Sign up for a SimpleWorker account, it's free to try!
gem install simple_worker
You really just need your access keys.
SimpleWorker.configure do |config|
config.access_key = ACCESS_KEY
config.secret_key = SECRET_KEY
end
If you're rolling on Rails, use the following:
rails generate simple_worker worker_class_in_underscore_format
This will set you up with a worker in app/workers
However, this is not autorequired, so a line similar to this will be required in your initializer:
Dir.glob("#{Rails.root}/app/workers/*.rb").each{|w| require w}
This one-liner requires all .rb files in app/workers
Here's an example worker that sends an email:
require 'simple_worker'
class EmailWorker < SimpleWorker::Base
attr_accessor :to, :subject, :body
# This is the method that will be run
def run
send_email(:to=>to, :subject=>subject, :body=>body)
end
def send_email
# Put sending code here
end
end
Let's say someone does something in your app and you want to send an email about it.
worker = EmailWorker.new
worker.to = current_user.email
worker.subject = "Here is your mail!"
worker.body = "This is the body"
worker.run_local
Once you've got it working locally, the next step is to run it on the SimpleWorker cloud.
Let's say someone does something in your app and you want to send an email about it.
worker = EmailWorker.new
worker.to = current_user.email
worker.subject = "Here is your mail!"
worker.body = "This is the body"
worker.queue
This will send it off to the SimpleWorker cloud.
Simply define the priority in your queue command.
worker.queue(:priority=>1)
Default priority is 0 and we currently support priority 0, 1, 2. See pricing page for more information on priorites.
There are two scenarios here, one is the scenario where you want something to happen due to a user action in your application. This is almost the same as queuing your worker.
worker = EmailWorker.new
worker.to = current_user.email
worker.subject = "Here is your mail!"
worker.body = "This is the body"
worker.schedule(:start_at=>1.hours.since)
By default, if you call schedule more than once for the same worker class, the new schedule will replace the old one. If you'd
like multiple schedules for the same class, provide a :name parameter:
worker.schedule(:name=>"EmailWorkerDaily", :start_at=>......)
You can also set priority for scheduled jobs:
worker.schedule(:priority=>1, ....)
If you still have access to the worker object, just call:
worker.status
If you only have the job ID, call:
SimpleWorker.service.status(job_id)
This will return a hash like:
{"task_id"=>"ece460ce-12d8-11e0-8e15-12313b0440c6",
"status"=>"running",
"msg"=>nil,
"start_time"=>"2010-12-28T23:19:36+00:00",
"end_time"=>nil,
"duration"=>nil,
"progress"=>{"percent"=>25}}
There is also a convenience method worker.wait_until_complete that will wait until the status returned is completed or error.
In your worker, just call the log method with the string you want logged:
log "Starting to do something..."
The log will be available for viewing via the SimpleWorker UI or via log in the API:
SimpleWorker.service.log(job_id)
or if you still have a handle to your worker object:
worker.get_log
This is just a way to let your users know where the job is at if required.
set_progress(:percent => 25, :message => "We are a quarter of the way there!")
You can actually put anything in this hash and it will be returned with a call to status. We recommend using the format above for consistency and to get some additional features where we look for these values.
The alternative is when you want to user it like Cron. In this case you'll probably want to write a script that will schedule, you don't want to schedule it everytime your app starts or anything so best to keep it external.
Create a file called 'schedule_email_worker.rb' and add this:
require 'simple_worker'
require_relative 'email_worker'
worker = EmailWorker.new
worker.to = current_user.email
worker.subject = "Here is your mail!"
worker.body = "This is the body"
worker.schedule(:start_at=>1.hours.since, :run_every=>3600)
Now run it and your worker will be scheduled to run every hour.
Rails 2.X:
config.gem 'simple_worker'
Rails 3.X:
gem 'simple_worker'
Now you can use your workers like they're part of your app! We recommend putting your worker classes in /app/workers path.
Although you could easily do this in your worker, this makes it a bit more convenient and more importantly it will create the connection for you. If you are using Rails 3, you just need to add one line:
config.database = Rails.configuration.database_configuration[Rails.env]
For non Rails 3, you would add the following to your SimpleWorker config:
config.database = {
:adapter => "mysql2",
:host => "localhost",
:database => "appdb",
:username => "appuser",
:password => "secret"
}
Then before you job is run, SimpleWorker will establish the ActiveRecord connection.
If you need to inclue other classes in for your worker to use (which is very common), then you'll want to use the merge functions. For example:
class AvgWorker < SimpleWorker::Base
attr_accessor :aws_access_key,
:aws_secret_key,
:s3_suffix
merge File.join(File.dirname(__FILE__), "..", "app", "models", "user.rb")
merge File.join(File.dirname(__FILE__), "..", "app", "models", "account")
Or simpler yet, try using relative paths:
merge "../models/user"
merge "../models/account.rb"
Also you could use wildcards and path in merge_folder
merge_folder "./foldername/" #will merge all *.rb files from foldername
merge_folder "../lib/**/" # will merge all *.rb files from lib and all subdirectories
The opposite can be done as well with "unmerge" and can be useful when using Rails to exclude classes that are automatically merged.
Merging other workers is a bit different than merging other code like above because they will be uploaded separately and treated as distinctly separate workers.
merge_worker "./other_worker.rb", "OtherWorker"
You could easily merge mailers you're using in your application.
merge_mailer 'mailer_file' #if your mailer's templates are placed in default path
#or
merge_mailer 'mailer_file', {:path_to_templates=>"templates_path"}#if you're using mailer outside of rails with custom templates path
if you already set auto_merge=true all your mailers already merged.
Configuring a Mailer Connection
If you are using Rails 3,your action_mailer connection would be configured automatically from your config
For non Rails 3 or if you want to use different mailer configs, you should add the following to your SimpleWorker config:
config.mailer = {
:address => "smtp.gmail.com",
:port => 587,
:domain => 'gmail.com',
:user_name => GMAIL_USERNAME
:password => GMAIL_PASSWORD
:authentication => 'plain',
:enable_starttls_auto => true}
Then before you job is run, SimpleWorker will establish the ActionMailer connection.
This allows you to use any gem you'd like with SimpleWorker. This uses the same syntax as bundler gem files.
# merge latest version of the gem
merge_gem "some_gem"
# or specify specific version
merge_gem "some_gem_with_version", "1.2.3"
# or if gem has poor naming scheme
merge_gem 'mongoid_i18n', :require => 'mongoid/i18n'
# of if gem requires other directories outside lib
merge_gem 'prawn', :include_dirs=>['data']
Check here for more info on merge_gem.
If you want to merge items for all of your workers, you can do merges in your SimpleWorker.configure block:
config.merge 'my file' config.merge_gem 'httparty'
These are attributes that can be set as part of your config block then will be set on all your worker objects automatically. This is particularly good for things like database connection info or things that you would need to use across the board.
Eg:
config.global_attributes[:db_user] = "sa"
config.global_attributes[:db_pass] = "pass"
Then in your worker, you would have the attributes defined:
attr_accessor :db_user, :db_pass
Join the discussion group at: https://groups.google.com/forum/?hl=en#!forum/simple_worker