22

✏️ Editing: multiprocessing.html

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<html xmlns="http://www.w3.org/1999/xhtml">
 <head>
 <meta http-equiv="X-UA-Compatible" content="IE=Edge" />
 <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
 <title>16.6. multiprocessing — Process-based “threading” interface &#8212; Python 2.7.16 documentation</title>
 <link rel="stylesheet" href="../_static/classic.css" type="text/css" />
 <link rel="stylesheet" href="../_static/pygments.css" type="text/css" />
 <script type="text/javascript" id="documentation_options" data-url_root="../" src="../_static/documentation_options.js"></script>
 <script type="text/javascript" src="../_static/jquery.js"></script>
 <script type="text/javascript" src="../_static/underscore.js"></script>
 <script type="text/javascript" src="../_static/doctools.js"></script>
 <script type="text/javascript" src="../_static/sidebar.js"></script>
 <link rel="search" type="application/opensearchdescription+xml"
 title="Search within Python 2.7.16 documentation"
 href="../_static/opensearch.xml"/>
 <link rel="author" title="About these documents" href="../about.html" />
 <link rel="index" title="Index" href="../genindex.html" />
 <link rel="search" title="Search" href="../search.html" />
 <link rel="copyright" title="Copyright" href="../copyright.html" />
 <link rel="next" title="16.7. mmap — Memory-mapped file support" href="mmap.html" />
 <link rel="prev" title="16.5. dummy_thread — Drop-in replacement for the thread module" href="dummy_thread.html" />
 <link rel="shortcut icon" type="image/png" href="../_static/py.png" />
 <link rel="canonical" href="https://docs.python.org/2/library/multiprocessing.html" />
 <script type="text/javascript" src="../_static/copybutton.js"></script>

</head><body> 
 <div class="related" role="navigation" aria-label="related navigation">
 <h3>Navigation</h3>
 <ul>
 <li class="right" style="margin-right: 10px">
 <a href="../genindex.html" title="General Index"
 accesskey="I">index</a></li>
 <li class="right" >
 <a href="../py-modindex.html" title="Python Module Index"
 >modules</a> |</li>
 <li class="right" >
 <a href="mmap.html" title="16.7. mmap — Memory-mapped file support"
 accesskey="N">next</a> |</li>
 <li class="right" >
 <a href="dummy_thread.html" title="16.5. dummy_thread — Drop-in replacement for the thread module"
 accesskey="P">previous</a> |</li>
 <li><img src="../_static/py.png" alt=""
 style="vertical-align: middle; margin-top: -1px"/></li>
 <li><a href="https://www.python.org/">Python</a> &#187;</li>
 <li>
 <a href="../index.html">Python 2.7.16 documentation</a> &#187;
 </li>

<li class="nav-item nav-item-1"><a href="index.html" >The Python Standard Library</a> &#187;</li>
 <li class="nav-item nav-item-2"><a href="someos.html" accesskey="U">16. Optional Operating System Services</a> &#187;</li> 
 </ul>
 </div>

<div class="document">
 <div class="documentwrapper">
 <div class="bodywrapper">
 <div class="body" role="main">
 
 <div class="section" id="module-multiprocessing">
<h1>16.6. <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> — Process-based “threading” interface<a class="headerlink" href="#module-multiprocessing" title="Permalink to this headline">¶</a></h1>
<div class="versionadded">
New in version 2.6.
</div>
<div class="section" id="introduction">
<h2>16.6.1. Introduction<a class="headerlink" href="#introduction" title="Permalink to this headline">¶</a></h2>
<a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> is a package that supports spawning processes using an
API similar to the <a class="reference internal" href="threading.html#module-threading" title="threading: Higher-level threading interface."><code class="xref py py-mod docutils literal notranslate">threading</code></a> module. The <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> package
offers both local and remote concurrency, effectively side-stepping the
<a class="reference internal" href="../glossary.html#term-global-interpreter-lock">Global Interpreter Lock</a> by using subprocesses instead of threads. Due
to this, the <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> module allows the programmer to fully
leverage multiple processors on a given machine. It runs on both Unix and
Windows.
The <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> module also introduces APIs which do not have
analogs in the <a class="reference internal" href="threading.html#module-threading" title="threading: Higher-level threading interface."><code class="xref py py-mod docutils literal notranslate">threading</code></a> module. A prime example of this is the
<code class="xref py py-class docutils literal notranslate">Pool</code> object which offers a convenient means of parallelizing the
execution of a function across multiple input values, distributing the
input data across processes (data parallelism). The following example
demonstrates the common practice of defining such functions in a module so
that child processes can successfully import that module. This basic example
of data parallelism using <code class="xref py py-class docutils literal notranslate">Pool</code>,
<div class="highlight-default notranslate"><div class="highlight"><pre>from multiprocessing import Pool

def f(name):
 print &#39;hello&#39;, name

def info(title):
 print title
 print &#39;module name:&#39;, __name__
 if hasattr(os, &#39;getppid&#39;): # only available on Unix
 print &#39;parent process:&#39;, os.getppid()
 print &#39;process id:&#39;, os.getpid()

def f(name):
 info(&#39;function f&#39;)
 print &#39;hello&#39;, name

def f(q):
 q.put([42, None, &#39;hello&#39;])

def f(conn):
 conn.send([42, None, &#39;hello&#39;])
 conn.close()

def f(l, i):
 l.acquire()
 print &#39;hello world&#39;, i
 l.release()

for num in range(10):
 Process(target=f, args=(lock, num)).start()
</pre></div>
</div>
Without using the lock output from the different processes is liable to get all
mixed up.
</div>
<div class="section" id="sharing-state-between-processes">
<h3>16.6.1.4. Sharing state between processes<a class="headerlink" href="#sharing-state-between-processes" title="Permalink to this headline">¶</a></h3>
As mentioned above, when doing concurrent programming it is usually best to
avoid using shared state as far as possible. This is particularly true when
using multiple processes.
However, if you really do need to use some shared data then
<a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> provides a couple of ways of doing so.
Shared memory
<blockquote>
<div>Data can be stored in a shared memory map using <a class="reference internal" href="#multiprocessing.Value" title="multiprocessing.Value"><code class="xref py py-class docutils literal notranslate">Value</code></a> or
<a class="reference internal" href="#multiprocessing.Array" title="multiprocessing.Array"><code class="xref py py-class docutils literal notranslate">Array</code></a>. For example, the following code
<div class="highlight-default notranslate"><div class="highlight"><pre>from multiprocessing import Process, Value, Array

def f(n, a):
 n.value = 3.1415927
 for i in range(len(a)):
 a[i] = -a[i]

print num.value
 print arr[:]
</pre></div>
</div>
will print
<div class="highlight-default notranslate"><div class="highlight"><pre>3.1415927
[0, -1, -2, -3, -4, -5, -6, -7, -8, -9]
</pre></div>
</div>
The <code class="docutils literal notranslate">'d'</code> and <code class="docutils literal notranslate">'i'</code> arguments used when creating <code class="docutils literal notranslate">num</code> and <code class="docutils literal notranslate">arr</code> are
typecodes of the kind used by the <a class="reference internal" href="array.html#module-array" title="array: Space efficient arrays of uniformly typed numeric values."><code class="xref py py-mod docutils literal notranslate">array</code></a> module: <code class="docutils literal notranslate">'d'</code> indicates a
double precision float and <code class="docutils literal notranslate">'i'</code> indicates a signed integer. These shared
objects will be process and thread-safe.
For more flexibility in using shared memory one can use the
<a class="reference internal" href="#module-multiprocessing.sharedctypes" title="multiprocessing.sharedctypes: Allocate ctypes objects from shared memory."><code class="xref py py-mod docutils literal notranslate">multiprocessing.sharedctypes</code></a> module which supports the creation of
arbitrary ctypes objects allocated from shared memory.
</div></blockquote>
Server process
<blockquote>
<div>A manager object returned by <code class="xref py py-func docutils literal notranslate">Manager()</code> controls a server process which
holds Python objects and allows other processes to manipulate them using
proxies.
A manager returned by <code class="xref py py-func docutils literal notranslate">Manager()</code> will support types <code class="xref py py-class docutils literal notranslate">list</code>,
<a class="reference internal" href="stdtypes.html#dict" title="dict"><code class="xref py py-class docutils literal notranslate">dict</code></a>, <a class="reference internal" href="#multiprocessing.managers.Namespace" title="multiprocessing.managers.Namespace"><code class="xref py py-class docutils literal notranslate">Namespace</code></a>, <a class="reference internal" href="#multiprocessing.Lock" title="multiprocessing.Lock"><code class="xref py py-class docutils literal notranslate">Lock</code></a>, <a class="reference internal" href="#multiprocessing.RLock" title="multiprocessing.RLock"><code class="xref py py-class docutils literal notranslate">RLock</code></a>,
<a class="reference internal" href="#multiprocessing.Semaphore" title="multiprocessing.Semaphore"><code class="xref py py-class docutils literal notranslate">Semaphore</code></a>, <a class="reference internal" href="#multiprocessing.BoundedSemaphore" title="multiprocessing.BoundedSemaphore"><code class="xref py py-class docutils literal notranslate">BoundedSemaphore</code></a>, <a class="reference internal" href="#multiprocessing.Condition" title="multiprocessing.Condition"><code class="xref py py-class docutils literal notranslate">Condition</code></a>,
<a class="reference internal" href="#multiprocessing.Event" title="multiprocessing.Event"><code class="xref py py-class docutils literal notranslate">Event</code></a>, <a class="reference internal" href="#multiprocessing.Queue" title="multiprocessing.Queue"><code class="xref py py-class docutils literal notranslate">Queue</code></a>, <a class="reference internal" href="#multiprocessing.Value" title="multiprocessing.Value"><code class="xref py py-class docutils literal notranslate">Value</code></a> and <a class="reference internal" href="#multiprocessing.Array" title="multiprocessing.Array"><code class="xref py py-class docutils literal notranslate">Array</code></a>. For
example,
<div class="highlight-default notranslate"><div class="highlight"><pre>from multiprocessing import Process, Manager

def f(d, l):
 d[1] = &#39;1&#39;
 d[&#39;2&#39;] = 2
 d[0.25] = None
 l.reverse()

d = manager.dict()
 l = manager.list(range(10))

print d
 print l
</pre></div>
</div>
will print
<div class="highlight-default notranslate"><div class="highlight"><pre>{0.25: None, 1: &#39;1&#39;, &#39;2&#39;: 2}
[9, 8, 7, 6, 5, 4, 3, 2, 1, 0]
</pre></div>
</div>
Server process managers are more flexible than using shared memory objects
because they can be made to support arbitrary object types. Also, a single
manager can be shared by processes on different computers over a network.
They are, however, slower than using shared memory.
</div></blockquote>
</div>
<div class="section" id="using-a-pool-of-workers">
<h3>16.6.1.5. Using a pool of workers<a class="headerlink" href="#using-a-pool-of-workers" title="Permalink to this headline">¶</a></h3>
The <code class="xref py py-class docutils literal notranslate">Pool</code> class represents a pool of worker
processes. It has methods which allows tasks to be offloaded to the worker
processes in a few different ways.
For example:
<div class="highlight-default notranslate"><div class="highlight"><pre>from multiprocessing import Pool, TimeoutError
import time
import os

# print &quot;[0, 1, 4,..., 81]&quot;
 print pool.map(f, range(10))

# print same numbers in arbitrary order
 for i in pool.imap_unordered(f, range(10)):
 print i

# evaluate &quot;f(20)&quot; asynchronously
 res = pool.apply_async(f, (20,)) # runs in *only* one process
 print res.get(timeout=1) # prints &quot;400&quot;

# evaluate &quot;os.getpid()&quot; asynchronously
 res = pool.apply_async(os.getpid, ()) # runs in *only* one process
 print res.get(timeout=1) # prints the PID of that process

# launching multiple evaluations asynchronously *may* use more processes
 multiple_results = [pool.apply_async(os.getpid, ()) for i in range(4)]
 print [res.get(timeout=1) for res in multiple_results]

# make a single worker sleep for 10 secs
 res = pool.apply_async(time.sleep, (10,))
 try:
 print res.get(timeout=1)
 except TimeoutError:
 print &quot;We lacked patience and got a multiprocessing.TimeoutError&quot;
</pre></div>
</div>
Note that the methods of a pool should only ever be used by the
process which created it.
<div class="admonition note">
Note
Functionality within this package requires that the <code class="docutils literal notranslate">__main__</code> module be
importable by the children. This is covered in <a class="reference internal" href="#multiprocessing-programming">Programming guidelines</a>
however it is worth pointing out here. This means that some examples, such
as the <code class="xref py py-class docutils literal notranslate">Pool</code> examples will not work in the interactive interpreter.
For example:
<div class="highlight-default notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from multiprocessing import Pool
&gt;&gt;&gt; p = Pool(5)
&gt;&gt;&gt; def f(x):
... return x*x
...
&gt;&gt;&gt; p.map(f, [1,2,3])
Process PoolWorker-1:
Process PoolWorker-2:
Process PoolWorker-3:
Traceback (most recent call last):
AttributeError: &#39;module&#39; object has no attribute &#39;f&#39;
AttributeError: &#39;module&#39; object has no attribute &#39;f&#39;
AttributeError: &#39;module&#39; object has no attribute &#39;f&#39;
</pre></div>
</div>
(If you try this it will actually output three full tracebacks
interleaved in a semi-random fashion, and then you may have to
stop the master process somehow.)
</div>
</div>
</div>
<div class="section" id="reference">
<h2>16.6.2. Reference<a class="headerlink" href="#reference" title="Permalink to this headline">¶</a></h2>
The <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> package mostly replicates the API of the
<a class="reference internal" href="threading.html#module-threading" title="threading: Higher-level threading interface."><code class="xref py py-mod docutils literal notranslate">threading</code></a> module.
<div class="section" id="process-and-exceptions">
<h3>16.6.2.1. <a class="reference internal" href="#multiprocessing.Process" title="multiprocessing.Process"><code class="xref py py-class docutils literal notranslate">Process</code></a> and exceptions<a class="headerlink" href="#process-and-exceptions" title="Permalink to this headline">¶</a></h3>
<dl class="class">
<dt id="multiprocessing.Process">
class <code class="descclassname">multiprocessing.</code><code class="descname">Process</code>(group=None, target=None, name=None, args=(), kwargs={})<a class="headerlink" href="#multiprocessing.Process" title="Permalink to this definition">¶</a></dt>
<dd>Process objects represent activity that is run in a separate process. The
<a class="reference internal" href="#multiprocessing.Process" title="multiprocessing.Process"><code class="xref py py-class docutils literal notranslate">Process</code></a> class has equivalents of all the methods of
<a class="reference internal" href="threading.html#threading.Thread" title="threading.Thread"><code class="xref py py-class docutils literal notranslate">threading.Thread</code></a>.
The constructor should always be called with keyword arguments. group
should always be <code class="docutils literal notranslate">None</code>; it exists solely for compatibility with
<a class="reference internal" href="threading.html#threading.Thread" title="threading.Thread"><code class="xref py py-class docutils literal notranslate">threading.Thread</code></a>. target is the callable object to be invoked by
the <a class="reference internal" href="#multiprocessing.Process.run" title="multiprocessing.Process.run"><code class="xref py py-meth docutils literal notranslate">run()</code></a> method. It defaults to <code class="docutils literal notranslate">None</code>, meaning nothing is
called. name is the process name. By default, a unique name is constructed
of the form ‘Process-N1:N2:…:Nk’ where N1,N2,…,Nk is a sequence of integers whose length
is determined by the generation of the process. args is the argument
tuple for the target invocation. kwargs is a dictionary of keyword
arguments for the target invocation. By default, no arguments are passed to
target.
If a subclass overrides the constructor, it must make sure it invokes the
base class constructor (<code class="xref py py-meth docutils literal notranslate">Process.__init__()</code>) before doing anything else
to the process.
<dl class="method">
<dt id="multiprocessing.Process.run">
<code class="descname">run</code>()<a class="headerlink" href="#multiprocessing.Process.run" title="Permalink to this definition">¶</a></dt>
<dd>Method representing the process’s activity.
You may override this method in a subclass. The standard <a class="reference internal" href="#multiprocessing.Process.run" title="multiprocessing.Process.run"><code class="xref py py-meth docutils literal notranslate">run()</code></a>
method invokes the callable object passed to the object’s constructor as
the target argument, if any, with sequential and keyword arguments taken
from the args and kwargs arguments, respectively.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Process.start">
<code class="descname">start</code>()<a class="headerlink" href="#multiprocessing.Process.start" title="Permalink to this definition">¶</a></dt>
<dd>Start the process’s activity.
This must be called at most once per process object. It arranges for the
object’s <a class="reference internal" href="#multiprocessing.Process.run" title="multiprocessing.Process.run"><code class="xref py py-meth docutils literal notranslate">run()</code></a> method to be invoked in a separate process.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Process.join">
<code class="descname">join</code>([timeout])<a class="headerlink" href="#multiprocessing.Process.join" title="Permalink to this definition">¶</a></dt>
<dd>Block the calling thread until the process whose <a class="reference internal" href="#multiprocessing.Process.join" title="multiprocessing.Process.join"><code class="xref py py-meth docutils literal notranslate">join()</code></a> method is
called terminates or until the optional timeout occurs.
If timeout is <code class="docutils literal notranslate">None</code> then there is no timeout.
A process can be joined many times.
A process cannot join itself because this would cause a deadlock. It is
an error to attempt to join a process before it has been started.
</dd></dl>

<dl class="attribute">
<dt id="multiprocessing.Process.name">
<code class="descname">name</code><a class="headerlink" href="#multiprocessing.Process.name" title="Permalink to this definition">¶</a></dt>
<dd>The process’s name.
The name is a string used for identification purposes only. It has no
semantics. Multiple processes may be given the same name. The initial
name is set by the constructor.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Process.is_alive">
<code class="descname">is_alive</code>()<a class="headerlink" href="#multiprocessing.Process.is_alive" title="Permalink to this definition">¶</a></dt>
<dd>Return whether the process is alive.
Roughly, a process object is alive from the moment the <a class="reference internal" href="#multiprocessing.Process.start" title="multiprocessing.Process.start"><code class="xref py py-meth docutils literal notranslate">start()</code></a>
method returns until the child process terminates.
</dd></dl>

<dl class="attribute">
<dt id="multiprocessing.Process.daemon">
<code class="descname">daemon</code><a class="headerlink" href="#multiprocessing.Process.daemon" title="Permalink to this definition">¶</a></dt>
<dd>The process’s daemon flag, a Boolean value. This must be set before
<a class="reference internal" href="#multiprocessing.Process.start" title="multiprocessing.Process.start"><code class="xref py py-meth docutils literal notranslate">start()</code></a> is called.
The initial value is inherited from the creating process.
When a process exits, it attempts to terminate all of its daemonic child
processes.
Note that a daemonic process is not allowed to create child processes.
Otherwise a daemonic process would leave its children orphaned if it gets
terminated when its parent process exits. Additionally, these are not
Unix daemons or services, they are normal processes that will be
terminated (and not joined) if non-daemonic processes have exited.
</dd></dl>

In addition to the <a class="reference internal" href="threading.html#threading.Thread" title="threading.Thread"><code class="xref py py-class docutils literal notranslate">threading.Thread</code></a> API, <a class="reference internal" href="#multiprocessing.Process" title="multiprocessing.Process"><code class="xref py py-class docutils literal notranslate">Process</code></a> objects
also support the following attributes and methods:
<dl class="attribute">
<dt id="multiprocessing.Process.pid">
<code class="descname">pid</code><a class="headerlink" href="#multiprocessing.Process.pid" title="Permalink to this definition">¶</a></dt>
<dd>Return the process ID. Before the process is spawned, this will be
<code class="docutils literal notranslate">None</code>.
</dd></dl>

<dl class="attribute">
<dt id="multiprocessing.Process.exitcode">
<code class="descname">exitcode</code><a class="headerlink" href="#multiprocessing.Process.exitcode" title="Permalink to this definition">¶</a></dt>
<dd>The child’s exit code. This will be <code class="docutils literal notranslate">None</code> if the process has not yet
terminated. A negative value -N indicates that the child was terminated
by signal N.
</dd></dl>

