Module: Resque::Plugins::Lock

Defined in:
lib/resque/plugins/lock.rb

Overview

If you want only one instance of your job queued at a time, extend it with this module.

For example:

require ‘resque/plugins/lock’

class UpdateNetworkGraph

extend Resque::Plugins::Lock

def self.perform(repo_id)
  heavy_lifting
end

end

No other UpdateNetworkGraph jobs will be placed on the queue, the QueueLock class will check Redis to see if any others are queued with the same arguments before queueing. If another is queued the enqueue will be aborted.

If you want to define the key yourself you can override the ‘lock` class method in your subclass, e.g.

class UpdateNetworkGraph

extend Resque::Plugins::Lock

# Run only one at a time, regardless of repo_id.
def self.lock(repo_id)
  "network-graph"
end

def self.perform(repo_id)
  heavy_lifting
end

end

The above modification will ensure only one job of class UpdateNetworkGraph is running at a time, regardless of the repo_id. Normally a job is locked using a combination of its class name and arguments.

Instance Method Summary collapse

Instance Method Details

#around_perform_lock(*args) ⇒ Object 



55
56
57
58
59
60
61
62
63
 
# File 'lib/resque/plugins/lock.rb', line 55

def around_perform_lock(*args)
  begin
    yield
  ensure
    # Always clear the lock when we're done, even if there is an
    # error.
    Resque.redis.del(lock(*args))
  end
end

#before_enqueue_lock(*args) ⇒ Object 



51
52
53
 
# File 'lib/resque/plugins/lock.rb', line 51

def before_enqueue_lock(*args)
  Resque.redis.setnx(lock(*args), true)
end

#lock(*args) ⇒ Object

Override in your job to control the lock key. It is passed the same arguments as ‘perform`, that is, your job’s payload.



47
48
49
 
# File 'lib/resque/plugins/lock.rb', line 47

def lock(*args)
  "lock:#{name}-#{args.to_s}"
end