Skip to main content

Exclusive process locking to ensure that your code does not execute concurrently, using POSIX file locking.

Project description

This is a simple Python 3.4+ module for ensuring that your code does not execute concurrently in multiple processes, using POSIX file locking.

The lock can be acquired easily using with syntax or as a decorator.


If you have long-running processes that would cause corruption if it’s executed multiple times concurrently, this package is for you. You might use this in scripts that make backups, perform database migrations, or other long-running processes that need to abort if they are already running.

How it works under the hood

The module uses POSIX file locking and a PID file:

  • A file is selected to hold lock information, typically /var/lock/py_exclusivelock_yournamehere.lock, called the lockfile, based on a name you provide. The name is sanitized before being used in the filename.

  • If the lockfile already exists and it contains the PID of a running process (including the current process), then a CannotAcquireLock exception is thrown.

  • Otherwise the lockfile is created (or overwritten) and this process’s integer process ID is written to the file. POSIX file locking (lockf) and open with O_EXCL is used to prevent race conditions.

  • The lockfile is deleted when the with block or decorated function exits. Or when used with .forever() (see below), at program exit.

How to use it

First install this package:

pip3 install exclusiveprocess

Then in your Python file import the package:

from exclusiveprocess import Lock, CannotAcquireLock

You can use it in a with statement:

    with Lock(name="myprocess"):
        print("This block cannot be executed concurrently!")
except CannotAcquireLock:
    print("Well, that's bad.")

Or as a decorator:

# lock name chosen based on __file__
def myfunc():
    print("This function cannot be executed concurrently!")

# lock name is "myprocess"
def myfunc():
    print("This function cannot be executed concurrently!")

The name is up to you. The lock is specific to the name. The name is system global (as global as the file system is).

There are also some handy features for locking your whole program.

  1. The name argument is optional and defaults to the filename of the module that contains the function that called Lock (i.e. your Python source file), using inspect.stack(), which results in the Lock being automatically exclusive to all invocations of your application.

  2. When you set the optional die keyword argument to True, Lock will print an error to STDERR and exit the process immediately with exit code 1 if the lock cannot be acquired, instead of rasing an exception.

  3. The lock can be acquired with .forever(), instead of with or decorator syntax, in which case the lock will be released only at program exit using atexit.

With these features, you can make your whole program exclusive by placing the following line at the start of your program:

# At program start.
# program exits here if lock could not be acquired

If two such programs are run conncurrently you will see on STDERR:

Another '/home/user/' process is already running (pid 27922).


The with object can be captured if you want to see where the lockfile is stored:

with Lock(name="test1") as lock:

# outputs:

The Lock class logs every lock acquired and released to

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

exclusiveprocess-0.9.4.tar.gz (9.0 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page