<dl class="attribute">
<dt id="multiprocessing.Process.authkey">
<code class="descname">authkey</code><a class="headerlink" href="#multiprocessing.Process.authkey" title="Permalink to this definition">¶</a></dt>
<dd>The process’s authentication key (a byte string).
When <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> is initialized the main process is assigned a
random string using <a class="reference internal" href="os.html#os.urandom" title="os.urandom"><code class="xref py py-func docutils literal notranslate">os.urandom()</code></a>.
When a <a class="reference internal" href="#multiprocessing.Process" title="multiprocessing.Process"><code class="xref py py-class docutils literal notranslate">Process</code></a> object is created, it will inherit the
authentication key of its parent process, although this may be changed by
setting <a class="reference internal" href="#multiprocessing.Process.authkey" title="multiprocessing.Process.authkey"><code class="xref py py-attr docutils literal notranslate">authkey</code></a> to another byte string.
See <a class="reference internal" href="#multiprocessing-auth-keys">Authentication keys</a>.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Process.terminate">
<code class="descname">terminate</code>()<a class="headerlink" href="#multiprocessing.Process.terminate" title="Permalink to this definition">¶</a></dt>
<dd>Terminate the process. On Unix this is done using the <code class="docutils literal notranslate">SIGTERM</code> signal;
on Windows <code class="xref c c-func docutils literal notranslate">TerminateProcess()</code> is used. Note that exit handlers and
finally clauses, etc., will not be executed.
Note that descendant processes of the process will not be terminated –
they will simply become orphaned.
<div class="admonition warning">
Warning
If this method is used when the associated process is using a pipe or
queue then the pipe or queue is liable to become corrupted and may
become unusable by other process. Similarly, if the process has
acquired a lock or semaphore etc. then terminating it is liable to
cause other processes to deadlock.
</div>
</dd></dl>

Note that the <a class="reference internal" href="#multiprocessing.Process.start" title="multiprocessing.Process.start"><code class="xref py py-meth docutils literal notranslate">start()</code></a>, <a class="reference internal" href="#multiprocessing.Process.join" title="multiprocessing.Process.join"><code class="xref py py-meth docutils literal notranslate">join()</code></a>, <a class="reference internal" href="#multiprocessing.Process.is_alive" title="multiprocessing.Process.is_alive"><code class="xref py py-meth docutils literal notranslate">is_alive()</code></a>,
<a class="reference internal" href="#multiprocessing.Process.terminate" title="multiprocessing.Process.terminate"><code class="xref py py-meth docutils literal notranslate">terminate()</code></a> and <a class="reference internal" href="#multiprocessing.Process.exitcode" title="multiprocessing.Process.exitcode"><code class="xref py py-attr docutils literal notranslate">exitcode</code></a> methods should only be called by
the process that created the process object.
Example usage of some of the methods of <a class="reference internal" href="#multiprocessing.Process" title="multiprocessing.Process"><code class="xref py py-class docutils literal notranslate">Process</code></a>:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; import multiprocessing, time, signal
&gt;&gt;&gt; p = multiprocessing.Process(target=time.sleep, args=(1000,))
&gt;&gt;&gt; print p, p.is_alive()
&lt;Process(Process-1, initial)&gt; False
&gt;&gt;&gt; p.start()
&gt;&gt;&gt; print p, p.is_alive()
&lt;Process(Process-1, started)&gt; True
&gt;&gt;&gt; p.terminate()
&gt;&gt;&gt; time.sleep(0.1)
&gt;&gt;&gt; print p, p.is_alive()
&lt;Process(Process-1, stopped[SIGTERM])&gt; False
&gt;&gt;&gt; p.exitcode == -signal.SIGTERM
True
</pre></div>
</div>
</dd></dl>

<dl class="exception">
<dt id="multiprocessing.BufferTooShort">
exception <code class="descclassname">multiprocessing.</code><code class="descname">BufferTooShort</code><a class="headerlink" href="#multiprocessing.BufferTooShort" title="Permalink to this definition">¶</a></dt>
<dd>Exception raised by <a class="reference internal" href="#Connection.recv_bytes_into" title="Connection.recv_bytes_into"><code class="xref py py-meth docutils literal notranslate">Connection.recv_bytes_into()</code></a> when the supplied
buffer object is too small for the message read.
If <code class="docutils literal notranslate">e</code> is an instance of <a class="reference internal" href="#multiprocessing.BufferTooShort" title="multiprocessing.BufferTooShort"><code class="xref py py-exc docutils literal notranslate">BufferTooShort</code></a> then <code class="docutils literal notranslate">e.args[0]</code> will give
the message as a byte string.
</dd></dl>

</div>
<div class="section" id="pipes-and-queues">
<h3>16.6.2.2. Pipes and Queues<a class="headerlink" href="#pipes-and-queues" title="Permalink to this headline">¶</a></h3>
When using multiple processes, one generally uses message passing for
communication between processes and avoids having to use any synchronization
primitives like locks.
For passing messages one can use <a class="reference internal" href="#multiprocessing.Pipe" title="multiprocessing.Pipe"><code class="xref py py-func docutils literal notranslate">Pipe()</code></a> (for a connection between two
processes) or a queue (which allows multiple producers and consumers).
The <a class="reference internal" href="#multiprocessing.Queue" title="multiprocessing.Queue"><code class="xref py py-class docutils literal notranslate">Queue</code></a>, <a class="reference internal" href="#multiprocessing.multiprocessing.queues.SimpleQueue" title="multiprocessing.multiprocessing.queues.SimpleQueue"><code class="xref py py-class docutils literal notranslate">multiprocessing.queues.SimpleQueue</code></a> and <a class="reference internal" href="#multiprocessing.JoinableQueue" title="multiprocessing.JoinableQueue"><code class="xref py py-class docutils literal notranslate">JoinableQueue</code></a> types are multi-producer,
multi-consumer FIFO queues modelled on the <a class="reference internal" href="queue.html#Queue.Queue" title="Queue.Queue"><code class="xref py py-class docutils literal notranslate">Queue.Queue</code></a> class in the
standard library. They differ in that <a class="reference internal" href="#multiprocessing.Queue" title="multiprocessing.Queue"><code class="xref py py-class docutils literal notranslate">Queue</code></a> lacks the
<a class="reference internal" href="queue.html#Queue.Queue.task_done" title="Queue.Queue.task_done"><code class="xref py py-meth docutils literal notranslate">task_done()</code></a> and <a class="reference internal" href="queue.html#Queue.Queue.join" title="Queue.Queue.join"><code class="xref py py-meth docutils literal notranslate">join()</code></a> methods introduced
into Python 2.5’s <a class="reference internal" href="queue.html#Queue.Queue" title="Queue.Queue"><code class="xref py py-class docutils literal notranslate">Queue.Queue</code></a> class.
If you use <a class="reference internal" href="#multiprocessing.JoinableQueue" title="multiprocessing.JoinableQueue"><code class="xref py py-class docutils literal notranslate">JoinableQueue</code></a> then you must call
<a class="reference internal" href="#multiprocessing.JoinableQueue.task_done" title="multiprocessing.JoinableQueue.task_done"><code class="xref py py-meth docutils literal notranslate">JoinableQueue.task_done()</code></a> for each task removed from the queue or else the
semaphore used to count the number of unfinished tasks may eventually overflow,
raising an exception.
Note that one can also create a shared queue by using a manager object – see
<a class="reference internal" href="#multiprocessing-managers">Managers</a>.
<div class="admonition note">
Note
<a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> uses the usual <a class="reference internal" href="queue.html#Queue.Empty" title="Queue.Empty"><code class="xref py py-exc docutils literal notranslate">Queue.Empty</code></a> and
<a class="reference internal" href="queue.html#Queue.Full" title="Queue.Full"><code class="xref py py-exc docutils literal notranslate">Queue.Full</code></a> exceptions to signal a timeout. They are not available in
the <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> namespace so you need to import them from
<a class="reference internal" href="queue.html#module-Queue" title="Queue: A synchronized queue class."><code class="xref py py-mod docutils literal notranslate">Queue</code></a>.
</div>
<div class="admonition note">
Note
When an object is put on a queue, the object is pickled and a
background thread later flushes the pickled data to an underlying
pipe. This has some consequences which are a little surprising,
but should not cause any practical difficulties – if they really
bother you then you can instead use a queue created with a
<a class="reference internal" href="#multiprocessing-managers">manager</a>.
<ol class="last arabic simple">
<li>After putting an object on an empty queue there may be an
infinitesimal delay before the queue’s <a class="reference internal" href="#multiprocessing.Queue.empty" title="multiprocessing.Queue.empty"><code class="xref py py-meth docutils literal notranslate">empty()</code></a>
method returns <a class="reference internal" href="constants.html#False" title="False"><code class="xref py py-const docutils literal notranslate">False</code></a> and <a class="reference internal" href="#multiprocessing.Queue.get_nowait" title="multiprocessing.Queue.get_nowait"><code class="xref py py-meth docutils literal notranslate">get_nowait()</code></a> can
return without raising <a class="reference internal" href="queue.html#Queue.Empty" title="Queue.Empty"><code class="xref py py-exc docutils literal notranslate">Queue.Empty</code></a>.</li>
<li>If multiple processes are enqueuing objects, it is possible for
the objects to be received at the other end out-of-order.
However, objects enqueued by the same process will always be in
the expected order with respect to each other.</li>
</ol>
</div>
<div class="admonition warning">
Warning
If a process is killed using <a class="reference internal" href="#multiprocessing.Process.terminate" title="multiprocessing.Process.terminate"><code class="xref py py-meth docutils literal notranslate">Process.terminate()</code></a> or <a class="reference internal" href="os.html#os.kill" title="os.kill"><code class="xref py py-func docutils literal notranslate">os.kill()</code></a>
while it is trying to use a <a class="reference internal" href="#multiprocessing.Queue" title="multiprocessing.Queue"><code class="xref py py-class docutils literal notranslate">Queue</code></a>, then the data in the queue is
likely to become corrupted. This may cause any other process to get an
exception when it tries to use the queue later on.
</div>
<div class="admonition warning">
Warning
As mentioned above, if a child process has put items on a queue (and it has
not used <a class="reference internal" href="#multiprocessing.Queue.cancel_join_thread" title="multiprocessing.Queue.cancel_join_thread"><code class="xref py py-meth docutils literal notranslate">JoinableQueue.cancel_join_thread</code></a>), then that process will
not terminate until all buffered items have been flushed to the pipe.
This means that if you try joining that process you may get a deadlock unless
you are sure that all items which have been put on the queue have been
consumed. Similarly, if the child process is non-daemonic then the parent
process may hang on exit when it tries to join all its non-daemonic children.
Note that a queue created using a manager does not have this issue. See
<a class="reference internal" href="#multiprocessing-programming">Programming guidelines</a>.
</div>
For an example of the usage of queues for interprocess communication see
<a class="reference internal" href="#multiprocessing-examples">Examples</a>.
<dl class="function">
<dt id="multiprocessing.Pipe">
<code class="descclassname">multiprocessing.</code><code class="descname">Pipe</code>([duplex])<a class="headerlink" href="#multiprocessing.Pipe" title="Permalink to this definition">¶</a></dt>
<dd>Returns a pair <code class="docutils literal notranslate">(conn1, conn2)</code> of <a class="reference internal" href="#Connection" title="Connection"><code class="xref py py-class docutils literal notranslate">Connection</code></a> objects representing
the ends of a pipe.
If duplex is <code class="docutils literal notranslate">True</code> (the default) then the pipe is bidirectional. If
duplex is <code class="docutils literal notranslate">False</code> then the pipe is unidirectional: <code class="docutils literal notranslate">conn1</code> can only be
used for receiving messages and <code class="docutils literal notranslate">conn2</code> can only be used for sending
messages.
</dd></dl>

<dl class="class">
<dt id="multiprocessing.Queue">
class <code class="descclassname">multiprocessing.</code><code class="descname">Queue</code>([maxsize])<a class="headerlink" href="#multiprocessing.Queue" title="Permalink to this definition">¶</a></dt>
<dd>Returns a process shared queue implemented using a pipe and a few
locks/semaphores. When a process first puts an item on the queue a feeder
thread is started which transfers objects from a buffer into the pipe.
The usual <a class="reference internal" href="queue.html#Queue.Empty" title="Queue.Empty"><code class="xref py py-exc docutils literal notranslate">Queue.Empty</code></a> and <a class="reference internal" href="queue.html#Queue.Full" title="Queue.Full"><code class="xref py py-exc docutils literal notranslate">Queue.Full</code></a> exceptions from the
standard library’s <a class="reference internal" href="queue.html#module-Queue" title="Queue: A synchronized queue class."><code class="xref py py-mod docutils literal notranslate">Queue</code></a> module are raised to signal timeouts.
<a class="reference internal" href="#multiprocessing.Queue" title="multiprocessing.Queue"><code class="xref py py-class docutils literal notranslate">Queue</code></a> implements all the methods of <a class="reference internal" href="queue.html#Queue.Queue" title="Queue.Queue"><code class="xref py py-class docutils literal notranslate">Queue.Queue</code></a> except for
<a class="reference internal" href="queue.html#Queue.Queue.task_done" title="Queue.Queue.task_done"><code class="xref py py-meth docutils literal notranslate">task_done()</code></a> and <a class="reference internal" href="queue.html#Queue.Queue.join" title="Queue.Queue.join"><code class="xref py py-meth docutils literal notranslate">join()</code></a>.
<dl class="method">
<dt id="multiprocessing.Queue.qsize">
<code class="descname">qsize</code>()<a class="headerlink" href="#multiprocessing.Queue.qsize" title="Permalink to this definition">¶</a></dt>
<dd>Return the approximate size of the queue. Because of
multithreading/multiprocessing semantics, this number is not reliable.
Note that this may raise <a class="reference internal" href="exceptions.html#exceptions.NotImplementedError" title="exceptions.NotImplementedError"><code class="xref py py-exc docutils literal notranslate">NotImplementedError</code></a> on Unix platforms like
Mac OS X where <code class="docutils literal notranslate">sem_getvalue()</code> is not implemented.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Queue.empty">
<code class="descname">empty</code>()<a class="headerlink" href="#multiprocessing.Queue.empty" title="Permalink to this definition">¶</a></dt>
<dd>Return <code class="docutils literal notranslate">True</code> if the queue is empty, <code class="docutils literal notranslate">False</code> otherwise. Because of
multithreading/multiprocessing semantics, this is not reliable.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Queue.full">
<code class="descname">full</code>()<a class="headerlink" href="#multiprocessing.Queue.full" title="Permalink to this definition">¶</a></dt>
<dd>Return <code class="docutils literal notranslate">True</code> if the queue is full, <code class="docutils literal notranslate">False</code> otherwise. Because of
multithreading/multiprocessing semantics, this is not reliable.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Queue.put">
<code class="descname">put</code>(obj[, block[, timeout]])<a class="headerlink" href="#multiprocessing.Queue.put" title="Permalink to this definition">¶</a></dt>
<dd>Put obj into the queue. If the optional argument block is <code class="docutils literal notranslate">True</code>
(the default) and timeout is <code class="docutils literal notranslate">None</code> (the default), block if necessary until
a free slot is available. If timeout is a positive number, it blocks at
most timeout seconds and raises the <a class="reference internal" href="queue.html#Queue.Full" title="Queue.Full"><code class="xref py py-exc docutils literal notranslate">Queue.Full</code></a> exception if no
free slot was available within that time. Otherwise (block is
<code class="docutils literal notranslate">False</code>), put an item on the queue if a free slot is immediately
available, else raise the <a class="reference internal" href="queue.html#Queue.Full" title="Queue.Full"><code class="xref py py-exc docutils literal notranslate">Queue.Full</code></a> exception (timeout is
ignored in that case).
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Queue.put_nowait">
<code class="descname">put_nowait</code>(obj)<a class="headerlink" href="#multiprocessing.Queue.put_nowait" title="Permalink to this definition">¶</a></dt>
<dd>Equivalent to <code class="docutils literal notranslate">put(obj, False)</code>.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Queue.get">
<code class="descname">get</code>([block[, timeout]])<a class="headerlink" href="#multiprocessing.Queue.get" title="Permalink to this definition">¶</a></dt>
<dd>Remove and return an item from the queue. If optional args block is
<code class="docutils literal notranslate">True</code> (the default) and timeout is <code class="docutils literal notranslate">None</code> (the default), block if
necessary until an item is available. If timeout is a positive number,
it blocks at most timeout seconds and raises the <a class="reference internal" href="queue.html#Queue.Empty" title="Queue.Empty"><code class="xref py py-exc docutils literal notranslate">Queue.Empty</code></a>
exception if no item was available within that time. Otherwise (block is
<code class="docutils literal notranslate">False</code>), return an item if one is immediately available, else raise the
<a class="reference internal" href="queue.html#Queue.Empty" title="Queue.Empty"><code class="xref py py-exc docutils literal notranslate">Queue.Empty</code></a> exception (timeout is ignored in that case).
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Queue.get_nowait">
<code class="descname">get_nowait</code>()<a class="headerlink" href="#multiprocessing.Queue.get_nowait" title="Permalink to this definition">¶</a></dt>
<dd>Equivalent to <code class="docutils literal notranslate">get(False)</code>.
</dd></dl>

<a class="reference internal" href="#multiprocessing.Queue" title="multiprocessing.Queue"><code class="xref py py-class docutils literal notranslate">Queue</code></a> has a few additional methods not found in
<a class="reference internal" href="queue.html#Queue.Queue" title="Queue.Queue"><code class="xref py py-class docutils literal notranslate">Queue.Queue</code></a>. These methods are usually unnecessary for most
code:
<dl class="method">
<dt id="multiprocessing.Queue.close">
<code class="descname">close</code>()<a class="headerlink" href="#multiprocessing.Queue.close" title="Permalink to this definition">¶</a></dt>
<dd>Indicate that no more data will be put on this queue by the current
process. The background thread will quit once it has flushed all buffered
data to the pipe. This is called automatically when the queue is garbage
collected.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Queue.join_thread">
<code class="descname">join_thread</code>()<a class="headerlink" href="#multiprocessing.Queue.join_thread" title="Permalink to this definition">¶</a></dt>
<dd>Join the background thread. This can only be used after <a class="reference internal" href="#multiprocessing.Queue.close" title="multiprocessing.Queue.close"><code class="xref py py-meth docutils literal notranslate">close()</code></a> has
been called. It blocks until the background thread exits, ensuring that
all data in the buffer has been flushed to the pipe.
By default if a process is not the creator of the queue then on exit it
will attempt to join the queue’s background thread. The process can call
<a class="reference internal" href="#multiprocessing.Queue.cancel_join_thread" title="multiprocessing.Queue.cancel_join_thread"><code class="xref py py-meth docutils literal notranslate">cancel_join_thread()</code></a> to make <a class="reference internal" href="#multiprocessing.Queue.join_thread" title="multiprocessing.Queue.join_thread"><code class="xref py py-meth docutils literal notranslate">join_thread()</code></a> do nothing.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Queue.cancel_join_thread">
<code class="descname">cancel_join_thread</code>()<a class="headerlink" href="#multiprocessing.Queue.cancel_join_thread" title="Permalink to this definition">¶</a></dt>
<dd>Prevent <a class="reference internal" href="#multiprocessing.Queue.join_thread" title="multiprocessing.Queue.join_thread"><code class="xref py py-meth docutils literal notranslate">join_thread()</code></a> from blocking. In particular, this prevents
the background thread from being joined automatically when the process
exits – see <a class="reference internal" href="#multiprocessing.Queue.join_thread" title="multiprocessing.Queue.join_thread"><code class="xref py py-meth docutils literal notranslate">join_thread()</code></a>.
A better name for this method might be
<code class="docutils literal notranslate">allow_exit_without_flush()</code>. It is likely to cause enqueued
data to lost, and you almost certainly will not need to use it.
It is really only there if you need the current process to exit
immediately without waiting to flush enqueued data to the
underlying pipe, and you don’t care about lost data.
</dd></dl>

<div class="admonition note">
Note
This class’s functionality requires a functioning shared semaphore
implementation on the host operating system. Without one, the
functionality in this class will be disabled, and attempts to
instantiate a <a class="reference internal" href="queue.html#module-Queue" title="Queue: A synchronized queue class."><code class="xref py py-class docutils literal notranslate">Queue</code></a> will result in an <a class="reference internal" href="exceptions.html#exceptions.ImportError" title="exceptions.ImportError"><code class="xref py py-exc docutils literal notranslate">ImportError</code></a>. See
<a class="reference external" href="https://bugs.python.org/issue3770">bpo-3770</a> for additional information. The same holds true for any
of the specialized queue types listed below.
</div>
</dd></dl>

<dl class="class">
<dt id="multiprocessing.multiprocessing.queues.SimpleQueue">
class <code class="descclassname">multiprocessing.queues.</code><code class="descname">SimpleQueue</code><a class="headerlink" href="#multiprocessing.multiprocessing.queues.SimpleQueue" title="Permalink to this definition">¶</a></dt>
<dd>It is a simplified <a class="reference internal" href="#multiprocessing.Queue" title="multiprocessing.Queue"><code class="xref py py-class docutils literal notranslate">Queue</code></a> type, very close to a locked <a class="reference internal" href="#multiprocessing.Pipe" title="multiprocessing.Pipe"><code class="xref py py-class docutils literal notranslate">Pipe</code></a>.
<dl class="method">
<dt id="multiprocessing.multiprocessing.queues.SimpleQueue.empty">
<code class="descname">empty</code>()<a class="headerlink" href="#multiprocessing.multiprocessing.queues.SimpleQueue.empty" title="Permalink to this definition">¶</a></dt>
<dd>Return <code class="docutils literal notranslate">True</code> if the queue is empty, <code class="docutils literal notranslate">False</code> otherwise.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.multiprocessing.queues.SimpleQueue.get">
<code class="descname">get</code>()<a class="headerlink" href="#multiprocessing.multiprocessing.queues.SimpleQueue.get" title="Permalink to this definition">¶</a></dt>
<dd>Remove and return an item from the queue.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.multiprocessing.queues.SimpleQueue.put">
<code class="descname">put</code>(item)<a class="headerlink" href="#multiprocessing.multiprocessing.queues.SimpleQueue.put" title="Permalink to this definition">¶</a></dt>
<dd>Put item into the queue.
</dd></dl>

