Ansible role for Celery

This role will install celery, a distributed task queue. You can also install Flower, a web based monitoring dashboard for celery clusters.

In order to daemonize Celery and/or Flower, supervisor is used.

Tested On

• Ubuntu 14.04

Defaults

The defaults/main.yml should be pretty clear on the usage and values. The version used by defaults are:

• celery_version: 3.1.18
• celery_flower_version: 0.8.0
• celery_user: nobody
• celery_group: nogroup

Usage

The tests in the respository should be pretty straight forward, but here are some examples:

Celery

This following will generate a supervisor configuration file in /etc/supervisor/conf.d named worker1. The attributes of the key worker1 represent configuration params to supervisor, so you are free to specify the ones you need

- name: Install and run Celery
  hosts: localhost
  sudo: True

  vars:
    - supervisor_services: 
        worker1: 
          directory:    /path/to/celery/task
          process_name: celery_worker1
          num_procs:    1
          user:         "{{celery_user}}"
          command: |
            "celery -A tasks worker --loglevel=debug -f /tmp/worker1.log"
  
  roles:
    - { role: ansible-role-celery }

Flower

Running Flower is very similar:

- name: Install and run Celery
  hosts: localhost
  sudo: True

  vars:
    - celery_flower_service: 
        flower: 
          directory:    /path/to/celery/task
          process_name: celery_flower
          num_procs:    1
          user:         "{{celery_user}}"
          command: |
            "celery -A tasks --loglevel=debug -f /tmp/flower.log --broker=redis://localhost"
  
  roles:
    - { role: ansible-role-celery }

TODO

• Extract supervisor to it's own role
• Extend to other distros