stakhanov
AMQP worker library
Last updated 6 months ago by cptechadmin .
MIT · Repository · Bugs · Original npm · Tarball · package.json
$ cnpm install stakhanov 
SYNC missed versions from official npm registry.

Stakhanov

CircleCI codecov package License

Stakhanov allows you to easily create workers that handle tasks from an AMQP queue. It handles common concerns related to implementing a worker with AMQP:

  • Number of tasks handled at the same time
  • Timeout after which the task is considered as failed
  • Handle disconnections from the AMQP server
  • Validating the schema of the message specifying the task

WARNING

The queueName configuration must be unique for each worker, otherwise messages won't necessarily be routed to the good consumer.

When listening, the lib will create a queue of the form queueName.routingKey for each routing key/handler. That is why the queueName config must really be unique, typically of the form application.workername. This will generate a unique queue name per routing key, of the form application.workername.routingkey.

Example:

    function* handle(msg) { ... };
    function* validate(msg) { ... };

    const worker = workerlib.createWorkers([{
      handle: handle,
      validate: validate,
      routingKey: 'application.something_happened'
    }], {
      workerName: 'my worker',
      amqpUrl: 'amqp://guest:guest@localhost:5672',
      exchangeName: 'bus',
      queueName: 'example.simple_worker'
    }, {
      channelPrefetch: 50,
      taskTimeout: 30000,
      processExitTimeout: 3000,
      channelCloseTimeout: 500,
    });

Spec:

createWorkers(handlers, config, options = {})

Create and return a worker instance.

handlers

An array of handlers to handle incoming messages, each handler is an object with the following keys:

  • routingKey: name of the routing key to bind to
  • handle: function to handle incoming messages (see example below)
  • validate: function validating incoming messages bodies (see example below). If the function does not throw, the message is considered valid

config

Configuration of the worker:

  • amqpUrl: url of the AMQP broker
  • exchangeName: name of the exchange to listen on
  • exchangeType: type of the exchange (default to topic)
  • queueName: name of the queue to bind to / create
  • queueoptions: queue options (default to {})
  • workerName: name of the worker, used for logging purposes

options

Optional parameters for the worker:

  • heartbeat: if provided, will override default heartbeat value (in seconds, default 10)
  • taskTimeout: task timeout (maximum time in milliseconds allowed to be spent on message handling, default 30000)
  • processExitTimeout: process exit timeout (maximum time in milliseconds the worker will wait for connections to close before forcing exit, default 3000)
  • channelCloseTimeout: timeout between channel cancelation and close (maximum time in milliseconds the worker will let unfinished messages processing before nacking them, default 500)
  • closeOnSignals: listen to SIGINT/SIGTERM and call worker.close for a graceful shutdown (default false)
  • channelPrefetch: channel prefetch value (default 100)
  • logger: logger object implementing common logging functions (debug, info, warn, error)

Basic use

To listen on channel:

    yield worker.listen();

To shutdown worker:

    yield worker.close();

Remark: for testing purpose, as the close function will eventually execute a process.exit, you can add the following parameter to the close function:

    yield worker.close(false);

Events

To organize your own tests, the worker instance allows you to wait for specific event with the wait method.

Example:

yield worker.wait('task.completed');

Spec:

wait(event, timeout)

The wait method returns a Promise which will be completed when the expected event occurs for the first time, or rejected after a timeout.

event: name of the event, required. Must be one of the available events:

  • task.completed: emitted when a task has been completed.
  • task.retried: emitted when a task is going to be retried because the handler has failed.
  • task.failed: emitted when a task has failed because the handler has failed on a task which was already being retried.
  • task.closed: emitted when the worker has been closed.

timeout: timeout in milliseconds after which the promise will be rejected. Defaults to 1000 ms.

Dev Requirements

Install Node 6.

For nvm users, just move to the project directory and run :

nvm i

If you already have installed Node 6 before, just type:

nvm use

Current Tags

  • 0.7.3                                ...           latest (6 months ago)

15 Versions

  • 0.7.3                                ...           6 months ago
  • 0.7.2                                ...           a year ago
  • 0.7.1                                ...           a year ago
  • 0.7.0                                ...           a year ago
  • 0.6.0                                ...           a year ago
  • 0.5.0                                ...           a year ago
  • 0.4.0                                ...           a year ago
  • 0.3.2                                ...           a year ago
  • 0.3.1                                ...           2 years ago
  • 0.3.0                                ...           2 years ago
  • 0.2.2                                ...           2 years ago
  • 0.2.1                                ...           2 years ago
  • 0.2.0                                ...           2 years ago
  • 0.1.1                                ...           2 years ago
  • 0.1.0                                ...           2 years ago
Maintainers (1)
Downloads
Today 0
This Week 0
This Month 15
Last Day 0
Last Week 15
Last Month 2
Dependencies (4)
Dev Dependencies (17)
Dependents (1)

Copyright 2014 - 2016 © taobao.org |