</dd></dl>

<dl class="class">
<dt id="multiprocessing.JoinableQueue">
class <code class="descclassname">multiprocessing.</code><code class="descname">JoinableQueue</code>([maxsize])<a class="headerlink" href="#multiprocessing.JoinableQueue" title="Permalink to this definition">¶</a></dt>
<dd><a class="reference internal" href="#multiprocessing.JoinableQueue" title="multiprocessing.JoinableQueue"><code class="xref py py-class docutils literal notranslate">JoinableQueue</code></a>, a <a class="reference internal" href="#multiprocessing.Queue" title="multiprocessing.Queue"><code class="xref py py-class docutils literal notranslate">Queue</code></a> subclass, is a queue which
additionally has <a class="reference internal" href="#multiprocessing.JoinableQueue.task_done" title="multiprocessing.JoinableQueue.task_done"><code class="xref py py-meth docutils literal notranslate">task_done()</code></a> and <a class="reference internal" href="#multiprocessing.JoinableQueue.join" title="multiprocessing.JoinableQueue.join"><code class="xref py py-meth docutils literal notranslate">join()</code></a> methods.
<dl class="method">
<dt id="multiprocessing.JoinableQueue.task_done">
<code class="descname">task_done</code>()<a class="headerlink" href="#multiprocessing.JoinableQueue.task_done" title="Permalink to this definition">¶</a></dt>
<dd>Indicate that a formerly enqueued task is complete. Used by queue consumer
threads. For each <a class="reference internal" href="#multiprocessing.Queue.get" title="multiprocessing.Queue.get"><code class="xref py py-meth docutils literal notranslate">get()</code></a> used to fetch a task, a subsequent
call to <a class="reference internal" href="#multiprocessing.JoinableQueue.task_done" title="multiprocessing.JoinableQueue.task_done"><code class="xref py py-meth docutils literal notranslate">task_done()</code></a> tells the queue that the processing on the task
is complete.
If a <a class="reference internal" href="queue.html#Queue.Queue.join" title="Queue.Queue.join"><code class="xref py py-meth docutils literal notranslate">join()</code></a> is currently blocking, it will resume when all
items have been processed (meaning that a <a class="reference internal" href="#multiprocessing.JoinableQueue.task_done" title="multiprocessing.JoinableQueue.task_done"><code class="xref py py-meth docutils literal notranslate">task_done()</code></a> call was
received for every item that had been <a class="reference internal" href="#multiprocessing.Queue.put" title="multiprocessing.Queue.put"><code class="xref py py-meth docutils literal notranslate">put()</code></a> into the queue).
Raises a <a class="reference internal" href="exceptions.html#exceptions.ValueError" title="exceptions.ValueError"><code class="xref py py-exc docutils literal notranslate">ValueError</code></a> if called more times than there were items
placed in the queue.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.JoinableQueue.join">
<code class="descname">join</code>()<a class="headerlink" href="#multiprocessing.JoinableQueue.join" title="Permalink to this definition">¶</a></dt>
<dd>Block until all items in the queue have been gotten and processed.
The count of unfinished tasks goes up whenever an item is added to the
queue. The count goes down whenever a consumer thread calls
<a class="reference internal" href="#multiprocessing.JoinableQueue.task_done" title="multiprocessing.JoinableQueue.task_done"><code class="xref py py-meth docutils literal notranslate">task_done()</code></a> to indicate that the item was retrieved and all work on
it is complete. When the count of unfinished tasks drops to zero,
<a class="reference internal" href="queue.html#Queue.Queue.join" title="Queue.Queue.join"><code class="xref py py-meth docutils literal notranslate">join()</code></a> unblocks.
</dd></dl>

</dd></dl>

</div>
<div class="section" id="miscellaneous">
<h3>16.6.2.3. Miscellaneous<a class="headerlink" href="#miscellaneous" title="Permalink to this headline">¶</a></h3>
<dl class="function">
<dt id="multiprocessing.active_children">
<code class="descclassname">multiprocessing.</code><code class="descname">active_children</code>()<a class="headerlink" href="#multiprocessing.active_children" title="Permalink to this definition">¶</a></dt>
<dd>Return list of all live children of the current process.
Calling this has the side effect of “joining” any processes which have
already finished.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.cpu_count">
<code class="descclassname">multiprocessing.</code><code class="descname">cpu_count</code>()<a class="headerlink" href="#multiprocessing.cpu_count" title="Permalink to this definition">¶</a></dt>
<dd>Return the number of CPUs in the system. May raise
<a class="reference internal" href="exceptions.html#exceptions.NotImplementedError" title="exceptions.NotImplementedError"><code class="xref py py-exc docutils literal notranslate">NotImplementedError</code></a>.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.current_process">
<code class="descclassname">multiprocessing.</code><code class="descname">current_process</code>()<a class="headerlink" href="#multiprocessing.current_process" title="Permalink to this definition">¶</a></dt>
<dd>Return the <a class="reference internal" href="#multiprocessing.Process" title="multiprocessing.Process"><code class="xref py py-class docutils literal notranslate">Process</code></a> object corresponding to the current process.
An analogue of <a class="reference internal" href="threading.html#threading.current_thread" title="threading.current_thread"><code class="xref py py-func docutils literal notranslate">threading.current_thread()</code></a>.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.freeze_support">
<code class="descclassname">multiprocessing.</code><code class="descname">freeze_support</code>()<a class="headerlink" href="#multiprocessing.freeze_support" title="Permalink to this definition">¶</a></dt>
<dd>Add support for when a program which uses <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> has been
frozen to produce a Windows executable. (Has been tested with py2exe,
PyInstaller and cx_Freeze.)
One needs to call this function straight after the <code class="docutils literal notranslate">if __name__ ==
'__main__'</code> line of the main module. For example:
<div class="highlight-default notranslate"><div class="highlight"><pre>from multiprocessing import Process, freeze_support

def f():
 print &#39;hello world!&#39;

<dl class="function">
<dt id="multiprocessing.set_executable">
<code class="descclassname">multiprocessing.</code><code class="descname">set_executable</code>()<a class="headerlink" href="#multiprocessing.set_executable" title="Permalink to this definition">¶</a></dt>
<dd>Sets the path of the Python interpreter to use when starting a child process.
(By default <a class="reference internal" href="sys.html#sys.executable" title="sys.executable"><code class="xref py py-data docutils literal notranslate">sys.executable</code></a> is used). Embedders will probably need to
do some thing like
<div class="highlight-default notranslate"><div class="highlight"><pre>set_executable(os.path.join(sys.exec_prefix, &#39;pythonw.exe&#39;))
</pre></div>
</div>
before they can create child processes. (Windows only)
</dd></dl>

<div class="admonition note">
Note
<a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> contains no analogues of
<a class="reference internal" href="threading.html#threading.active_count" title="threading.active_count"><code class="xref py py-func docutils literal notranslate">threading.active_count()</code></a>, <a class="reference internal" href="threading.html#threading.enumerate" title="threading.enumerate"><code class="xref py py-func docutils literal notranslate">threading.enumerate()</code></a>,
<a class="reference internal" href="threading.html#threading.settrace" title="threading.settrace"><code class="xref py py-func docutils literal notranslate">threading.settrace()</code></a>, <a class="reference internal" href="threading.html#threading.setprofile" title="threading.setprofile"><code class="xref py py-func docutils literal notranslate">threading.setprofile()</code></a>,
<a class="reference internal" href="threading.html#threading.Timer" title="threading.Timer"><code class="xref py py-class docutils literal notranslate">threading.Timer</code></a>, or <a class="reference internal" href="threading.html#threading.local" title="threading.local"><code class="xref py py-class docutils literal notranslate">threading.local</code></a>.
</div>
</div>
<div class="section" id="connection-objects">
<h3>16.6.2.4. Connection Objects<a class="headerlink" href="#connection-objects" title="Permalink to this headline">¶</a></h3>
Connection objects allow the sending and receiving of picklable objects or
strings. They can be thought of as message oriented connected sockets.
Connection objects are usually created using
<a class="reference internal" href="#multiprocessing.Pipe" title="multiprocessing.Pipe"><code class="xref py py-func docutils literal notranslate">Pipe</code></a> – see also
<a class="reference internal" href="#multiprocessing-listeners-clients">Listeners and Clients</a>.
<dl class="class">
<dt id="Connection">
class <code class="descname">Connection</code><a class="headerlink" href="#Connection" title="Permalink to this definition">¶</a></dt>
<dd><dl class="method">
<dt id="Connection.send">
<code class="descname">send</code>(obj)<a class="headerlink" href="#Connection.send" title="Permalink to this definition">¶</a></dt>
<dd>Send an object to the other end of the connection which should be read
using <a class="reference internal" href="#Connection.recv" title="Connection.recv"><code class="xref py py-meth docutils literal notranslate">recv()</code></a>.
The object must be picklable. Very large pickles (approximately 32 MB+,
though it depends on the OS) may raise a <a class="reference internal" href="exceptions.html#exceptions.ValueError" title="exceptions.ValueError"><code class="xref py py-exc docutils literal notranslate">ValueError</code></a> exception.
</dd></dl>

<dl class="method">
<dt id="Connection.recv">
<code class="descname">recv</code>()<a class="headerlink" href="#Connection.recv" title="Permalink to this definition">¶</a></dt>
<dd>Return an object sent from the other end of the connection using
<a class="reference internal" href="#Connection.send" title="Connection.send"><code class="xref py py-meth docutils literal notranslate">send()</code></a>. Blocks until there is something to receive. Raises
<a class="reference internal" href="exceptions.html#exceptions.EOFError" title="exceptions.EOFError"><code class="xref py py-exc docutils literal notranslate">EOFError</code></a> if there is nothing left to receive
and the other end was closed.
</dd></dl>

<dl class="method">
<dt id="Connection.fileno">
<code class="descname">fileno</code>()<a class="headerlink" href="#Connection.fileno" title="Permalink to this definition">¶</a></dt>
<dd>Return the file descriptor or handle used by the connection.
</dd></dl>

<dl class="method">
<dt id="Connection.close">
<code class="descname">close</code>()<a class="headerlink" href="#Connection.close" title="Permalink to this definition">¶</a></dt>
<dd>Close the connection.
This is called automatically when the connection is garbage collected.
</dd></dl>

<dl class="method">
<dt id="Connection.poll">
<code class="descname">poll</code>([timeout])<a class="headerlink" href="#Connection.poll" title="Permalink to this definition">¶</a></dt>
<dd>Return whether there is any data available to be read.
If timeout is not specified then it will return immediately. If
timeout is a number then this specifies the maximum time in seconds to
block. If timeout is <code class="docutils literal notranslate">None</code> then an infinite timeout is used.
</dd></dl>

<dl class="method">
<dt id="Connection.send_bytes">
<code class="descname">send_bytes</code>(buffer[, offset[, size]])<a class="headerlink" href="#Connection.send_bytes" title="Permalink to this definition">¶</a></dt>
<dd>Send byte data from an object supporting the buffer interface as a
complete message.
If offset is given then data is read from that position in buffer. If
size is given then that many bytes will be read from buffer. Very large
buffers (approximately 32 MB+, though it depends on the OS) may raise a
<a class="reference internal" href="exceptions.html#exceptions.ValueError" title="exceptions.ValueError"><code class="xref py py-exc docutils literal notranslate">ValueError</code></a> exception
</dd></dl>

<dl class="method">
<dt id="Connection.recv_bytes">
<code class="descname">recv_bytes</code>([maxlength])<a class="headerlink" href="#Connection.recv_bytes" title="Permalink to this definition">¶</a></dt>
<dd>Return a complete message of byte data sent from the other end of the
connection as a string. Blocks until there is something to receive.
Raises <a class="reference internal" href="exceptions.html#exceptions.EOFError" title="exceptions.EOFError"><code class="xref py py-exc docutils literal notranslate">EOFError</code></a> if there is nothing left
to receive and the other end has closed.
If maxlength is specified and the message is longer than maxlength
then <a class="reference internal" href="exceptions.html#exceptions.IOError" title="exceptions.IOError"><code class="xref py py-exc docutils literal notranslate">IOError</code></a> is raised and the connection will no longer be
readable.
</dd></dl>

<dl class="method">
<dt id="Connection.recv_bytes_into">
<code class="descname">recv_bytes_into</code>(buffer[, offset])<a class="headerlink" href="#Connection.recv_bytes_into" title="Permalink to this definition">¶</a></dt>
<dd>Read into buffer a complete message of byte data sent from the other end
of the connection and return the number of bytes in the message. Blocks
until there is something to receive. Raises
<a class="reference internal" href="exceptions.html#exceptions.EOFError" title="exceptions.EOFError"><code class="xref py py-exc docutils literal notranslate">EOFError</code></a> if there is nothing left to receive and the other end was
closed.
buffer must be an object satisfying the writable buffer interface. If
offset is given then the message will be written into the buffer from
that position. Offset must be a non-negative integer less than the
length of buffer (in bytes).
If the buffer is too short then a <code class="xref py py-exc docutils literal notranslate">BufferTooShort</code> exception is
raised and the complete message is available as <code class="docutils literal notranslate">e.args[0]</code> where <code class="docutils literal notranslate">e</code>
is the exception instance.
</dd></dl>

</dd></dl>

For example:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from multiprocessing import Pipe
&gt;&gt;&gt; a, b = Pipe()
&gt;&gt;&gt; a.send([1, &#39;hello&#39;, None])
&gt;&gt;&gt; b.recv()
[1, &#39;hello&#39;, None]
&gt;&gt;&gt; b.send_bytes(&#39;thank you&#39;)
&gt;&gt;&gt; a.recv_bytes()
&#39;thank you&#39;
&gt;&gt;&gt; import array
&gt;&gt;&gt; arr1 = array.array(&#39;i&#39;, range(5))
&gt;&gt;&gt; arr2 = array.array(&#39;i&#39;, [0] * 10)
&gt;&gt;&gt; a.send_bytes(arr1)
&gt;&gt;&gt; count = b.recv_bytes_into(arr2)
&gt;&gt;&gt; assert count == len(arr1) * arr1.itemsize
&gt;&gt;&gt; arr2
array(&#39;i&#39;, [0, 1, 2, 3, 4, 0, 0, 0, 0, 0])
</pre></div>
</div>
<div class="admonition warning">
Warning
The <a class="reference internal" href="#Connection.recv" title="Connection.recv"><code class="xref py py-meth docutils literal notranslate">Connection.recv()</code></a> method automatically unpickles the data it
receives, which can be a security risk unless you can trust the process
which sent the message.
Therefore, unless the connection object was produced using <code class="xref py py-func docutils literal notranslate">Pipe()</code> you
should only use the <a class="reference internal" href="#Connection.recv" title="Connection.recv"><code class="xref py py-meth docutils literal notranslate">recv()</code></a> and <a class="reference internal" href="#Connection.send" title="Connection.send"><code class="xref py py-meth docutils literal notranslate">send()</code></a>
methods after performing some sort of authentication. See
<a class="reference internal" href="#multiprocessing-auth-keys">Authentication keys</a>.
</div>
<div class="admonition warning">
Warning
If a process is killed while it is trying to read or write to a pipe then
the data in the pipe is likely to become corrupted, because it may become
impossible to be sure where the message boundaries lie.
</div>
</div>
<div class="section" id="synchronization-primitives">
<h3>16.6.2.5. Synchronization primitives<a class="headerlink" href="#synchronization-primitives" title="Permalink to this headline">¶</a></h3>
Generally synchronization primitives are not as necessary in a multiprocess
program as they are in a multithreaded program. See the documentation for
<a class="reference internal" href="threading.html#module-threading" title="threading: Higher-level threading interface."><code class="xref py py-mod docutils literal notranslate">threading</code></a> module.
Note that one can also create synchronization primitives by using a manager
object – see <a class="reference internal" href="#multiprocessing-managers">Managers</a>.
<dl class="class">
<dt id="multiprocessing.BoundedSemaphore">
class <code class="descclassname">multiprocessing.</code><code class="descname">BoundedSemaphore</code>([value])<a class="headerlink" href="#multiprocessing.BoundedSemaphore" title="Permalink to this definition">¶</a></dt>
<dd>A bounded semaphore object: a close analog of
<a class="reference internal" href="threading.html#threading.BoundedSemaphore" title="threading.BoundedSemaphore"><code class="xref py py-class docutils literal notranslate">threading.BoundedSemaphore</code></a>.
A solitary difference from its close analog exists: its <code class="docutils literal notranslate">acquire</code> method’s
first argument is named block and it supports an optional second argument
timeout, as is consistent with <a class="reference internal" href="#multiprocessing.Lock.acquire" title="multiprocessing.Lock.acquire"><code class="xref py py-meth docutils literal notranslate">Lock.acquire()</code></a>.
<div class="admonition note">
Note
On Mac OS X, this is indistinguishable from <a class="reference internal" href="#multiprocessing.Semaphore" title="multiprocessing.Semaphore"><code class="xref py py-class docutils literal notranslate">Semaphore</code></a> because
<code class="docutils literal notranslate">sem_getvalue()</code> is not implemented on that platform.
</div>
</dd></dl>

<dl class="class">
<dt id="multiprocessing.Condition">
class <code class="descclassname">multiprocessing.</code><code class="descname">Condition</code>([lock])<a class="headerlink" href="#multiprocessing.Condition" title="Permalink to this definition">¶</a></dt>
<dd>A condition variable: a clone of <a class="reference internal" href="threading.html#threading.Condition" title="threading.Condition"><code class="xref py py-class docutils literal notranslate">threading.Condition</code></a>.
If lock is specified then it should be a <a class="reference internal" href="#multiprocessing.Lock" title="multiprocessing.Lock"><code class="xref py py-class docutils literal notranslate">Lock</code></a> or <a class="reference internal" href="#multiprocessing.RLock" title="multiprocessing.RLock"><code class="xref py py-class docutils literal notranslate">RLock</code></a>
object from <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a>.
</dd></dl>

<dl class="class">
<dt id="multiprocessing.Event">
class <code class="descclassname">multiprocessing.</code><code class="descname">Event</code><a class="headerlink" href="#multiprocessing.Event" title="Permalink to this definition">¶</a></dt>
<dd>A clone of <a class="reference internal" href="threading.html#threading.Event" title="threading.Event"><code class="xref py py-class docutils literal notranslate">threading.Event</code></a>.
This method returns the state of the internal semaphore on exit, so it
will always return <code class="docutils literal notranslate">True</code> except if a timeout is given and the operation
times out.
<div class="versionchanged">
Changed in version 2.7: Previously, the method always returned <code class="docutils literal notranslate">None</code>.
</div>
</dd></dl>

