Distributed Lock

Python distributed lock (currently only with memcached)

How to configure

In your setup file (in django, settings.py), configure locking:

import distributedlock
distributedlock.DEFAULT_MEMCACHED_CLIENT = memcache.Client(['127.0.0.1:11211'])
distributedlock.DEFAULT_TIMEOUT=60
distributedlock.DEFAULT_BLOCKING=False

If you setting up memcached in Django, you can use it abstraction of memcached.

from django.core.cache import cache
distributedlock.DEFAULT_MEMCACHED_CLIENT = cache

You can configure this settings in each lock, as parameter.

How to use

Using minimal configuration, as decorator:

from distributedlock import distributedlock
@distributedlock()
def hello_world():
    print 'running'

Or as with block:

from distributedlock import distributedlock

#... my code before
with distributedlock('hello'):
    print 'running'
#... my code after

You can use with conventional threading.Lock (only in process locking)

import threading
with distributedlock('hello', lock=threading.Lock())
    print 'running'

Arguments

def distributedlock(key, lock=None, blocking=None)

• key: name of key in memcached. Avoid long names, because memcached supports only 255 characters in key. Using decorator key name will be class name + method name if not specified.
• lock: If you desire use another lock strategy, like threading.Lock() or threading.RLock(). defaults to distributedlock.memcachedlock.MemcachedLock
• blocking: If another process has lock, wait until have a lock or abort immediately, raising LockNotAcquiredError. Defaults to distributedlock.DEFAULT_BLOCKING

Tips

• If you have a dynamic key, use lock with block to compose your key. For example:

def synchronized_method(arg1)
    with distributedlock('sync_process_%d' % arg1.id):
        # do something