Gem Version Gem Downloads Build Status

Extension for Delayed::Job that allows better tracking of jobs!


  • add to Gemfile: gem 'delayed_job_progress'
  • bundle install
  • rails g delayed_job:progress
  • rake db:migrate

Configuration and Usage

Consider this:

class User < ActiveRecord::Base
  # convenient relationship to grab associated jobs
  has_many :jobs, :as => :record, :class_name => 'Delayed::Job'

Creating a delayed job:

user = User.find(123)

If you're using custom jobs you'll need to do something like this:

class CustomUserJob <
  def enqueue(job)
    job.record            = User.find(user_id)
    job.identifier        = 'unique_identifier'
    job.progress_max      = 100
    job.progress_current  = 0

  def before(job)
    @job  = job
    @user = job.record

  def perform
    @job.update_column(:message, 'working')
    (0..100).each do |i|
      @job.update_column(:progress_current, i)
    @job.update_column(:message, 'complete')


This will create a Delayed::Job record:

=> [#<Delayed::Job>]

That job knows about object that spawned it:

-> Delayed::Job.last.record
=> #<User>

Delayed::Job records now have new attributes:

`progress_max`      - default is `100`. You can change it to whatever during `enqueue`.
`progress_current`  - default is `0`. You can manually increment it while job is running. Will be set to `process_max` when job completes.
`message`           - default is `nil`. Optional informational string.
`error_message`     - error message without backtrace. Also useful to set your own message.
`completed_at`      - when job is done this timestamp is recorded.
`identifier`        - you can assign something during `enqueue` so you can fetch that job later for display.

This extension also introduces worker setting that keeps completed jobs around. This way you can keep list of completed jobs for a while. If you want to remove them, you need to .destroy(:force) them.

Delayed::Worker.destroy_completed_jobs = false

Jobs Controller

GET /jobs               - List all jobs. Can filter based on associated record via `record_type` and `record_id` parameters. `identifier` parameter can be used as well
GET /jobs/<id>          - Status of a job. Will see all the Delayed::Job attributes including things like progress
DELETE /jobs/<id>`      - If job is stuck/failed, we can remove it
POST /jobs/<id>/reload` - Restart failed job