<dl class="class">
<dt id="multiprocessing.Lock">
class <code class="descclassname">multiprocessing.</code><code class="descname">Lock</code><a class="headerlink" href="#multiprocessing.Lock" title="Permalink to this definition">¶</a></dt>
<dd>A non-recursive lock object: a close analog of <a class="reference internal" href="threading.html#threading.Lock" title="threading.Lock"><code class="xref py py-class docutils literal notranslate">threading.Lock</code></a>.
Once a process or thread has acquired a lock, subsequent attempts to
acquire it from any process or thread will block until it is released;
any process or thread may release it. The concepts and behaviors of
<a class="reference internal" href="threading.html#threading.Lock" title="threading.Lock"><code class="xref py py-class docutils literal notranslate">threading.Lock</code></a> as it applies to threads are replicated here in
<a class="reference internal" href="#multiprocessing.Lock" title="multiprocessing.Lock"><code class="xref py py-class docutils literal notranslate">multiprocessing.Lock</code></a> as it applies to either processes or threads,
except as noted.
Note that <a class="reference internal" href="#multiprocessing.Lock" title="multiprocessing.Lock"><code class="xref py py-class docutils literal notranslate">Lock</code></a> is actually a factory function which returns an
instance of <code class="docutils literal notranslate">multiprocessing.synchronize.Lock</code> initialized with a
default context.
<a class="reference internal" href="#multiprocessing.Lock" title="multiprocessing.Lock"><code class="xref py py-class docutils literal notranslate">Lock</code></a> supports the <a class="reference internal" href="../glossary.html#term-context-manager">context manager</a> protocol and thus may be
used in <a class="reference internal" href="../reference/compound_stmts.html#with"><code class="xref std std-keyword docutils literal notranslate">with</code></a> statements.
<dl class="method">
<dt id="multiprocessing.Lock.acquire">
<code class="descname">acquire</code>(block=True, timeout=None)<a class="headerlink" href="#multiprocessing.Lock.acquire" title="Permalink to this definition">¶</a></dt>
<dd>Acquire a lock, blocking or non-blocking.
With the block argument set to <code class="docutils literal notranslate">True</code> (the default), the method call
will block until the lock is in an unlocked state, then set it to locked
and return <code class="docutils literal notranslate">True</code>. Note that the name of this first argument differs
from that in <a class="reference internal" href="threading.html#threading.Lock.acquire" title="threading.Lock.acquire"><code class="xref py py-meth docutils literal notranslate">threading.Lock.acquire()</code></a>.
With the block argument set to <code class="docutils literal notranslate">False</code>, the method call does not
block. If the lock is currently in a locked state, return <code class="docutils literal notranslate">False</code>;
otherwise set the lock to a locked state and return <code class="docutils literal notranslate">True</code>.
When invoked with a positive, floating-point value for timeout, block
for at most the number of seconds specified by timeout as long as
the lock can not be acquired. Invocations with a negative value for
timeout are equivalent to a timeout of zero. Invocations with a
timeout value of <code class="docutils literal notranslate">None</code> (the default) set the timeout period to
infinite. The timeout argument has no practical implications if the
block argument is set to <code class="docutils literal notranslate">False</code> and is thus ignored. Returns
<code class="docutils literal notranslate">True</code> if the lock has been acquired or <code class="docutils literal notranslate">False</code> if the timeout period
has elapsed. Note that the timeout argument does not exist in this
method’s analog, <a class="reference internal" href="threading.html#threading.Lock.acquire" title="threading.Lock.acquire"><code class="xref py py-meth docutils literal notranslate">threading.Lock.acquire()</code></a>.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.Lock.release">
<code class="descname">release</code>()<a class="headerlink" href="#multiprocessing.Lock.release" title="Permalink to this definition">¶</a></dt>
<dd>Release a lock. This can be called from any process or thread, not only
the process or thread which originally acquired the lock.
Behavior is the same as in <a class="reference internal" href="threading.html#threading.Lock.release" title="threading.Lock.release"><code class="xref py py-meth docutils literal notranslate">threading.Lock.release()</code></a> except that
when invoked on an unlocked lock, a <a class="reference internal" href="exceptions.html#exceptions.ValueError" title="exceptions.ValueError"><code class="xref py py-exc docutils literal notranslate">ValueError</code></a> is raised.
</dd></dl>

</dd></dl>

<dl class="class">
<dt id="multiprocessing.RLock">
class <code class="descclassname">multiprocessing.</code><code class="descname">RLock</code><a class="headerlink" href="#multiprocessing.RLock" title="Permalink to this definition">¶</a></dt>
<dd>A recursive lock object: a close analog of <a class="reference internal" href="threading.html#threading.RLock" title="threading.RLock"><code class="xref py py-class docutils literal notranslate">threading.RLock</code></a>. A
recursive lock must be released by the process or thread that acquired it.
Once a process or thread has acquired a recursive lock, the same process
or thread may acquire it again without blocking; that process or thread
must release it once for each time it has been acquired.
Note that <a class="reference internal" href="#multiprocessing.RLock" title="multiprocessing.RLock"><code class="xref py py-class docutils literal notranslate">RLock</code></a> is actually a factory function which returns an
instance of <code class="docutils literal notranslate">multiprocessing.synchronize.RLock</code> initialized with a
default context.
<a class="reference internal" href="#multiprocessing.RLock" title="multiprocessing.RLock"><code class="xref py py-class docutils literal notranslate">RLock</code></a> supports the <a class="reference internal" href="../glossary.html#term-context-manager">context manager</a> protocol and thus may be
used in <a class="reference internal" href="../reference/compound_stmts.html#with"><code class="xref std std-keyword docutils literal notranslate">with</code></a> statements.
<dl class="method">
<dt id="multiprocessing.RLock.acquire">
<code class="descname">acquire</code>(block=True, timeout=None)<a class="headerlink" href="#multiprocessing.RLock.acquire" title="Permalink to this definition">¶</a></dt>
<dd>Acquire a lock, blocking or non-blocking.
When invoked with the block argument set to <code class="docutils literal notranslate">True</code>, block until the
lock is in an unlocked state (not owned by any process or thread) unless
the lock is already owned by the current process or thread. The current
process or thread then takes ownership of the lock (if it does not
already have ownership) and the recursion level inside the lock increments
by one, resulting in a return value of <code class="docutils literal notranslate">True</code>. Note that there are
several differences in this first argument’s behavior compared to the
implementation of <a class="reference internal" href="threading.html#threading.RLock.acquire" title="threading.RLock.acquire"><code class="xref py py-meth docutils literal notranslate">threading.RLock.acquire()</code></a>, starting with the name
of the argument itself.
When invoked with the block argument set to <code class="docutils literal notranslate">False</code>, do not block.
If the lock has already been acquired (and thus is owned) by another
process or thread, the current process or thread does not take ownership
and the recursion level within the lock is not changed, resulting in
a return value of <code class="docutils literal notranslate">False</code>. If the lock is in an unlocked state, the
current process or thread takes ownership and the recursion level is
incremented, resulting in a return value of <code class="docutils literal notranslate">True</code>.
Use and behaviors of the timeout argument are the same as in
<a class="reference internal" href="#multiprocessing.Lock.acquire" title="multiprocessing.Lock.acquire"><code class="xref py py-meth docutils literal notranslate">Lock.acquire()</code></a>. Note that the timeout argument does
not exist in this method’s analog, <a class="reference internal" href="threading.html#threading.RLock.acquire" title="threading.RLock.acquire"><code class="xref py py-meth docutils literal notranslate">threading.RLock.acquire()</code></a>.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.RLock.release">
<code class="descname">release</code>()<a class="headerlink" href="#multiprocessing.RLock.release" title="Permalink to this definition">¶</a></dt>
<dd>Release a lock, decrementing the recursion level. If after the
decrement the recursion level is zero, reset the lock to unlocked (not
owned by any process or thread) and if any other processes or threads
are blocked waiting for the lock to become unlocked, allow exactly one
of them to proceed. If after the decrement the recursion level is still
nonzero, the lock remains locked and owned by the calling process or
thread.
Only call this method when the calling process or thread owns the lock.
An <a class="reference internal" href="exceptions.html#exceptions.AssertionError" title="exceptions.AssertionError"><code class="xref py py-exc docutils literal notranslate">AssertionError</code></a> is raised if this method is called by a process
or thread other than the owner or if the lock is in an unlocked (unowned)
state. Note that the type of exception raised in this situation
differs from the implemented behavior in <a class="reference internal" href="threading.html#threading.RLock.release" title="threading.RLock.release"><code class="xref py py-meth docutils literal notranslate">threading.RLock.release()</code></a>.
</dd></dl>

</dd></dl>

<dl class="class">
<dt id="multiprocessing.Semaphore">
class <code class="descclassname">multiprocessing.</code><code class="descname">Semaphore</code>([value])<a class="headerlink" href="#multiprocessing.Semaphore" title="Permalink to this definition">¶</a></dt>
<dd>A semaphore object: a close analog of <a class="reference internal" href="threading.html#threading.Semaphore" title="threading.Semaphore"><code class="xref py py-class docutils literal notranslate">threading.Semaphore</code></a>.
A solitary difference from its close analog exists: its <code class="docutils literal notranslate">acquire</code> method’s
first argument is named block and it supports an optional second argument
timeout, as is consistent with <a class="reference internal" href="#multiprocessing.Lock.acquire" title="multiprocessing.Lock.acquire"><code class="xref py py-meth docutils literal notranslate">Lock.acquire()</code></a>.
</dd></dl>

<div class="admonition note">
Note
The <code class="xref py py-meth docutils literal notranslate">acquire()</code> method of <a class="reference internal" href="#multiprocessing.BoundedSemaphore" title="multiprocessing.BoundedSemaphore"><code class="xref py py-class docutils literal notranslate">BoundedSemaphore</code></a>, <a class="reference internal" href="#multiprocessing.Lock" title="multiprocessing.Lock"><code class="xref py py-class docutils literal notranslate">Lock</code></a>,
<a class="reference internal" href="#multiprocessing.RLock" title="multiprocessing.RLock"><code class="xref py py-class docutils literal notranslate">RLock</code></a> and <a class="reference internal" href="#multiprocessing.Semaphore" title="multiprocessing.Semaphore"><code class="xref py py-class docutils literal notranslate">Semaphore</code></a> has a timeout parameter not supported
by the equivalents in <a class="reference internal" href="threading.html#module-threading" title="threading: Higher-level threading interface."><code class="xref py py-mod docutils literal notranslate">threading</code></a>. The signature is
<code class="docutils literal notranslate">acquire(block=True, timeout=None)</code> with keyword parameters being
acceptable. If block is <code class="docutils literal notranslate">True</code> and timeout is not <code class="docutils literal notranslate">None</code> then it
specifies a timeout in seconds. If block is <code class="docutils literal notranslate">False</code> then timeout is
ignored.
On Mac OS X, <code class="docutils literal notranslate">sem_timedwait</code> is unsupported, so calling <code class="docutils literal notranslate">acquire()</code> with
a timeout will emulate that function’s behavior using a sleeping loop.
</div>
<div class="admonition note">
Note
If the SIGINT signal generated by <kbd class="kbd docutils literal notranslate">Ctrl-C</kbd> arrives while the main thread is
blocked by a call to <code class="xref py py-meth docutils literal notranslate">BoundedSemaphore.acquire()</code>, <a class="reference internal" href="#multiprocessing.Lock.acquire" title="multiprocessing.Lock.acquire"><code class="xref py py-meth docutils literal notranslate">Lock.acquire()</code></a>,
<a class="reference internal" href="#multiprocessing.RLock.acquire" title="multiprocessing.RLock.acquire"><code class="xref py py-meth docutils literal notranslate">RLock.acquire()</code></a>, <code class="xref py py-meth docutils literal notranslate">Semaphore.acquire()</code>, <code class="xref py py-meth docutils literal notranslate">Condition.acquire()</code>
or <code class="xref py py-meth docutils literal notranslate">Condition.wait()</code> then the call will be immediately interrupted and
<a class="reference internal" href="exceptions.html#exceptions.KeyboardInterrupt" title="exceptions.KeyboardInterrupt"><code class="xref py py-exc docutils literal notranslate">KeyboardInterrupt</code></a> will be raised.
This differs from the behaviour of <a class="reference internal" href="threading.html#module-threading" title="threading: Higher-level threading interface."><code class="xref py py-mod docutils literal notranslate">threading</code></a> where SIGINT will be
ignored while the equivalent blocking calls are in progress.
</div>
<div class="admonition note">
Note
Some of this package’s functionality requires a functioning shared semaphore
implementation on the host operating system. Without one, the
<code class="xref py py-mod docutils literal notranslate">multiprocessing.synchronize</code> module will be disabled, and attempts to
import it will result in an <a class="reference internal" href="exceptions.html#exceptions.ImportError" title="exceptions.ImportError"><code class="xref py py-exc docutils literal notranslate">ImportError</code></a>. See
<a class="reference external" href="https://bugs.python.org/issue3770">bpo-3770</a> for additional information.
</div>
</div>
<div class="section" id="shared-ctypes-objects">
<h3>16.6.2.6. Shared <a class="reference internal" href="ctypes.html#module-ctypes" title="ctypes: A foreign function library for Python."><code class="xref py py-mod docutils literal notranslate">ctypes</code></a> Objects<a class="headerlink" href="#shared-ctypes-objects" title="Permalink to this headline">¶</a></h3>
It is possible to create shared objects using shared memory which can be
inherited by child processes.
<dl class="function">
<dt id="multiprocessing.Value">
<code class="descclassname">multiprocessing.</code><code class="descname">Value</code>(typecode_or_type, *args[, lock])<a class="headerlink" href="#multiprocessing.Value" title="Permalink to this definition">¶</a></dt>
<dd>Return a <a class="reference internal" href="ctypes.html#module-ctypes" title="ctypes: A foreign function library for Python."><code class="xref py py-mod docutils literal notranslate">ctypes</code></a> object allocated from shared memory. By default the
return value is actually a synchronized wrapper for the object.
typecode_or_type determines the type of the returned object: it is either a
ctypes type or a one character typecode of the kind used by the <a class="reference internal" href="array.html#module-array" title="array: Space efficient arrays of uniformly typed numeric values."><code class="xref py py-mod docutils literal notranslate">array</code></a>
module. *args is passed on to the constructor for the type.
If lock is <code class="docutils literal notranslate">True</code> (the default) then a new recursive lock
object is created to synchronize access to the value. If lock is
a <a class="reference internal" href="#multiprocessing.Lock" title="multiprocessing.Lock"><code class="xref py py-class docutils literal notranslate">Lock</code></a> or <a class="reference internal" href="#multiprocessing.RLock" title="multiprocessing.RLock"><code class="xref py py-class docutils literal notranslate">RLock</code></a> object then that will be used to
synchronize access to the value. If lock is <code class="docutils literal notranslate">False</code> then
access to the returned object will not be automatically protected
by a lock, so it will not necessarily be “process-safe”.
Operations like <code class="docutils literal notranslate">+=</code> which involve a read and write are not
atomic. So if, for instance, you want to atomically increment a
shared value it is insufficient to just do
<div class="highlight-default notranslate"><div class="highlight"><pre>counter.value += 1
</pre></div>
</div>
Assuming the associated lock is recursive (which it is by default)
you can instead do
<div class="highlight-default notranslate"><div class="highlight"><pre>with counter.get_lock():
 counter.value += 1
</pre></div>
</div>
Note that lock is a keyword-only argument.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.Array">
<code class="descclassname">multiprocessing.</code><code class="descname">Array</code>(typecode_or_type, size_or_initializer, *, lock=True)<a class="headerlink" href="#multiprocessing.Array" title="Permalink to this definition">¶</a></dt>
<dd>Return a ctypes array allocated from shared memory. By default the return
value is actually a synchronized wrapper for the array.
typecode_or_type determines the type of the elements of the returned array:
it is either a ctypes type or a one character typecode of the kind used by
the <a class="reference internal" href="array.html#module-array" title="array: Space efficient arrays of uniformly typed numeric values."><code class="xref py py-mod docutils literal notranslate">array</code></a> module. If size_or_initializer is an integer, then it
determines the length of the array, and the array will be initially zeroed.
Otherwise, size_or_initializer is a sequence which is used to initialize
the array and whose length determines the length of the array.
If lock is <code class="docutils literal notranslate">True</code> (the default) then a new lock object is created to
synchronize access to the value. If lock is a <a class="reference internal" href="#multiprocessing.Lock" title="multiprocessing.Lock"><code class="xref py py-class docutils literal notranslate">Lock</code></a> or
<a class="reference internal" href="#multiprocessing.RLock" title="multiprocessing.RLock"><code class="xref py py-class docutils literal notranslate">RLock</code></a> object then that will be used to synchronize access to the
value. If lock is <code class="docutils literal notranslate">False</code> then access to the returned object will not be
automatically protected by a lock, so it will not necessarily be
“process-safe”.
Note that lock is a keyword only argument.
Note that an array of <a class="reference internal" href="ctypes.html#ctypes.c_char" title="ctypes.c_char"><code class="xref py py-data docutils literal notranslate">ctypes.c_char</code></a> has value and raw
attributes which allow one to use it to store and retrieve strings.
</dd></dl>

<div class="section" id="module-multiprocessing.sharedctypes">
<h4>16.6.2.6.1. The <a class="reference internal" href="#module-multiprocessing.sharedctypes" title="multiprocessing.sharedctypes: Allocate ctypes objects from shared memory."><code class="xref py py-mod docutils literal notranslate">multiprocessing.sharedctypes</code></a> module<a class="headerlink" href="#module-multiprocessing.sharedctypes" title="Permalink to this headline">¶</a></h4>
The <a class="reference internal" href="#module-multiprocessing.sharedctypes" title="multiprocessing.sharedctypes: Allocate ctypes objects from shared memory."><code class="xref py py-mod docutils literal notranslate">multiprocessing.sharedctypes</code></a> module provides functions for allocating
<a class="reference internal" href="ctypes.html#module-ctypes" title="ctypes: A foreign function library for Python."><code class="xref py py-mod docutils literal notranslate">ctypes</code></a> objects from shared memory which can be inherited by child
processes.
<div class="admonition note">
Note
Although it is possible to store a pointer in shared memory remember that
this will refer to a location in the address space of a specific process.
However, the pointer is quite likely to be invalid in the context of a second
process and trying to dereference the pointer from the second process may
cause a crash.
</div>
<dl class="function">
<dt id="multiprocessing.sharedctypes.RawArray">
<code class="descclassname">multiprocessing.sharedctypes.</code><code class="descname">RawArray</code>(typecode_or_type, size_or_initializer)<a class="headerlink" href="#multiprocessing.sharedctypes.RawArray" title="Permalink to this definition">¶</a></dt>
<dd>Return a ctypes array allocated from shared memory.
typecode_or_type determines the type of the elements of the returned array:
it is either a ctypes type or a one character typecode of the kind used by
the <a class="reference internal" href="array.html#module-array" title="array: Space efficient arrays of uniformly typed numeric values."><code class="xref py py-mod docutils literal notranslate">array</code></a> module. If size_or_initializer is an integer then it
determines the length of the array, and the array will be initially zeroed.
Otherwise size_or_initializer is a sequence which is used to initialize the
array and whose length determines the length of the array.
Note that setting and getting an element is potentially non-atomic – use
<a class="reference internal" href="#multiprocessing.sharedctypes.Array" title="multiprocessing.sharedctypes.Array"><code class="xref py py-func docutils literal notranslate">Array()</code></a> instead to make sure that access is automatically synchronized
using a lock.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.sharedctypes.RawValue">
<code class="descclassname">multiprocessing.sharedctypes.</code><code class="descname">RawValue</code>(typecode_or_type, *args)<a class="headerlink" href="#multiprocessing.sharedctypes.RawValue" title="Permalink to this definition">¶</a></dt>
<dd>Return a ctypes object allocated from shared memory.
typecode_or_type determines the type of the returned object: it is either a
ctypes type or a one character typecode of the kind used by the <a class="reference internal" href="array.html#module-array" title="array: Space efficient arrays of uniformly typed numeric values."><code class="xref py py-mod docutils literal notranslate">array</code></a>
module. *args is passed on to the constructor for the type.
Note that setting and getting the value is potentially non-atomic – use
<a class="reference internal" href="#multiprocessing.sharedctypes.Value" title="multiprocessing.sharedctypes.Value"><code class="xref py py-func docutils literal notranslate">Value()</code></a> instead to make sure that access is automatically synchronized
using a lock.
Note that an array of <a class="reference internal" href="ctypes.html#ctypes.c_char" title="ctypes.c_char"><code class="xref py py-data docutils literal notranslate">ctypes.c_char</code></a> has <code class="docutils literal notranslate">value</code> and <code class="docutils literal notranslate">raw</code>
attributes which allow one to use it to store and retrieve strings – see
documentation for <a class="reference internal" href="ctypes.html#module-ctypes" title="ctypes: A foreign function library for Python."><code class="xref py py-mod docutils literal notranslate">ctypes</code></a>.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.sharedctypes.Array">
<code class="descclassname">multiprocessing.sharedctypes.</code><code class="descname">Array</code>(typecode_or_type, size_or_initializer, *args[, lock])<a class="headerlink" href="#multiprocessing.sharedctypes.Array" title="Permalink to this definition">¶</a></dt>
<dd>The same as <a class="reference internal" href="#multiprocessing.sharedctypes.RawArray" title="multiprocessing.sharedctypes.RawArray"><code class="xref py py-func docutils literal notranslate">RawArray()</code></a> except that depending on the value of lock a
process-safe synchronization wrapper may be returned instead of a raw ctypes
array.
If lock is <code class="docutils literal notranslate">True</code> (the default) then a new lock object is created to
synchronize access to the value. If lock is a
<a class="reference internal" href="#multiprocessing.Lock" title="multiprocessing.Lock"><code class="xref py py-class docutils literal notranslate">Lock</code></a> or <a class="reference internal" href="#multiprocessing.RLock" title="multiprocessing.RLock"><code class="xref py py-class docutils literal notranslate">RLock</code></a> object
then that will be used to synchronize access to the
value. If lock is <code class="docutils literal notranslate">False</code> then access to the returned object will not be
automatically protected by a lock, so it will not necessarily be
“process-safe”.
Note that lock is a keyword-only argument.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.sharedctypes.Value">
<code class="descclassname">multiprocessing.sharedctypes.</code><code class="descname">Value</code>(typecode_or_type, *args[, lock])<a class="headerlink" href="#multiprocessing.sharedctypes.Value" title="Permalink to this definition">¶</a></dt>
<dd>The same as <a class="reference internal" href="#multiprocessing.sharedctypes.RawValue" title="multiprocessing.sharedctypes.RawValue"><code class="xref py py-func docutils literal notranslate">RawValue()</code></a> except that depending on the value of lock a
process-safe synchronization wrapper may be returned instead of a raw ctypes
object.
If lock is <code class="docutils literal notranslate">True</code> (the default) then a new lock object is created to
synchronize access to the value. If lock is a <a class="reference internal" href="#multiprocessing.Lock" title="multiprocessing.Lock"><code class="xref py py-class docutils literal notranslate">Lock</code></a> or
<a class="reference internal" href="#multiprocessing.RLock" title="multiprocessing.RLock"><code class="xref py py-class docutils literal notranslate">RLock</code></a> object then that will be used to synchronize access to the
value. If lock is <code class="docutils literal notranslate">False</code> then access to the returned object will not be
automatically protected by a lock, so it will not necessarily be
“process-safe”.
Note that lock is a keyword-only argument.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.sharedctypes.copy">
<code class="descclassname">multiprocessing.sharedctypes.</code><code class="descname">copy</code>(obj)<a class="headerlink" href="#multiprocessing.sharedctypes.copy" title="Permalink to this definition">¶</a></dt>
<dd>Return a ctypes object allocated from shared memory which is a copy of the
ctypes object obj.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.sharedctypes.synchronized">
<code class="descclassname">multiprocessing.sharedctypes.</code><code class="descname">synchronized</code>(obj[, lock])<a class="headerlink" href="#multiprocessing.sharedctypes.synchronized" title="Permalink to this definition">¶</a></dt>
<dd>Return a process-safe wrapper object for a ctypes object which uses lock to
synchronize access. If lock is <code class="docutils literal notranslate">None</code> (the default) then a
<a class="reference internal" href="#multiprocessing.RLock" title="multiprocessing.RLock"><code class="xref py py-class docutils literal notranslate">multiprocessing.RLock</code></a> object is created automatically.
A synchronized wrapper will have two methods in addition to those of the
object it wraps: <code class="xref py py-meth docutils literal notranslate">get_obj()</code> returns the wrapped object and
<code class="xref py py-meth docutils literal notranslate">get_lock()</code> returns the lock object used for synchronization.
Note that accessing the ctypes object through the wrapper can be a lot slower
than accessing the raw ctypes object.
</dd></dl>

