the job class.

Methods

  • Dispatcher#dispatch(string job, any ...params)

Examples

const EkoQueue = require('eko-queue')(jobProvider, {});
const dispatcher = EkoQueue.dispatcher();

// Using additional parameters to pass through to the handle method
dispatcher.dispatch('email', 'andrew@ekoapp.com', 'Welcome Back!');

// Or you can just send an object through with your parameters
dispatcher.dispatch('email', {
  email: 'andrew@ekoapp.com',
  subject: 'Welcome Back!',
});

Processing jobs

The queue processor has two (2) methods to process jobs. You can process a single job or process all the jobs. If running the processAll method it will also subscribe to dispatched jobs even after it is started.

Methods

  • Processor#process(string job)
  • Processor#processAll()

Examples

const EkoQueue = require('eko-queue')(jobProvider, {});
const processor = EkoQueue.processor();

// Process individual job
processor.process('email');

// Process all jobs
processor.processAll();

Job classes

The job classes are what the processor uses to process the job. Job classes must all live in the same directory and have a suffix of job in the filename. An example would be a job named email would have a filename of email-job.js.

The job class require a handle and failed method that both accept optional parameters that will be passed from the dispatcher. The job classes can also contain optional concurrency and priority properties.

The handle and failed methods also allow you to yield to promises so you can write your code in a more produral manner. You can yield to generators or promises.

Example

'use strict';

class EmailJob {
  /**
   * Number of jobs to run concurrently
   * @return {number}
   */
  get concurrency() {
    return 1;
  }

  /**
   * Priority of job 10 is lowest, -10 is highest
   * @return {number}
   */
  get priority() {
    return 0;
  }

  /**
   * Handle job
   * @param {string} email
   * @param {string} subject
   */
  * handle(email, subject) {
    console.log(`Sending email to ${email} with the subject ${subject}`);
  	const results = yield someEmailClient.send(email, subject, 'Hello from EkoQueue');
  }

  /**
   * Handle job failure
   * @param {string} email
   * @param {string} subject
   */
  * failed(email, subject) {
    console.error(`Failed sending email to ${email} with the subject ${subject}`);
  }
}

module.exports = EmailJob;

Running tests, linter

  • Run npm test to run the test suite
  • Run npm run lint to run the linter

Running examples

Checkout this repository and run the following in two (2) different terminal sessions. The order does not matter.

$ DEBUG=eko* nodemon examples/process.js
$ DEBUG=eko* nodemon examples/dispatch.js

Running the server

Todo: Export server from constructor. In the mean time you can checkout this repository and run npm run server.