Module: Puppet::Agent::Locker

Included in:
Puppet::Agent
Defined in:
lib/puppet/agent/locker.rb

Overview

This module is responsible for encapsulating the logic for “locking” the puppet agent during a catalog run; in other words, keeping track of enough state to answer the question “is there a puppet agent currently applying a catalog?”

The implementation involves writing a lockfile whose contents are simply the PID of the running agent process. This is considered part of the public Puppet API because it used by external tools such as mcollective.

For more information, please see docs on the website.

http://links.puppet.com/agent_lockfiles

Instance Method Summary collapse

Instance Method Details

#lock ⇒ Object

Yield if we get a lock, else raise Puppet::LockError. Return value of block yielded.



18
19
20
21
22
23
24
25
26
27
28
 
# File 'lib/puppet/agent/locker.rb', line 18

def lock
  if lockfile.lock
    begin
      yield
    ensure
      lockfile.unlock
    end
  else
    fail Puppet::LockError, _('Failed to acquire lock')
  end
end

#lockfile_path ⇒ Object 



40
41
42
 
# File 'lib/puppet/agent/locker.rb', line 40

def lockfile_path
  @lockfile_path ||= Puppet[:agent_catalog_run_lockfile]
end

#running? ⇒ Boolean

Deprecated.

Returns:

  • (Boolean) 


31
32
33
34
35
36
37
38
 
# File 'lib/puppet/agent/locker.rb', line 31

def running?
  Puppet.deprecation_warning _(<<-ENDHEREDOC)
Puppet::Agent::Locker.running? is deprecated as it is inherently unsafe.
The only safe way to know if the lock is locked is to try lock and perform some
action and then handle the LockError that may result.
ENDHEREDOC
  lockfile.locked?
end