The table below compares the syntax for creating shared ctypes objects from
shared memory with the normal ctypes syntax. (In the table <code class="docutils literal notranslate">MyStruct</code> is some
subclass of <a class="reference internal" href="ctypes.html#ctypes.Structure" title="ctypes.Structure"><code class="xref py py-class docutils literal notranslate">ctypes.Structure</code></a>.)
<table border="1" class="docutils">
<colgroup>
<col width="27%" />
<col width="36%" />
<col width="37%" />
</colgroup>
<thead valign="bottom">
<tr class="row-odd"><th class="head">ctypes</th>
<th class="head">sharedctypes using type</th>
<th class="head">sharedctypes using typecode</th>
</tr>
</thead>
<tbody valign="top">
<tr class="row-even"><td>c_double(2.4)</td>
<td>RawValue(c_double, 2.4)</td>
<td>RawValue(‘d’, 2.4)</td>
</tr>
<tr class="row-odd"><td>MyStruct(4, 6)</td>
<td>RawValue(MyStruct, 4, 6)</td>
<td>&#160;</td>
</tr>
<tr class="row-even"><td>(c_short * 7)()</td>
<td>RawArray(c_short, 7)</td>
<td>RawArray(‘h’, 7)</td>
</tr>
<tr class="row-odd"><td>(c_int * 3)(9, 2, 8)</td>
<td>RawArray(c_int, (9, 2, 8))</td>
<td>RawArray(‘i’, (9, 2, 8))</td>
</tr>
</tbody>
</table>
Below is an example where a number of ctypes objects are modified by a child
process:
<div class="highlight-default notranslate"><div class="highlight"><pre>from multiprocessing import Process, Lock
from multiprocessing.sharedctypes import Value, Array
from ctypes import Structure, c_double

class Point(Structure):
 _fields_ = [(&#39;x&#39;, c_double), (&#39;y&#39;, c_double)]

def modify(n, x, s, A):
 n.value **= 2
 x.value **= 2
 s.value = s.value.upper()
 for a in A:
 a.x **= 2
 a.y **= 2

n = Value(&#39;i&#39;, 7)
 x = Value(c_double, 1.0/3.0, lock=False)
 s = Array(&#39;c&#39;, &#39;hello world&#39;, lock=lock)
 A = Array(Point, [(1.875,-6.25), (-5.75,2.0), (2.375,9.5)], lock=lock)

p = Process(target=modify, args=(n, x, s, A))
 p.start()
 p.join()

print n.value
 print x.value
 print s.value
 print [(a.x, a.y) for a in A]
</pre></div>
</div>
The results printed are
<div class="highlight-none notranslate"><div class="highlight"><pre>49
0.1111111111111111
HELLO WORLD
[(3.515625, 39.0625), (33.0625, 4.0), (5.640625, 90.25)]
</pre></div>
</div>
</div>
</div>
<div class="section" id="managers">
<h3>16.6.2.7. Managers<a class="headerlink" href="#managers" title="Permalink to this headline">¶</a></h3>
Managers provide a way to create data which can be shared between different
processes. A manager object controls a server process which manages shared
objects. Other processes can access the shared objects by using proxies.
<dl class="function">
<dt id="multiprocessing.sharedctypes.multiprocessing.Manager">
<code class="descclassname">multiprocessing.</code><code class="descname">Manager</code>()<a class="headerlink" href="#multiprocessing.sharedctypes.multiprocessing.Manager" title="Permalink to this definition">¶</a></dt>
<dd>Returns a started <a class="reference internal" href="#multiprocessing.managers.SyncManager" title="multiprocessing.managers.SyncManager"><code class="xref py py-class docutils literal notranslate">SyncManager</code></a> object which
can be used for sharing objects between processes. The returned manager
object corresponds to a spawned child process and has methods which will
create shared objects and return corresponding proxies.
</dd></dl>

Manager processes will be shutdown as soon as they are garbage collected or
their parent process exits. The manager classes are defined in the
<a class="reference internal" href="#module-multiprocessing.managers" title="multiprocessing.managers: Share data between process with shared objects."><code class="xref py py-mod docutils literal notranslate">multiprocessing.managers</code></a> module:
<dl class="class">
<dt id="multiprocessing.managers.BaseManager">
class <code class="descclassname">multiprocessing.managers.</code><code class="descname">BaseManager</code>([address[, authkey]])<a class="headerlink" href="#multiprocessing.managers.BaseManager" title="Permalink to this definition">¶</a></dt>
<dd>Create a BaseManager object.
Once created one should call <a class="reference internal" href="#multiprocessing.managers.BaseManager.start" title="multiprocessing.managers.BaseManager.start"><code class="xref py py-meth docutils literal notranslate">start()</code></a> or <code class="docutils literal notranslate">get_server().serve_forever()</code> to ensure
that the manager object refers to a started manager process.
address is the address on which the manager process listens for new
connections. If address is <code class="docutils literal notranslate">None</code> then an arbitrary one is chosen.
authkey is the authentication key which will be used to check the validity
of incoming connections to the server process. If authkey is <code class="docutils literal notranslate">None</code> then
<code class="docutils literal notranslate">current_process().authkey</code>. Otherwise authkey is used and it
must be a string.
<dl class="method">
<dt id="multiprocessing.managers.BaseManager.start">
<code class="descname">start</code>([initializer[, initargs]])<a class="headerlink" href="#multiprocessing.managers.BaseManager.start" title="Permalink to this definition">¶</a></dt>
<dd>Start a subprocess to start the manager. If initializer is not <code class="docutils literal notranslate">None</code>
then the subprocess will call <code class="docutils literal notranslate">initializer(*initargs)</code> when it starts.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.BaseManager.get_server">
<code class="descname">get_server</code>()<a class="headerlink" href="#multiprocessing.managers.BaseManager.get_server" title="Permalink to this definition">¶</a></dt>
<dd>Returns a <code class="xref py py-class docutils literal notranslate">Server</code> object which represents the actual server under
the control of the Manager. The <code class="xref py py-class docutils literal notranslate">Server</code> object supports the
<code class="xref py py-meth docutils literal notranslate">serve_forever()</code> method:
<div class="highlight-python notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from multiprocessing.managers import BaseManager
&gt;&gt;&gt; manager = BaseManager(address=(&#39;&#39;, 50000), authkey=&#39;abc&#39;)
&gt;&gt;&gt; server = manager.get_server()
&gt;&gt;&gt; server.serve_forever()
</pre></div>
</div>
<code class="xref py py-class docutils literal notranslate">Server</code> additionally has an <a class="reference internal" href="#multiprocessing.managers.BaseManager.address" title="multiprocessing.managers.BaseManager.address"><code class="xref py py-attr docutils literal notranslate">address</code></a> attribute.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.BaseManager.connect">
<code class="descname">connect</code>()<a class="headerlink" href="#multiprocessing.managers.BaseManager.connect" title="Permalink to this definition">¶</a></dt>
<dd>Connect a local manager object to a remote manager process:
<div class="highlight-python notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from multiprocessing.managers import BaseManager
&gt;&gt;&gt; m = BaseManager(address=(&#39;127.0.0.1&#39;, 5000), authkey=&#39;abc&#39;)
&gt;&gt;&gt; m.connect()
</pre></div>
</div>
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.BaseManager.shutdown">
<code class="descname">shutdown</code>()<a class="headerlink" href="#multiprocessing.managers.BaseManager.shutdown" title="Permalink to this definition">¶</a></dt>
<dd>Stop the process used by the manager. This is only available if
<a class="reference internal" href="#multiprocessing.managers.BaseManager.start" title="multiprocessing.managers.BaseManager.start"><code class="xref py py-meth docutils literal notranslate">start()</code></a> has been used to start the server process.
This can be called multiple times.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.BaseManager.register">
<code class="descname">register</code>(typeid[, callable[, proxytype[, exposed[, method_to_typeid[, create_method]]]]])<a class="headerlink" href="#multiprocessing.managers.BaseManager.register" title="Permalink to this definition">¶</a></dt>
<dd>A classmethod which can be used for registering a type or callable with
the manager class.
typeid is a “type identifier” which is used to identify a particular
type of shared object. This must be a string.
callable is a callable used for creating objects for this type
identifier. If a manager instance will be created using the
<code class="xref py py-meth docutils literal notranslate">from_address()</code> classmethod or if the create_method argument is
<code class="docutils literal notranslate">False</code> then this can be left as <code class="docutils literal notranslate">None</code>.
proxytype is a subclass of <a class="reference internal" href="#multiprocessing.managers.BaseProxy" title="multiprocessing.managers.BaseProxy"><code class="xref py py-class docutils literal notranslate">BaseProxy</code></a> which is used to create
proxies for shared objects with this typeid. If <code class="docutils literal notranslate">None</code> then a proxy
class is created automatically.
exposed is used to specify a sequence of method names which proxies for
this typeid should be allowed to access using
<a class="reference internal" href="#multiprocessing.managers.BaseProxy._callmethod" title="multiprocessing.managers.BaseProxy._callmethod"><code class="xref py py-meth docutils literal notranslate">BaseProxy._callmethod()</code></a>. (If exposed is <code class="docutils literal notranslate">None</code> then
<code class="xref py py-attr docutils literal notranslate">proxytype._exposed_</code> is used instead if it exists.) In the case
where no exposed list is specified, all “public methods” of the shared
object will be accessible. (Here a “public method” means any attribute
which has a <a class="reference internal" href="../reference/datamodel.html#object.__call__" title="object.__call__"><code class="xref py py-meth docutils literal notranslate">__call__()</code></a> method and whose name does not begin
with <code class="docutils literal notranslate">'_'</code>.)
method_to_typeid is a mapping used to specify the return type of those
exposed methods which should return a proxy. It maps method names to
typeid strings. (If method_to_typeid is <code class="docutils literal notranslate">None</code> then
<code class="xref py py-attr docutils literal notranslate">proxytype._method_to_typeid_</code> is used instead if it exists.) If a
method’s name is not a key of this mapping or if the mapping is <code class="docutils literal notranslate">None</code>
then the object returned by the method will be copied by value.
create_method determines whether a method should be created with name
typeid which can be used to tell the server process to create a new
shared object and return a proxy for it. By default it is <code class="docutils literal notranslate">True</code>.
</dd></dl>

<a class="reference internal" href="#multiprocessing.managers.BaseManager" title="multiprocessing.managers.BaseManager"><code class="xref py py-class docutils literal notranslate">BaseManager</code></a> instances also have one read-only property:
<dl class="attribute">
<dt id="multiprocessing.managers.BaseManager.address">
<code class="descname">address</code><a class="headerlink" href="#multiprocessing.managers.BaseManager.address" title="Permalink to this definition">¶</a></dt>
<dd>The address used by the manager.
</dd></dl>

</dd></dl>

<dl class="class">
<dt id="multiprocessing.managers.SyncManager">
class <code class="descclassname">multiprocessing.managers.</code><code class="descname">SyncManager</code><a class="headerlink" href="#multiprocessing.managers.SyncManager" title="Permalink to this definition">¶</a></dt>
<dd>A subclass of <a class="reference internal" href="#multiprocessing.managers.BaseManager" title="multiprocessing.managers.BaseManager"><code class="xref py py-class docutils literal notranslate">BaseManager</code></a> which can be used for the synchronization
of processes. Objects of this type are returned by
<code class="xref py py-func docutils literal notranslate">multiprocessing.Manager()</code>.
It also supports creation of shared lists and dictionaries.
<dl class="method">
<dt id="multiprocessing.managers.SyncManager.BoundedSemaphore">
<code class="descname">BoundedSemaphore</code>([value])<a class="headerlink" href="#multiprocessing.managers.SyncManager.BoundedSemaphore" title="Permalink to this definition">¶</a></dt>
<dd>Create a shared <a class="reference internal" href="threading.html#threading.BoundedSemaphore" title="threading.BoundedSemaphore"><code class="xref py py-class docutils literal notranslate">threading.BoundedSemaphore</code></a> object and return a
proxy for it.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.SyncManager.Condition">
<code class="descname">Condition</code>([lock])<a class="headerlink" href="#multiprocessing.managers.SyncManager.Condition" title="Permalink to this definition">¶</a></dt>
<dd>Create a shared <a class="reference internal" href="threading.html#threading.Condition" title="threading.Condition"><code class="xref py py-class docutils literal notranslate">threading.Condition</code></a> object and return a proxy for
it.
If lock is supplied then it should be a proxy for a
<a class="reference internal" href="threading.html#threading.Lock" title="threading.Lock"><code class="xref py py-class docutils literal notranslate">threading.Lock</code></a> or <a class="reference internal" href="threading.html#threading.RLock" title="threading.RLock"><code class="xref py py-class docutils literal notranslate">threading.RLock</code></a> object.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.SyncManager.Event">
<code class="descname">Event</code>()<a class="headerlink" href="#multiprocessing.managers.SyncManager.Event" title="Permalink to this definition">¶</a></dt>
<dd>Create a shared <a class="reference internal" href="threading.html#threading.Event" title="threading.Event"><code class="xref py py-class docutils literal notranslate">threading.Event</code></a> object and return a proxy for it.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.SyncManager.Lock">
<code class="descname">Lock</code>()<a class="headerlink" href="#multiprocessing.managers.SyncManager.Lock" title="Permalink to this definition">¶</a></dt>
<dd>Create a shared <a class="reference internal" href="threading.html#threading.Lock" title="threading.Lock"><code class="xref py py-class docutils literal notranslate">threading.Lock</code></a> object and return a proxy for it.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.SyncManager.Namespace">
<code class="descname">Namespace</code>()<a class="headerlink" href="#multiprocessing.managers.SyncManager.Namespace" title="Permalink to this definition">¶</a></dt>
<dd>Create a shared <a class="reference internal" href="#multiprocessing.managers.Namespace" title="multiprocessing.managers.Namespace"><code class="xref py py-class docutils literal notranslate">Namespace</code></a> object and return a proxy for it.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.SyncManager.Queue">
<code class="descname">Queue</code>([maxsize])<a class="headerlink" href="#multiprocessing.managers.SyncManager.Queue" title="Permalink to this definition">¶</a></dt>
<dd>Create a shared <a class="reference internal" href="queue.html#Queue.Queue" title="Queue.Queue"><code class="xref py py-class docutils literal notranslate">Queue.Queue</code></a> object and return a proxy for it.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.SyncManager.RLock">
<code class="descname">RLock</code>()<a class="headerlink" href="#multiprocessing.managers.SyncManager.RLock" title="Permalink to this definition">¶</a></dt>
<dd>Create a shared <a class="reference internal" href="threading.html#threading.RLock" title="threading.RLock"><code class="xref py py-class docutils literal notranslate">threading.RLock</code></a> object and return a proxy for it.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.SyncManager.Semaphore">
<code class="descname">Semaphore</code>([value])<a class="headerlink" href="#multiprocessing.managers.SyncManager.Semaphore" title="Permalink to this definition">¶</a></dt>
<dd>Create a shared <a class="reference internal" href="threading.html#threading.Semaphore" title="threading.Semaphore"><code class="xref py py-class docutils literal notranslate">threading.Semaphore</code></a> object and return a proxy for
it.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.SyncManager.Array">
<code class="descname">Array</code>(typecode, sequence)<a class="headerlink" href="#multiprocessing.managers.SyncManager.Array" title="Permalink to this definition">¶</a></dt>
<dd>Create an array and return a proxy for it.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.SyncManager.Value">
<code class="descname">Value</code>(typecode, value)<a class="headerlink" href="#multiprocessing.managers.SyncManager.Value" title="Permalink to this definition">¶</a></dt>
<dd>Create an object with a writable <code class="docutils literal notranslate">value</code> attribute and return a proxy
for it.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.SyncManager.dict">
<code class="descname">dict</code>()<a class="headerlink" href="#multiprocessing.managers.SyncManager.dict" title="Permalink to this definition">¶</a></dt>
<dt>
<code class="descname">dict</code>(mapping)</dt>
<dt>
<code class="descname">dict</code>(sequence)</dt>
<dd>Create a shared <code class="docutils literal notranslate">dict</code> object and return a proxy for it.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.SyncManager.list">
<code class="descname">list</code>()<a class="headerlink" href="#multiprocessing.managers.SyncManager.list" title="Permalink to this definition">¶</a></dt>
<dt>
<code class="descname">list</code>(sequence)</dt>
<dd>Create a shared <code class="docutils literal notranslate">list</code> object and return a proxy for it.
</dd></dl>

<div class="admonition note">
Note
Modifications to mutable values or items in dict and list proxies will not
be propagated through the manager, because the proxy has no way of knowing
when its values or items are modified. To modify such an item, you can
re-assign the modified object to the container proxy:
<div class="last highlight-python notranslate"><div class="highlight"><pre># create a list proxy and append a mutable object (a dictionary)
lproxy = manager.list()
lproxy.append({})
# now mutate the dictionary
d = lproxy[0]
d[&#39;a&#39;] = 1
d[&#39;b&#39;] = 2
# at this point, the changes to d are not yet synced, but by
# reassigning the dictionary, the proxy is notified of the change
lproxy[0] = d
</pre></div>
</div>
</div>
</dd></dl>

<dl class="class">
<dt id="multiprocessing.managers.Namespace">
class <code class="descclassname">multiprocessing.managers.</code><code class="descname">Namespace</code><a class="headerlink" href="#multiprocessing.managers.Namespace" title="Permalink to this definition">¶</a></dt>
<dd>A type that can register with <a class="reference internal" href="#multiprocessing.managers.SyncManager" title="multiprocessing.managers.SyncManager"><code class="xref py py-class docutils literal notranslate">SyncManager</code></a>.
A namespace object has no public methods, but does have writable attributes.
Its representation shows the values of its attributes.
However, when using a proxy for a namespace object, an attribute beginning with
<code class="docutils literal notranslate">'_'</code> will be an attribute of the proxy and not an attribute of the referent:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; manager = multiprocessing.Manager()
&gt;&gt;&gt; Global = manager.Namespace()
&gt;&gt;&gt; Global.x = 10
&gt;&gt;&gt; Global.y = &#39;hello&#39;
&gt;&gt;&gt; Global._z = 12.3 # this is an attribute of the proxy
&gt;&gt;&gt; print Global
Namespace(x=10, y=&#39;hello&#39;)
</pre></div>
</div>
</dd></dl>

<div class="section" id="customized-managers">
<h4>16.6.2.7.1. Customized managers<a class="headerlink" href="#customized-managers" title="Permalink to this headline">¶</a></h4>
To create one’s own manager, one creates a subclass of <a class="reference internal" href="#multiprocessing.managers.BaseManager" title="multiprocessing.managers.BaseManager"><code class="xref py py-class docutils literal notranslate">BaseManager</code></a> and
uses the <a class="reference internal" href="#multiprocessing.managers.BaseManager.register" title="multiprocessing.managers.BaseManager.register"><code class="xref py py-meth docutils literal notranslate">register()</code></a> classmethod to register new types or
callables with the manager class. For example:
<div class="highlight-python notranslate"><div class="highlight"><pre>from multiprocessing.managers import BaseManager

class MathsClass(object):
 def add(self, x, y):
 return x + y
 def mul(self, x, y):
 return x * y

class MyManager(BaseManager):
 pass

MyManager.register(&#39;Maths&#39;, MathsClass)

if __name__ == &#39;__main__&#39;:
 manager = MyManager()
 manager.start()
 maths = manager.Maths()
 print maths.add(4, 3) # prints 7
 print maths.mul(7, 8) # prints 56
</pre></div>
</div>
</div>
<div class="section" id="using-a-remote-manager">
<h4>16.6.2.7.2. Using a remote manager<a class="headerlink" href="#using-a-remote-manager" title="Permalink to this headline">¶</a></h4>
It is possible to run a manager server on one machine and have clients use it
from other machines (assuming that the firewalls involved allow it).
Running the following commands creates a server for a single shared queue which
remote clients can access:
<div class="highlight-python notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from multiprocessing.managers import BaseManager
&gt;&gt;&gt; import Queue
&gt;&gt;&gt; queue = Queue.Queue()
&gt;&gt;&gt; class QueueManager(BaseManager): pass
&gt;&gt;&gt; QueueManager.register(&#39;get_queue&#39;, callable=lambda:queue)
&gt;&gt;&gt; m = QueueManager(address=(&#39;&#39;, 50000), authkey=&#39;abracadabra&#39;)
&gt;&gt;&gt; s = m.get_server()
&gt;&gt;&gt; s.serve_forever()
</pre></div>
</div>
One client can access the server as follows:
<div class="highlight-python notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from multiprocessing.managers import BaseManager
&gt;&gt;&gt; class QueueManager(BaseManager): pass
&gt;&gt;&gt; QueueManager.register(&#39;get_queue&#39;)
&gt;&gt;&gt; m = QueueManager(address=(&#39;foo.bar.org&#39;, 50000), authkey=&#39;abracadabra&#39;)
&gt;&gt;&gt; m.connect()
&gt;&gt;&gt; queue = m.get_queue()
&gt;&gt;&gt; queue.put(&#39;hello&#39;)
</pre></div>
</div>
Another client can also use it:
<div class="highlight-python notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from multiprocessing.managers import BaseManager
&gt;&gt;&gt; class QueueManager(BaseManager): pass
&gt;&gt;&gt; QueueManager.register(&#39;get_queue&#39;)
&gt;&gt;&gt; m = QueueManager(address=(&#39;foo.bar.org&#39;, 50000), authkey=&#39;abracadabra&#39;)
&gt;&gt;&gt; m.connect()
&gt;&gt;&gt; queue = m.get_queue()
&gt;&gt;&gt; queue.get()
&#39;hello&#39;
</pre></div>
</div>
Local processes can also access that queue, using the code from above on the
client to access it remotely:
<div class="highlight-python notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from multiprocessing import Process, Queue
&gt;&gt;&gt; from multiprocessing.managers import BaseManager
&gt;&gt;&gt; class Worker(Process):
... def __init__(self, q):
... self.q = q
... super(Worker, self).__init__()
... def run(self):
... self.q.put(&#39;local hello&#39;)
...
&gt;&gt;&gt; queue = Queue()
&gt;&gt;&gt; w = Worker(queue)
&gt;&gt;&gt; w.start()
&gt;&gt;&gt; class QueueManager(BaseManager): pass
...
&gt;&gt;&gt; QueueManager.register(&#39;get_queue&#39;, callable=lambda: queue)
&gt;&gt;&gt; m = QueueManager(address=(&#39;&#39;, 50000), authkey=&#39;abracadabra&#39;)
&gt;&gt;&gt; s = m.get_server()
&gt;&gt;&gt; s.serve_forever()
</pre></div>
</div>
</div>
</div>
<div class="section" id="proxy-objects">
<h3>16.6.2.8. Proxy Objects<a class="headerlink" href="#proxy-objects" title="Permalink to this headline">¶</a></h3>
A proxy is an object which refers to a shared object which lives (presumably)
in a different process. The shared object is said to be the referent of the
proxy. Multiple proxy objects may have the same referent.
A proxy object has methods which invoke corresponding methods of its referent
(although not every method of the referent will necessarily be available through
the proxy). A proxy can usually be used in most of the same ways that its
referent can:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from multiprocessing import Manager
&gt;&gt;&gt; manager = Manager()
&gt;&gt;&gt; l = manager.list([i*i for i in range(10)])
&gt;&gt;&gt; print l
[0, 1, 4, 9, 16, 25, 36, 49, 64, 81]
&gt;&gt;&gt; print repr(l)
&lt;ListProxy object, typeid &#39;list&#39; at 0x...&gt;
&gt;&gt;&gt; l[4]
16
&gt;&gt;&gt; l[2:5]
[4, 9, 16]
</pre></div>
</div>
Notice that applying <a class="reference internal" href="functions.html#str" title="str"><code class="xref py py-func docutils literal notranslate">str()</code></a> to a proxy will return the representation of
the referent, whereas applying <a class="reference internal" href="functions.html#repr" title="repr"><code class="xref py py-func docutils literal notranslate">repr()</code></a> will return the representation of
the proxy.
An important feature of proxy objects is that they are picklable so they can be
passed between processes. Note, however, that if a proxy is sent to the
corresponding manager’s process then unpickling it will produce the referent
itself. This means, for example, that one shared object can contain a second:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; a = manager.list()
&gt;&gt;&gt; b = manager.list()
&gt;&gt;&gt; a.append(b) # referent of a now contains referent of b
&gt;&gt;&gt; print a, b
[[]] []
&gt;&gt;&gt; b.append(&#39;hello&#39;)
&gt;&gt;&gt; print a, b
[[&#39;hello&#39;]] [&#39;hello&#39;]
</pre></div>
</div>
<div class="admonition note">
Note
The proxy types in <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> do nothing to support comparisons
by value. So, for instance, we have:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; manager.list([1,2,3]) == [1,2,3]
False
</pre></div>
</div>
One should just use a copy of the referent instead when making comparisons.
</div>
<dl class="class">
<dt id="multiprocessing.managers.BaseProxy">
class <code class="descclassname">multiprocessing.managers.</code><code class="descname">BaseProxy</code><a class="headerlink" href="#multiprocessing.managers.BaseProxy" title="Permalink to this definition">¶</a></dt>
<dd>Proxy objects are instances of subclasses of <a class="reference internal" href="#multiprocessing.managers.BaseProxy" title="multiprocessing.managers.BaseProxy"><code class="xref py py-class docutils literal notranslate">BaseProxy</code></a>.
<dl class="method">
<dt id="multiprocessing.managers.BaseProxy._callmethod">
<code class="descname">_callmethod</code>(methodname[, args[, kwds]])<a class="headerlink" href="#multiprocessing.managers.BaseProxy._callmethod" title="Permalink to this definition">¶</a></dt>
<dd>Call and return the result of a method of the proxy’s referent.
If <code class="docutils literal notranslate">proxy</code> is a proxy whose referent is <code class="docutils literal notranslate">obj</code> then the expression
<div class="highlight-python notranslate"><div class="highlight"><pre>proxy._callmethod(methodname, args, kwds)
</pre></div>
</div>
will evaluate the expression
<div class="highlight-python notranslate"><div class="highlight"><pre>getattr(obj, methodname)(*args, **kwds)
</pre></div>
</div>
in the manager’s process.
The returned value will be a copy of the result of the call or a proxy to
a new shared object – see documentation for the method_to_typeid
argument of <a class="reference internal" href="#multiprocessing.managers.BaseManager.register" title="multiprocessing.managers.BaseManager.register"><code class="xref py py-meth docutils literal notranslate">BaseManager.register()</code></a>.
If an exception is raised by the call, then is re-raised by
<a class="reference internal" href="#multiprocessing.managers.BaseProxy._callmethod" title="multiprocessing.managers.BaseProxy._callmethod"><code class="xref py py-meth docutils literal notranslate">_callmethod()</code></a>. If some other exception is raised in the manager’s
process then this is converted into a <code class="xref py py-exc docutils literal notranslate">RemoteError</code> exception and is
raised by <a class="reference internal" href="#multiprocessing.managers.BaseProxy._callmethod" title="multiprocessing.managers.BaseProxy._callmethod"><code class="xref py py-meth docutils literal notranslate">_callmethod()</code></a>.
Note in particular that an exception will be raised if methodname has
not been exposed.
An example of the usage of <a class="reference internal" href="#multiprocessing.managers.BaseProxy._callmethod" title="multiprocessing.managers.BaseProxy._callmethod"><code class="xref py py-meth docutils literal notranslate">_callmethod()</code></a>:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; l = manager.list(range(10))
&gt;&gt;&gt; l._callmethod(&#39;__len__&#39;)
10
&gt;&gt;&gt; l._callmethod(&#39;__getslice__&#39;, (2, 7)) # equiv to `l[2:7]`
[2, 3, 4, 5, 6]
&gt;&gt;&gt; l._callmethod(&#39;__getitem__&#39;, (20,)) # equiv to `l[20]`
Traceback (most recent call last):
...
IndexError: list index out of range
</pre></div>
</div>
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.BaseProxy._getvalue">
<code class="descname">_getvalue</code>()<a class="headerlink" href="#multiprocessing.managers.BaseProxy._getvalue" title="Permalink to this definition">¶</a></dt>
<dd>Return a copy of the referent.
If the referent is unpicklable then this will raise an exception.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.BaseProxy.__repr__">
<code class="descname">__repr__</code>()<a class="headerlink" href="#multiprocessing.managers.BaseProxy.__repr__" title="Permalink to this definition">¶</a></dt>
<dd>Return a representation of the proxy object.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.managers.BaseProxy.__str__">
<code class="descname">__str__</code>()<a class="headerlink" href="#multiprocessing.managers.BaseProxy.__str__" title="Permalink to this definition">¶</a></dt>
<dd>Return the representation of the referent.
</dd></dl>

</dd></dl>

<div class="section" id="cleanup">
<h4>16.6.2.8.1. Cleanup<a class="headerlink" href="#cleanup" title="Permalink to this headline">¶</a></h4>
A proxy object uses a weakref callback so that when it gets garbage collected it
deregisters itself from the manager which owns its referent.
A shared object gets deleted from the manager process when there are no longer
any proxies referring to it.
</div>
</div>
<div class="section" id="module-multiprocessing.pool">
<h3>16.6.2.9. Process Pools<a class="headerlink" href="#module-multiprocessing.pool" title="Permalink to this headline">¶</a></h3>
One can create a pool of processes which will carry out tasks submitted to it
with the <code class="xref py py-class docutils literal notranslate">Pool</code> class.
<dl class="class">
<dt id="multiprocessing.pool.multiprocessing.Pool">
class <code class="descclassname">multiprocessing.</code><code class="descname">Pool</code>([processes[, initializer[, initargs[, maxtasksperchild]]]])<a class="headerlink" href="#multiprocessing.pool.multiprocessing.Pool" title="Permalink to this definition">¶</a></dt>
<dd>A process pool object which controls a pool of worker processes to which jobs
can be submitted. It supports asynchronous results with timeouts and
callbacks and has a parallel map implementation.
processes is the number of worker processes to use. If processes is
<code class="docutils literal notranslate">None</code> then the number returned by <code class="xref py py-func docutils literal notranslate">cpu_count()</code> is used. If
initializer is not <code class="docutils literal notranslate">None</code> then each worker process will call
<code class="docutils literal notranslate">initializer(*initargs)</code> when it starts.
Note that the methods of the pool object should only be called by
the process which created the pool.
<div class="versionadded">
New in version 2.7: maxtasksperchild is the number of tasks a worker process can complete
before it will exit and be replaced with a fresh worker process, to enable
unused resources to be freed. The default maxtasksperchild is <code class="docutils literal notranslate">None</code>, which
means worker processes will live as long as the pool.
</div>
<div class="admonition note">
Note
Worker processes within a <code class="xref py py-class docutils literal notranslate">Pool</code> typically live for the complete
duration of the Pool’s work queue. A frequent pattern found in other
systems (such as Apache, mod_wsgi, etc) to free resources held by
workers is to allow a worker within a pool to complete only a set
amount of work before being exiting, being cleaned up and a new
process spawned to replace the old one. The maxtasksperchild
argument to the <code class="xref py py-class docutils literal notranslate">Pool</code> exposes this ability to the end user.
</div>
<dl class="method">
<dt id="multiprocessing.pool.multiprocessing.Pool.apply">
<code class="descname">apply</code>(func[, args[, kwds]])<a class="headerlink" href="#multiprocessing.pool.multiprocessing.Pool.apply" title="Permalink to this definition">¶</a></dt>
<dd>Equivalent of the <a class="reference internal" href="functions.html#apply" title="apply"><code class="xref py py-func docutils literal notranslate">apply()</code></a> built-in function. It blocks until the
result is ready, so <a class="reference internal" href="#multiprocessing.pool.multiprocessing.Pool.apply_async" title="multiprocessing.pool.multiprocessing.Pool.apply_async"><code class="xref py py-meth docutils literal notranslate">apply_async()</code></a> is better suited for performing
work in parallel. Additionally, func is only executed in one of the
workers of the pool.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.pool.multiprocessing.Pool.apply_async">
<code class="descname">apply_async</code>(func[, args[, kwds[, callback]]])<a class="headerlink" href="#multiprocessing.pool.multiprocessing.Pool.apply_async" title="Permalink to this definition">¶</a></dt>
<dd>A variant of the <a class="reference internal" href="functions.html#apply" title="apply"><code class="xref py py-meth docutils literal notranslate">apply()</code></a> method which returns a result object.
If callback is specified then it should be a callable which accepts a
single argument. When the result becomes ready callback is applied to
it (unless the call failed). callback should complete immediately since
otherwise the thread which handles the results will get blocked.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.pool.multiprocessing.Pool.map">
<code class="descname">map</code>(func, iterable[, chunksize])<a class="headerlink" href="#multiprocessing.pool.multiprocessing.Pool.map" title="Permalink to this definition">¶</a></dt>
<dd>A parallel equivalent of the <a class="reference internal" href="functions.html#map" title="map"><code class="xref py py-func docutils literal notranslate">map()</code></a> built-in function (it supports only
one iterable argument though). It blocks until the result is ready.
This method chops the iterable into a number of chunks which it submits to
the process pool as separate tasks. The (approximate) size of these
chunks can be specified by setting chunksize to a positive integer.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.pool.multiprocessing.Pool.map_async">
<code class="descname">map_async</code>(func, iterable[, chunksize[, callback]])<a class="headerlink" href="#multiprocessing.pool.multiprocessing.Pool.map_async" title="Permalink to this definition">¶</a></dt>
<dd>A variant of the <a class="reference internal" href="#multiprocessing.pool.multiprocessing.Pool.map" title="multiprocessing.pool.multiprocessing.Pool.map"><code class="xref py py-meth docutils literal notranslate">map()</code></a> method which returns a result object.
If callback is specified then it should be a callable which accepts a
single argument. When the result becomes ready callback is applied to
it (unless the call failed). callback should complete immediately since
otherwise the thread which handles the results will get blocked.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.pool.multiprocessing.Pool.imap">
<code class="descname">imap</code>(func, iterable[, chunksize])<a class="headerlink" href="#multiprocessing.pool.multiprocessing.Pool.imap" title="Permalink to this definition">¶</a></dt>
<dd>An equivalent of <a class="reference internal" href="itertools.html#itertools.imap" title="itertools.imap"><code class="xref py py-func docutils literal notranslate">itertools.imap()</code></a>.
The chunksize argument is the same as the one used by the <a class="reference internal" href="#multiprocessing.pool.multiprocessing.Pool.map" title="multiprocessing.pool.multiprocessing.Pool.map"><code class="xref py py-meth docutils literal notranslate">map()</code></a>
method. For very long iterables using a large value for chunksize can
make the job complete much faster than using the default value of
<code class="docutils literal notranslate">1</code>.
Also if chunksize is <code class="docutils literal notranslate">1</code> then the <code class="xref py py-meth docutils literal notranslate">next()</code> method of the iterator
returned by the <a class="reference internal" href="#multiprocessing.pool.multiprocessing.Pool.imap" title="multiprocessing.pool.multiprocessing.Pool.imap"><code class="xref py py-meth docutils literal notranslate">imap()</code></a> method has an optional timeout parameter:
<code class="docutils literal notranslate">next(timeout)</code> will raise <code class="xref py py-exc docutils literal notranslate">multiprocessing.TimeoutError</code> if the
result cannot be returned within timeout seconds.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.pool.multiprocessing.Pool.imap_unordered">
<code class="descname">imap_unordered</code>(func, iterable[, chunksize])<a class="headerlink" href="#multiprocessing.pool.multiprocessing.Pool.imap_unordered" title="Permalink to this definition">¶</a></dt>
<dd>The same as <a class="reference internal" href="#multiprocessing.pool.multiprocessing.Pool.imap" title="multiprocessing.pool.multiprocessing.Pool.imap"><code class="xref py py-meth docutils literal notranslate">imap()</code></a> except that the ordering of the results from the
returned iterator should be considered arbitrary. (Only when there is
only one worker process is the order guaranteed to be “correct”.)
</dd></dl>

<dl class="method">
<dt id="multiprocessing.pool.multiprocessing.Pool.close">
<code class="descname">close</code>()<a class="headerlink" href="#multiprocessing.pool.multiprocessing.Pool.close" title="Permalink to this definition">¶</a></dt>
<dd>Prevents any more tasks from being submitted to the pool. Once all the
tasks have been completed the worker processes will exit.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.pool.multiprocessing.Pool.terminate">
<code class="descname">terminate</code>()<a class="headerlink" href="#multiprocessing.pool.multiprocessing.Pool.terminate" title="Permalink to this definition">¶</a></dt>
<dd>Stops the worker processes immediately without completing outstanding
work. When the pool object is garbage collected <a class="reference internal" href="#multiprocessing.pool.multiprocessing.Pool.terminate" title="multiprocessing.pool.multiprocessing.Pool.terminate"><code class="xref py py-meth docutils literal notranslate">terminate()</code></a> will be
called immediately.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.pool.multiprocessing.Pool.join">
<code class="descname">join</code>()<a class="headerlink" href="#multiprocessing.pool.multiprocessing.Pool.join" title="Permalink to this definition">¶</a></dt>
<dd>Wait for the worker processes to exit. One must call <a class="reference internal" href="#multiprocessing.pool.multiprocessing.Pool.close" title="multiprocessing.pool.multiprocessing.Pool.close"><code class="xref py py-meth docutils literal notranslate">close()</code></a> or
<a class="reference internal" href="#multiprocessing.pool.multiprocessing.Pool.terminate" title="multiprocessing.pool.multiprocessing.Pool.terminate"><code class="xref py py-meth docutils literal notranslate">terminate()</code></a> before using <a class="reference internal" href="#multiprocessing.pool.multiprocessing.Pool.join" title="multiprocessing.pool.multiprocessing.Pool.join"><code class="xref py py-meth docutils literal notranslate">join()</code></a>.
</dd></dl>

</dd></dl>

<dl class="class">
<dt id="multiprocessing.pool.AsyncResult">
class <code class="descclassname">multiprocessing.pool.</code><code class="descname">AsyncResult</code><a class="headerlink" href="#multiprocessing.pool.AsyncResult" title="Permalink to this definition">¶</a></dt>
<dd>The class of the result returned by <code class="xref py py-meth docutils literal notranslate">Pool.apply_async()</code> and
<code class="xref py py-meth docutils literal notranslate">Pool.map_async()</code>.
<dl class="method">
<dt id="multiprocessing.pool.AsyncResult.get">
<code class="descname">get</code>([timeout])<a class="headerlink" href="#multiprocessing.pool.AsyncResult.get" title="Permalink to this definition">¶</a></dt>
<dd>Return the result when it arrives. If timeout is not <code class="docutils literal notranslate">None</code> and the
result does not arrive within timeout seconds then
<code class="xref py py-exc docutils literal notranslate">multiprocessing.TimeoutError</code> is raised. If the remote call raised
an exception then that exception will be reraised by <a class="reference internal" href="#multiprocessing.pool.AsyncResult.get" title="multiprocessing.pool.AsyncResult.get"><code class="xref py py-meth docutils literal notranslate">get()</code></a>.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.pool.AsyncResult.wait">
<code class="descname">wait</code>([timeout])<a class="headerlink" href="#multiprocessing.pool.AsyncResult.wait" title="Permalink to this definition">¶</a></dt>
<dd>Wait until the result is available or until timeout seconds pass.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.pool.AsyncResult.ready">
<code class="descname">ready</code>()<a class="headerlink" href="#multiprocessing.pool.AsyncResult.ready" title="Permalink to this definition">¶</a></dt>
<dd>Return whether the call has completed.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.pool.AsyncResult.successful">
<code class="descname">successful</code>()<a class="headerlink" href="#multiprocessing.pool.AsyncResult.successful" title="Permalink to this definition">¶</a></dt>
<dd>Return whether the call completed without raising an exception. Will
raise <a class="reference internal" href="exceptions.html#exceptions.AssertionError" title="exceptions.AssertionError"><code class="xref py py-exc docutils literal notranslate">AssertionError</code></a> if the result is not ready.
</dd></dl>

</dd></dl>

The following example demonstrates the use of a pool:
<div class="highlight-python notranslate"><div class="highlight"><pre>from multiprocessing import Pool
import time

print pool.map(f, range(10)) # prints &quot;[0, 1, 4,..., 81]&quot;

result = pool.apply_async(time.sleep, (10,))
 print result.get(timeout=1) # raises multiprocessing.TimeoutError
</pre></div>
</div>
</div>
<div class="section" id="module-multiprocessing.connection">
<h3>16.6.2.10. Listeners and Clients<a class="headerlink" href="#module-multiprocessing.connection" title="Permalink to this headline">¶</a></h3>
Usually message passing between processes is done using queues or by using
<a class="reference internal" href="#Connection" title="Connection"><code class="xref py py-class docutils literal notranslate">Connection</code></a> objects returned by <a class="reference internal" href="#multiprocessing.Pipe" title="multiprocessing.Pipe"><code class="xref py py-func docutils literal notranslate">Pipe()</code></a>.
However, the <a class="reference internal" href="#module-multiprocessing.connection" title="multiprocessing.connection: API for dealing with sockets."><code class="xref py py-mod docutils literal notranslate">multiprocessing.connection</code></a> module allows some extra
flexibility. It basically gives a high level message oriented API for dealing
with sockets or Windows named pipes, and also has support for digest
authentication using the <a class="reference internal" href="hmac.html#module-hmac" title="hmac: Keyed-Hashing for Message Authentication (HMAC) implementation"><code class="xref py py-mod docutils literal notranslate">hmac</code></a> module.
<dl class="function">
<dt id="multiprocessing.connection.deliver_challenge">
<code class="descclassname">multiprocessing.connection.</code><code class="descname">deliver_challenge</code>(connection, authkey)<a class="headerlink" href="#multiprocessing.connection.deliver_challenge" title="Permalink to this definition">¶</a></dt>
<dd>Send a randomly generated message to the other end of the connection and wait
for a reply.
If the reply matches the digest of the message using authkey as the key
then a welcome message is sent to the other end of the connection. Otherwise
<a class="reference internal" href="#multiprocessing.connection.AuthenticationError" title="multiprocessing.connection.AuthenticationError"><code class="xref py py-exc docutils literal notranslate">AuthenticationError</code></a> is raised.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.connection.answer_challenge">
<code class="descclassname">multiprocessing.connection.</code><code class="descname">answer_challenge</code>(connection, authkey)<a class="headerlink" href="#multiprocessing.connection.answer_challenge" title="Permalink to this definition">¶</a></dt>
<dd>Receive a message, calculate the digest of the message using authkey as the
key, and then send the digest back.
If a welcome message is not received, then <a class="reference internal" href="#multiprocessing.connection.AuthenticationError" title="multiprocessing.connection.AuthenticationError"><code class="xref py py-exc docutils literal notranslate">AuthenticationError</code></a> is
raised.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.connection.Client">
<code class="descclassname">multiprocessing.connection.</code><code class="descname">Client</code>(address[, family[, authenticate[, authkey]]])<a class="headerlink" href="#multiprocessing.connection.Client" title="Permalink to this definition">¶</a></dt>
<dd>Attempt to set up a connection to the listener which is using address
address, returning a <a class="reference internal" href="#Connection" title="Connection"><code class="xref py py-class docutils literal notranslate">Connection</code></a>.
The type of the connection is determined by family argument, but this can
generally be omitted since it can usually be inferred from the format of
address. (See <a class="reference internal" href="#multiprocessing-address-formats">Address Formats</a>)
If authenticate is <code class="docutils literal notranslate">True</code> or authkey is a string then digest
authentication is used. The key used for authentication will be either
authkey or <code class="docutils literal notranslate">current_process().authkey)</code> if authkey is <code class="docutils literal notranslate">None</code>.
If authentication fails then <a class="reference internal" href="#multiprocessing.connection.AuthenticationError" title="multiprocessing.connection.AuthenticationError"><code class="xref py py-exc docutils literal notranslate">AuthenticationError</code></a> is raised. See
<a class="reference internal" href="#multiprocessing-auth-keys">Authentication keys</a>.
</dd></dl>

<dl class="class">
<dt id="multiprocessing.connection.Listener">
class <code class="descclassname">multiprocessing.connection.</code><code class="descname">Listener</code>([address[, family[, backlog[, authenticate[, authkey]]]]])<a class="headerlink" href="#multiprocessing.connection.Listener" title="Permalink to this definition">¶</a></dt>
<dd>A wrapper for a bound socket or Windows named pipe which is ‘listening’ for
connections.
address is the address to be used by the bound socket or named pipe of the
listener object.
<div class="admonition note">
Note
If an address of ‘0.0.0.0’ is used, the address will not be a connectable
end point on Windows. If you require a connectable end-point,
you should use ‘127.0.0.1’.
</div>
family is the type of socket (or named pipe) to use. This can be one of
the strings <code class="docutils literal notranslate">'AF_INET'</code> (for a TCP socket), <code class="docutils literal notranslate">'AF_UNIX'</code> (for a Unix
domain socket) or <code class="docutils literal notranslate">'AF_PIPE'</code> (for a Windows named pipe). Of these only
the first is guaranteed to be available. If family is <code class="docutils literal notranslate">None</code> then the
family is inferred from the format of address. If address is also
<code class="docutils literal notranslate">None</code> then a default is chosen. This default is the family which is
assumed to be the fastest available. See
<a class="reference internal" href="#multiprocessing-address-formats">Address Formats</a>. Note that if family is
<code class="docutils literal notranslate">'AF_UNIX'</code> and address is <code class="docutils literal notranslate">None</code> then the socket will be created in a
private temporary directory created using <a class="reference internal" href="tempfile.html#tempfile.mkstemp" title="tempfile.mkstemp"><code class="xref py py-func docutils literal notranslate">tempfile.mkstemp()</code></a>.
If the listener object uses a socket then backlog (1 by default) is passed
to the <a class="reference internal" href="socket.html#socket.socket.listen" title="socket.socket.listen"><code class="xref py py-meth docutils literal notranslate">listen()</code></a> method of the socket once it has been
bound.
If authenticate is <code class="docutils literal notranslate">True</code> (<code class="docutils literal notranslate">False</code> by default) or authkey is not
<code class="docutils literal notranslate">None</code> then digest authentication is used.
If authkey is a string then it will be used as the authentication key;
otherwise it must be <code class="docutils literal notranslate">None</code>.
If authkey is <code class="docutils literal notranslate">None</code> and authenticate is <code class="docutils literal notranslate">True</code> then
<code class="docutils literal notranslate">current_process().authkey</code> is used as the authentication key. If
authkey is <code class="docutils literal notranslate">None</code> and authenticate is <code class="docutils literal notranslate">False</code> then no
authentication is done. If authentication fails then
<a class="reference internal" href="#multiprocessing.connection.AuthenticationError" title="multiprocessing.connection.AuthenticationError"><code class="xref py py-exc docutils literal notranslate">AuthenticationError</code></a> is raised. See <a class="reference internal" href="#multiprocessing-auth-keys">Authentication keys</a>.
<dl class="method">
<dt id="multiprocessing.connection.Listener.accept">
<code class="descname">accept</code>()<a class="headerlink" href="#multiprocessing.connection.Listener.accept" title="Permalink to this definition">¶</a></dt>
<dd>Accept a connection on the bound socket or named pipe of the listener
object and return a <a class="reference internal" href="#Connection" title="Connection"><code class="xref py py-class docutils literal notranslate">Connection</code></a> object.
If authentication is attempted and fails, then
<code class="xref py py-exc docutils literal notranslate">AuthenticationError</code> is raised.
</dd></dl>

<dl class="method">
<dt id="multiprocessing.connection.Listener.close">
<code class="descname">close</code>()<a class="headerlink" href="#multiprocessing.connection.Listener.close" title="Permalink to this definition">¶</a></dt>
<dd>Close the bound socket or named pipe of the listener object. This is
called automatically when the listener is garbage collected. However it
is advisable to call it explicitly.
</dd></dl>

Listener objects have the following read-only properties:
<dl class="attribute">
<dt id="multiprocessing.connection.Listener.address">
<code class="descname">address</code><a class="headerlink" href="#multiprocessing.connection.Listener.address" title="Permalink to this definition">¶</a></dt>
<dd>The address which is being used by the Listener object.
</dd></dl>

<dl class="attribute">
<dt id="multiprocessing.connection.Listener.last_accepted">
<code class="descname">last_accepted</code><a class="headerlink" href="#multiprocessing.connection.Listener.last_accepted" title="Permalink to this definition">¶</a></dt>
<dd>The address from which the last accepted connection came. If this is
unavailable then it is <code class="docutils literal notranslate">None</code>.
</dd></dl>

</dd></dl>

The module defines the following exceptions:
<dl class="exception">
<dt id="multiprocessing.connection.ProcessError">
exception <code class="descclassname">multiprocessing.connection.</code><code class="descname">ProcessError</code><a class="headerlink" href="#multiprocessing.connection.ProcessError" title="Permalink to this definition">¶</a></dt>
<dd>The base class of all <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> exceptions.
</dd></dl>

<dl class="exception">
<dt id="multiprocessing.connection.BufferTooShort">
exception <code class="descclassname">multiprocessing.connection.</code><code class="descname">BufferTooShort</code><a class="headerlink" href="#multiprocessing.connection.BufferTooShort" title="Permalink to this definition">¶</a></dt>
<dd>Exception raised by <a class="reference internal" href="#Connection.recv_bytes_into" title="Connection.recv_bytes_into"><code class="xref py py-meth docutils literal notranslate">Connection.recv_bytes_into()</code></a> when the supplied
buffer object is too small for the message read.
</dd></dl>

<dl class="exception">
<dt id="multiprocessing.connection.AuthenticationError">
exception <code class="descclassname">multiprocessing.connection.</code><code class="descname">AuthenticationError</code><a class="headerlink" href="#multiprocessing.connection.AuthenticationError" title="Permalink to this definition">¶</a></dt>
<dd>Raised when there is an authentication error.
</dd></dl>

<dl class="exception">
<dt id="multiprocessing.connection.TimeoutError">
exception <code class="descclassname">multiprocessing.connection.</code><code class="descname">TimeoutError</code><a class="headerlink" href="#multiprocessing.connection.TimeoutError" title="Permalink to this definition">¶</a></dt>
<dd>Raised by methods with a timeout when the timeout expires.
</dd></dl>

Examples
The following server code creates a listener which uses <code class="docutils literal notranslate">'secret password'</code> as
an authentication key. It then waits for a connection and sends some data to
the client:
<div class="highlight-python notranslate"><div class="highlight"><pre>from multiprocessing.connection import Listener
from array import array

address = (&#39;localhost&#39;, 6000) # family is deduced to be &#39;AF_INET&#39;
listener = Listener(address, authkey=&#39;secret password&#39;)

conn = listener.accept()
print &#39;connection accepted from&#39;, listener.last_accepted

conn.send([2.25, None, &#39;junk&#39;, float])

conn.send_bytes(&#39;hello&#39;)

conn.send_bytes(array(&#39;i&#39;, [42, 1729]))

conn.close()
listener.close()
</pre></div>
</div>
The following code connects to the server and receives some data from the
server:
<div class="highlight-python notranslate"><div class="highlight"><pre>from multiprocessing.connection import Client
from array import array

address = (&#39;localhost&#39;, 6000)
conn = Client(address, authkey=&#39;secret password&#39;)

print conn.recv() # =&gt; [2.25, None, &#39;junk&#39;, float]

print conn.recv_bytes() # =&gt; &#39;hello&#39;

arr = array(&#39;i&#39;, [0, 0, 0, 0, 0])
print conn.recv_bytes_into(arr) # =&gt; 8
print arr # =&gt; array(&#39;i&#39;, [42, 1729, 0, 0, 0])

conn.close()
</pre></div>
</div>
<div class="section" id="address-formats">
<h4>16.6.2.10.1. Address Formats<a class="headerlink" href="#address-formats" title="Permalink to this headline">¶</a></h4>
<ul class="simple">
<li>An <code class="docutils literal notranslate">'AF_INET'</code> address is a tuple of the form <code class="docutils literal notranslate">(hostname, port)</code> where
hostname is a string and port is an integer.</li>
<li>An <code class="docutils literal notranslate">'AF_UNIX'</code> address is a string representing a filename on the
filesystem.</li>
<li><dl class="first docutils">
<dt>An <code class="docutils literal notranslate">'AF_PIPE'</code> address is a string of the form</dt>
<dd><code class="samp docutils literal notranslate">r'\\.\pipe\PipeName'</code>. To use <a class="reference internal" href="#multiprocessing.connection.Client" title="multiprocessing.connection.Client"><code class="xref py py-func docutils literal notranslate">Client()</code></a> to connect to a named
pipe on a remote computer called ServerName one should use an address of the
form <code class="samp docutils literal notranslate">r'\\ServerName\pipe\PipeName'</code> instead.</dd>
</dl>
</li>
</ul>
Note that any string beginning with two backslashes is assumed by default to be
an <code class="docutils literal notranslate">'AF_PIPE'</code> address rather than an <code class="docutils literal notranslate">'AF_UNIX'</code> address.
</div>
</div>
<div class="section" id="authentication-keys">
<h3>16.6.2.11. Authentication keys<a class="headerlink" href="#authentication-keys" title="Permalink to this headline">¶</a></h3>
When one uses <a class="reference internal" href="#Connection.recv" title="Connection.recv"><code class="xref py py-meth docutils literal notranslate">Connection.recv()</code></a>, the
data received is automatically
unpickled. Unfortunately unpickling data from an untrusted source is a security
risk. Therefore <a class="reference internal" href="#multiprocessing.connection.Listener" title="multiprocessing.connection.Listener"><code class="xref py py-class docutils literal notranslate">Listener</code></a> and <a class="reference internal" href="#multiprocessing.connection.Client" title="multiprocessing.connection.Client"><code class="xref py py-func docutils literal notranslate">Client()</code></a> use the <a class="reference internal" href="hmac.html#module-hmac" title="hmac: Keyed-Hashing for Message Authentication (HMAC) implementation"><code class="xref py py-mod docutils literal notranslate">hmac</code></a> module
to provide digest authentication.
An authentication key is a string which can be thought of as a password: once a
connection is established both ends will demand proof that the other knows the
authentication key. (Demonstrating that both ends are using the same key does
not involve sending the key over the connection.)
If authentication is requested but no authentication key is specified then the
return value of <code class="docutils literal notranslate">current_process().authkey</code> is used (see
<a class="reference internal" href="#multiprocessing.Process" title="multiprocessing.Process"><code class="xref py py-class docutils literal notranslate">Process</code></a>). This value will be automatically inherited by
any <a class="reference internal" href="#multiprocessing.Process" title="multiprocessing.Process"><code class="xref py py-class docutils literal notranslate">Process</code></a> object that the current process creates.
This means that (by default) all processes of a multi-process program will share
a single authentication key which can be used when setting up connections
between themselves.
Suitable authentication keys can also be generated by using <a class="reference internal" href="os.html#os.urandom" title="os.urandom"><code class="xref py py-func docutils literal notranslate">os.urandom()</code></a>.
</div>
<div class="section" id="logging">
<h3>16.6.2.12. Logging<a class="headerlink" href="#logging" title="Permalink to this headline">¶</a></h3>
Some support for logging is available. Note, however, that the <a class="reference internal" href="logging.html#module-logging" title="logging: Flexible event logging system for applications."><code class="xref py py-mod docutils literal notranslate">logging</code></a>
package does not use process shared locks so it is possible (depending on the
handler type) for messages from different processes to get mixed up.
<dl class="function">
<dt id="multiprocessing.get_logger">
<code class="descclassname">multiprocessing.</code><code class="descname">get_logger</code>()<a class="headerlink" href="#multiprocessing.get_logger" title="Permalink to this definition">¶</a></dt>
<dd>Returns the logger used by <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a>. If necessary, a new one
will be created.
When first created the logger has level <code class="xref py py-data docutils literal notranslate">logging.NOTSET</code> and no
default handler. Messages sent to this logger will not by default propagate
to the root logger.
Note that on Windows child processes will only inherit the level of the
parent process’s logger – any other customization of the logger will not be
inherited.
</dd></dl>

<dl class="function">
<dt id="multiprocessing.log_to_stderr">
<code class="descclassname">multiprocessing.</code><code class="descname">log_to_stderr</code>()<a class="headerlink" href="#multiprocessing.log_to_stderr" title="Permalink to this definition">¶</a></dt>
<dd>This function performs a call to <a class="reference internal" href="#multiprocessing.get_logger" title="multiprocessing.get_logger"><code class="xref py py-func docutils literal notranslate">get_logger()</code></a> but in addition to
returning the logger created by get_logger, it adds a handler which sends
output to <a class="reference internal" href="sys.html#sys.stderr" title="sys.stderr"><code class="xref py py-data docutils literal notranslate">sys.stderr</code></a> using format
<code class="docutils literal notranslate">'[%(levelname)s/%(processName)s] %(message)s'</code>.
</dd></dl>

Below is an example session with logging turned on:
<div class="highlight-python notranslate"><div class="highlight"><pre>&gt;&gt;&gt; import multiprocessing, logging
&gt;&gt;&gt; logger = multiprocessing.log_to_stderr()
&gt;&gt;&gt; logger.setLevel(logging.INFO)
&gt;&gt;&gt; logger.warning(&#39;doomed&#39;)
[WARNING/MainProcess] doomed
&gt;&gt;&gt; m = multiprocessing.Manager()
[INFO/SyncManager-...] child process calling self.run()
[INFO/SyncManager-...] created temp directory /.../pymp-...
[INFO/SyncManager-...] manager serving at &#39;/.../listener-...&#39;
&gt;&gt;&gt; del m
[INFO/MainProcess] sending shutdown message to manager
[INFO/SyncManager-...] manager exiting with exitcode 0
</pre></div>
</div>
In addition to having these two logging functions, the multiprocessing also
exposes two additional logging level attributes. These are <code class="xref py py-const docutils literal notranslate">SUBWARNING</code>
and <code class="xref py py-const docutils literal notranslate">SUBDEBUG</code>. The table below illustrates where theses fit in the
normal level hierarchy.
<table border="1" class="docutils">
<colgroup>
<col width="50%" />
<col width="50%" />
</colgroup>
<thead valign="bottom">
<tr class="row-odd"><th class="head">Level</th>
<th class="head">Numeric value</th>
</tr>
</thead>
<tbody valign="top">
<tr class="row-even"><td><code class="docutils literal notranslate">SUBWARNING</code></td>
<td>25</td>
</tr>
<tr class="row-odd"><td><code class="docutils literal notranslate">SUBDEBUG</code></td>
<td>5</td>
</tr>
</tbody>
</table>
For a full table of logging levels, see the <a class="reference internal" href="logging.html#module-logging" title="logging: Flexible event logging system for applications."><code class="xref py py-mod docutils literal notranslate">logging</code></a> module.
These additional logging levels are used primarily for certain debug messages
within the multiprocessing module. Below is the same example as above, except
with <code class="xref py py-const docutils literal notranslate">SUBDEBUG</code> enabled:
<div class="highlight-python notranslate"><div class="highlight"><pre>&gt;&gt;&gt; import multiprocessing, logging
&gt;&gt;&gt; logger = multiprocessing.log_to_stderr()
&gt;&gt;&gt; logger.setLevel(multiprocessing.SUBDEBUG)
&gt;&gt;&gt; logger.warning(&#39;doomed&#39;)
[WARNING/MainProcess] doomed
&gt;&gt;&gt; m = multiprocessing.Manager()
[INFO/SyncManager-...] child process calling self.run()
[INFO/SyncManager-...] created temp directory /.../pymp-...
[INFO/SyncManager-...] manager serving at &#39;/.../pymp-djGBXN/listener-...&#39;
&gt;&gt;&gt; del m
[SUBDEBUG/MainProcess] finalizer calling ...
[INFO/MainProcess] sending shutdown message to manager
[DEBUG/SyncManager-...] manager received shutdown message
[SUBDEBUG/SyncManager-...] calling &lt;Finalize object, callback=unlink, ...
[SUBDEBUG/SyncManager-...] finalizer calling &lt;built-in function unlink&gt; ...
[SUBDEBUG/SyncManager-...] calling &lt;Finalize object, dead&gt;
[SUBDEBUG/SyncManager-...] finalizer calling &lt;function rmtree at 0x5aa730&gt; ...
[INFO/SyncManager-...] manager exiting with exitcode 0
</pre></div>
</div>
</div>
<div class="section" id="module-multiprocessing.dummy">
<h3>16.6.2.13. The <a class="reference internal" href="#module-multiprocessing.dummy" title="multiprocessing.dummy: Dumb wrapper around threading."><code class="xref py py-mod docutils literal notranslate">multiprocessing.dummy</code></a> module<a class="headerlink" href="#module-multiprocessing.dummy" title="Permalink to this headline">¶</a></h3>
<a class="reference internal" href="#module-multiprocessing.dummy" title="multiprocessing.dummy: Dumb wrapper around threading."><code class="xref py py-mod docutils literal notranslate">multiprocessing.dummy</code></a> replicates the API of <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> but is
no more than a wrapper around the <a class="reference internal" href="threading.html#module-threading" title="threading: Higher-level threading interface."><code class="xref py py-mod docutils literal notranslate">threading</code></a> module.
</div>
</div>
<div class="section" id="programming-guidelines">
<h2>16.6.3. Programming guidelines<a class="headerlink" href="#programming-guidelines" title="Permalink to this headline">¶</a></h2>
There are certain guidelines and idioms which should be adhered to when using
<a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a>.
<div class="section" id="all-platforms">
<h3>16.6.3.1. All platforms<a class="headerlink" href="#all-platforms" title="Permalink to this headline">¶</a></h3>
Avoid shared state
<blockquote>
<div>As far as possible one should try to avoid shifting large amounts of data
between processes.
It is probably best to stick to using queues or pipes for communication
between processes rather than using the lower level synchronization
primitives from the <a class="reference internal" href="threading.html#module-threading" title="threading: Higher-level threading interface."><code class="xref py py-mod docutils literal notranslate">threading</code></a> module.
</div></blockquote>
Picklability
<blockquote>
<div>Ensure that the arguments to the methods of proxies are picklable.</div></blockquote>
Thread safety of proxies
<blockquote>
<div>Do not use a proxy object from more than one thread unless you protect it
with a lock.
(There is never a problem with different processes using the same proxy.)
</div></blockquote>
Joining zombie processes
<blockquote>
<div>On Unix when a process finishes but has not been joined it becomes a zombie.
There should never be very many because each time a new process starts (or
<a class="reference internal" href="#multiprocessing.active_children" title="multiprocessing.active_children"><code class="xref py py-func docutils literal notranslate">active_children()</code></a> is called) all completed processes
which have not yet been joined will be joined. Also calling a finished
process’s <a class="reference internal" href="#multiprocessing.Process.is_alive" title="multiprocessing.Process.is_alive"><code class="xref py py-meth docutils literal notranslate">Process.is_alive</code></a> will
join the process. Even so it is probably good
practice to explicitly join all the processes that you start.</div></blockquote>
Better to inherit than pickle/unpickle
<blockquote>
<div>On Windows many types from <a class="reference internal" href="#module-multiprocessing" title="multiprocessing: Process-based &quot;threading&quot; interface."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> need to be picklable so
that child processes can use them. However, one should generally avoid
sending shared objects to other processes using pipes or queues. Instead
you should arrange the program so that a process which needs access to a
shared resource created elsewhere can inherit it from an ancestor process.</div></blockquote>
Avoid terminating processes
<blockquote>
<div>Using the <a class="reference internal" href="#multiprocessing.Process.terminate" title="multiprocessing.Process.terminate"><code class="xref py py-meth docutils literal notranslate">Process.terminate</code></a>
method to stop a process is liable to
cause any shared resources (such as locks, semaphores, pipes and queues)
currently being used by the process to become broken or unavailable to other
processes.
Therefore it is probably best to only consider using
<a class="reference internal" href="#multiprocessing.Process.terminate" title="multiprocessing.Process.terminate"><code class="xref py py-meth docutils literal notranslate">Process.terminate</code></a> on processes
which never use any shared resources.
</div></blockquote>
Joining processes that use queues
<blockquote>
<div>Bear in mind that a process that has put items in a queue will wait before
terminating until all the buffered items are fed by the “feeder” thread to
the underlying pipe. (The child process can call the
<a class="reference internal" href="#multiprocessing.Queue.cancel_join_thread" title="multiprocessing.Queue.cancel_join_thread"><code class="xref py py-meth docutils literal notranslate">cancel_join_thread()</code></a> method of the queue to avoid this behaviour.)
This means that whenever you use a queue you need to make sure that all
items which have been put on the queue will eventually be removed before the
process is joined. Otherwise you cannot be sure that processes which have
put items on the queue will terminate. Remember also that non-daemonic
processes will be joined automatically.
An example which will deadlock is the following:
<div class="highlight-python notranslate"><div class="highlight"><pre>from multiprocessing import Process, Queue

def f(q):
 q.put(&#39;X&#39; * 1000000)

def f():
 ... do something using &quot;lock&quot; ...

def f(l):
 ... do something using &quot;l&quot; ...

def foo():
 print &#39;hello&#39;

p = Process(target=foo)
p.start()
</pre></div>
</div>
Instead one should protect the “entry point” of the program by using <code class="docutils literal notranslate">if
__name__ == '__main__':</code> as follows:
<div class="highlight-python notranslate"><div class="highlight"><pre>from multiprocessing import Process, freeze_support

def foo():
 print &#39;hello&#39;

from multiprocessing import freeze_support
from multiprocessing.managers import BaseManager, BaseProxy
import operator

##

class Foo(object):
 def f(self):
 print &#39;you called Foo.f()&#39;
 def g(self):
 print &#39;you called Foo.g()&#39;
 def _h(self):
 print &#39;you called Foo._h()&#39;

# A simple generator function
def baz():
 for i in xrange(10):
 yield i*i

# Proxy type for generator objects
class GeneratorProxy(BaseProxy):
 _exposed_ = (&#39;next&#39;, &#39;__next__&#39;)
 def __iter__(self):
 return self
 def next(self):
 return self._callmethod(&#39;next&#39;)
 def __next__(self):
 return self._callmethod(&#39;__next__&#39;)

# Function to return the operator module
def get_operator_module():
 return operator

##

class MyManager(BaseManager):
 pass

# register the Foo class; make `f()` and `g()` accessible via proxy
MyManager.register(&#39;Foo1&#39;, Foo)

# register the Foo class; make `g()` and `_h()` accessible via proxy
MyManager.register(&#39;Foo2&#39;, Foo, exposed=(&#39;g&#39;, &#39;_h&#39;))

# register the generator function baz; use `GeneratorProxy` to make proxies
MyManager.register(&#39;baz&#39;, baz, proxytype=GeneratorProxy)

# register get_operator_module(); make public functions accessible via proxy
MyManager.register(&#39;operator&#39;, get_operator_module)

##

def test():
 manager = MyManager()
 manager.start()

print &#39;-&#39; * 20

f1 = manager.Foo1()
 f1.f()
 f1.g()
 assert not hasattr(f1, &#39;_h&#39;)
 assert sorted(f1._exposed_) == sorted([&#39;f&#39;, &#39;g&#39;])

print &#39;-&#39; * 20

f2 = manager.Foo2()
 f2.g()
 f2._h()
 assert not hasattr(f2, &#39;f&#39;)
 assert sorted(f2._exposed_) == sorted([&#39;g&#39;, &#39;_h&#39;])

print &#39;-&#39; * 20

it = manager.baz()
 for i in it:
 print &#39;&lt;%d&gt;&#39; % i,
 print

print &#39;-&#39; * 20

op = manager.operator()
 print &#39;op.add(23, 45) =&#39;, op.add(23, 45)
 print &#39;op.pow(2, 94) =&#39;, op.pow(2, 94)
 print &#39;op.getslice(range(10), 2, 6) =&#39;, op.getslice(range(10), 2, 6)
 print &#39;op.repeat(range(5), 3) =&#39;, op.repeat(range(5), 3)
 print &#39;op._exposed_ =&#39;, op._exposed_

##

import multiprocessing
import time
import random
import sys

#
# Functions used by test code
#

def calculate(func, args):
 result = func(*args)
 return &#39;%s says that %s%s = %s&#39; % (
 multiprocessing.current_process().name,
 func.__name__, args, result
 )

def calculatestar(args):
 return calculate(*args)

def mul(a, b):
 time.sleep(0.5*random.random())
 return a * b

def plus(a, b):
 time.sleep(0.5*random.random())
 return a + b

def f(x):
 return 1.0 / (x-5.0)

def pow3(x):
 return x**3

def noop(x):
 pass

#
# Test code
#

def test():
 print &#39;cpu_count() = %d\n&#39; % multiprocessing.cpu_count()

#
 # Create pool
 #

PROCESSES = 4
 print &#39;Creating pool with %d processes\n&#39; % PROCESSES
 pool = multiprocessing.Pool(PROCESSES)
 print &#39;pool = %s&#39; % pool
 print

#
 # Tests
 #

TASKS = [(mul, (i, 7)) for i in range(10)] + \
 [(plus, (i, 8)) for i in range(10)]

results = [pool.apply_async(calculate, t) for t in TASKS]
 imap_it = pool.imap(calculatestar, TASKS)
 imap_unordered_it = pool.imap_unordered(calculatestar, TASKS)

print &#39;Ordered results using pool.apply_async():&#39;
 for r in results:
 print &#39;\t&#39;, r.get()
 print

print &#39;Ordered results using pool.imap():&#39;
 for x in imap_it:
 print &#39;\t&#39;, x
 print

print &#39;Unordered results using pool.imap_unordered():&#39;
 for x in imap_unordered_it:
 print &#39;\t&#39;, x
 print

print &#39;Ordered results using pool.map() --- will block till complete:&#39;
 for x in pool.map(calculatestar, TASKS):
 print &#39;\t&#39;, x
 print

#
 # Simple benchmarks
 #

N = 100000
 print &#39;def pow3(x): return x**3&#39;

t = time.time()
 B = pool.map(pow3, xrange(N))
 print &#39;\tpool.map(pow3, xrange(%d)):\n\t\t%s seconds&#39; % \
 (N, time.time() - t)

t = time.time()
 C = list(pool.imap(pow3, xrange(N), chunksize=N//8))
 print &#39;\tlist(pool.imap(pow3, xrange(%d), chunksize=%d)):\n\t\t%s&#39; \
 &#39; seconds&#39; % (N, N//8, time.time() - t)

assert A == B == C, (len(A), len(B), len(C))
 print

L = [None] * 1000000
 print &#39;def noop(x): pass&#39;
 print &#39;L = [None] * 1000000&#39;

t = time.time()
 C = list(pool.imap(noop, L, chunksize=len(L)//8))
 print &#39;\tlist(pool.imap(noop, L, chunksize=%d)):\n\t\t%s seconds&#39; % \
 (len(L)//8, time.time() - t)

del A, B, C, L

#
 # Test error handling
 #

print &#39;Testing error handling:&#39;

try:
 print pool.apply(f, (5,))
 except ZeroDivisionError:
 print &#39;\tGot ZeroDivisionError as expected from pool.apply()&#39;
 else:
 raise AssertionError(&#39;expected ZeroDivisionError&#39;)

try:
 print pool.map(f, range(10))
 except ZeroDivisionError:
 print &#39;\tGot ZeroDivisionError as expected from pool.map()&#39;
 else:
 raise AssertionError(&#39;expected ZeroDivisionError&#39;)

try:
 print list(pool.imap(f, range(10)))
 except ZeroDivisionError:
 print &#39;\tGot ZeroDivisionError as expected from list(pool.imap())&#39;
 else:
 raise AssertionError(&#39;expected ZeroDivisionError&#39;)

it = pool.imap(f, range(10))
 for i in range(10):
 try:
 x = it.next()
 except ZeroDivisionError:
 if i == 5:
 pass
 except StopIteration:
 break
 else:
 if i == 5:
 raise AssertionError(&#39;expected ZeroDivisionError&#39;)

assert i == 9
 print &#39;\tGot ZeroDivisionError as expected from IMapIterator.next()&#39;
 print

#
 # Testing timeouts
 #

print &#39;Testing ApplyResult.get() with timeout:&#39;,
 res = pool.apply_async(calculate, TASKS[0])
 while 1:
 sys.stdout.flush()
 try:
 sys.stdout.write(&#39;\n\t%s&#39; % res.get(0.02))
 break
 except multiprocessing.TimeoutError:
 sys.stdout.write(&#39;.&#39;)
 print
 print

print &#39;Testing IMapIterator.next() with timeout:&#39;,
 it = pool.imap(calculatestar, TASKS)
 while 1:
 sys.stdout.flush()
 try:
 sys.stdout.write(&#39;\n\t%s&#39; % it.next(0.02))
 except StopIteration:
 break
 except multiprocessing.TimeoutError:
 sys.stdout.write(&#39;.&#39;)
 print
 print

#
 # Testing callback
 #

print &#39;Testing callback:&#39;

A = []
 B = [56, 0, 1, 8, 27, 64, 125, 216, 343, 512, 729]

r = pool.apply_async(mul, (7, 8), callback=A.append)
 r.wait()

r = pool.map_async(pow3, range(10), callback=A.extend)
 r.wait()

if A == B:
 print &#39;\tcallbacks succeeded\n&#39;
 else:
 print &#39;\t*** callbacks failed\n\t\t%s != %s\n&#39; % (A, B)

#
 # Check there are no outstanding tasks
 #

assert not pool._cache, &#39;cache = %r&#39; % pool._cache

#
 # Check close() methods
 #

print &#39;Testing close():&#39;

for worker in pool._pool:
 assert worker.is_alive()

assert result.get() is None

for worker in pool._pool:
 assert not worker.is_alive()

print &#39;\tclose() succeeded\n&#39;

#
 # Check terminate() method
 #

print &#39;Testing terminate():&#39;

pool = multiprocessing.Pool(2)
 DELTA = 0.1
 ignore = pool.apply(pow3, [2])
 results = [pool.apply_async(time.sleep, [DELTA]) for i in range(100)]
 pool.terminate()
 pool.join()

for worker in pool._pool:
 assert not worker.is_alive()

print &#39;\tterminate() succeeded\n&#39;

#
 # Check garbage collection
 #

print &#39;Testing garbage collection:&#39;

pool = multiprocessing.Pool(2)
 DELTA = 0.1
 processes = pool._pool
 ignore = pool.apply(pow3, [2])
 results = [pool.apply_async(time.sleep, [DELTA]) for i in range(100)]

results = pool = None

time.sleep(DELTA * 2)

for worker in processes:
 assert not worker.is_alive()

print &#39;\tgarbage collection succeeded\n&#39;

assert len(sys.argv) in (1, 2)

test()
</pre></div>
</div>
Synchronization types like locks, conditions and queues:
<div class="highlight-python notranslate"><div class="highlight"><pre>#
# A test file for the `multiprocessing` package
#
# Copyright (c) 2006-2008, R Oudkerk
# All rights reserved.
#

import time, sys, random
from Queue import Empty

import multiprocessing # may get overwritten

#### TEST_VALUE

def value_func(running, mutex):
 random.seed()
 time.sleep(random.random()*4)

mutex.acquire()
 print &#39;\n\t\t\t&#39; + str(multiprocessing.current_process()) + &#39; has finished&#39;
 running.value -= 1
 mutex.release()

def test_value():
 TASKS = 10
 running = multiprocessing.Value(&#39;i&#39;, TASKS)
 mutex = multiprocessing.Lock()

for i in range(TASKS):
 p = multiprocessing.Process(target=value_func, args=(running, mutex))
 p.start()

while running.value &gt; 0:
 time.sleep(0.08)
 mutex.acquire()
 print running.value,
 sys.stdout.flush()
 mutex.release()

print
 print &#39;No more running processes&#39;

#### TEST_QUEUE

def queue_func(queue):
 for i in range(30):
 time.sleep(0.5 * random.random())
 queue.put(i*i)
 queue.put(&#39;STOP&#39;)

def test_queue():
 q = multiprocessing.Queue()

o = None
 while o != &#39;STOP&#39;:
 try:
 o = q.get(timeout=0.3)
 print o,
 sys.stdout.flush()
 except Empty:
 print &#39;TIMEOUT&#39;

print

#### TEST_CONDITION

def condition_func(cond):
 cond.acquire()
 print &#39;\t&#39; + str(cond)
 time.sleep(2)
 print &#39;\tchild is notifying&#39;
 print &#39;\t&#39; + str(cond)
 cond.notify()
 cond.release()

def test_condition():
 cond = multiprocessing.Condition()

cond.acquire()
 print cond
 cond.acquire()
 print cond

p.start()

print &#39;main is waiting&#39;
 cond.wait()
 print &#39;main has woken up&#39;

print cond
 cond.release()
 print cond
 cond.release()

p.join()
 print cond

#### TEST_SEMAPHORE

def semaphore_func(sema, mutex, running):
 sema.acquire()

mutex.acquire()
 running.value += 1
 print running.value, &#39;tasks are running&#39;
 mutex.release()

random.seed()
 time.sleep(random.random()*2)

mutex.acquire()
 running.value -= 1
 print &#39;%s has finished&#39; % multiprocessing.current_process()
 mutex.release()

sema.release()

def test_semaphore():
 sema = multiprocessing.Semaphore(3)
 mutex = multiprocessing.RLock()
 running = multiprocessing.Value(&#39;i&#39;, 0)

processes = [
 multiprocessing.Process(target=semaphore_func,
 args=(sema, mutex, running))
 for i in range(10)
 ]

for p in processes:
 p.start()

#### TEST_JOIN_TIMEOUT

def join_timeout_func():
 print &#39;\tchild sleeping&#39;
 time.sleep(5.5)
 print &#39;\n\tchild terminating&#39;

def test_join_timeout():
 p = multiprocessing.Process(target=join_timeout_func)
 p.start()

print &#39;waiting for process to finish&#39;

while 1:
 p.join(timeout=1)
 if not p.is_alive():
 break
 print &#39;.&#39;,
 sys.stdout.flush()

#### TEST_EVENT

def event_func(event):
 print &#39;\t%r is waiting&#39; % multiprocessing.current_process()
 event.wait()
 print &#39;\t%r has woken up&#39; % multiprocessing.current_process()

def test_event():
 event = multiprocessing.Event()

processes = [multiprocessing.Process(target=event_func, args=(event,))
 for i in range(5)]

print &#39;main is sleeping&#39;
 time.sleep(2)

print &#39;main is setting event&#39;
 event.set()

#### TEST_SHAREDVALUES

def sharedvalues_func(values, arrays, shared_values, shared_arrays):
 for i in range(len(values)):
 v = values[i][1]
 sv = shared_values[i].value
 assert v == sv

for i in range(len(values)):
 a = arrays[i][1]
 sa = list(shared_arrays[i][:])
 assert a == sa

print &#39;Tests passed&#39;

def test_sharedvalues():
 values = [
 (&#39;i&#39;, 10),
 (&#39;h&#39;, -2),
 (&#39;d&#39;, 1.25)
 ]
 arrays = [
 (&#39;i&#39;, range(100)),
 (&#39;d&#39;, [0.25 * i for i in range(100)]),
 (&#39;H&#39;, range(1000))
 ]

shared_values = [multiprocessing.Value(id, v) for id, v in values]
 shared_arrays = [multiprocessing.Array(id, a) for id, a in arrays]

p = multiprocessing.Process(
 target=sharedvalues_func,
 args=(values, arrays, shared_values, shared_arrays)
 )
 p.start()
 p.join()

assert p.exitcode == 0

####

def test(namespace=multiprocessing):
 global multiprocessing

multiprocessing = namespace

for func in [ test_value, test_queue, test_condition,
 test_semaphore, test_join_timeout, test_event,
 test_sharedvalues ]:

print &#39;\n\t######## %s\n&#39; % func.__name__
 func()

ignore = multiprocessing.active_children() # cleanup any old processes
 if hasattr(multiprocessing, &#39;_debug_info&#39;):
 info = multiprocessing._debug_info()
 if info:
 print info
 raise ValueError(&#39;there should be no positive refcounts left&#39;)

if len(sys.argv) == 1 or sys.argv[1] == &#39;processes&#39;:
 print &#39; Using processes &#39;.center(79, &#39;-&#39;)
 namespace = multiprocessing
 elif sys.argv[1] == &#39;manager&#39;:
 print &#39; Using processes and a manager &#39;.center(79, &#39;-&#39;)
 namespace = multiprocessing.Manager()
 namespace.Process = multiprocessing.Process
 namespace.current_process = multiprocessing.current_process
 namespace.active_children = multiprocessing.active_children
 elif sys.argv[1] == &#39;threads&#39;:
 print &#39; Using threads &#39;.center(79, &#39;-&#39;)
 import multiprocessing.dummy as namespace
 else:
 print &#39;Usage:\n\t%s [processes | manager | threads]&#39; % sys.argv[0]
 raise SystemExit(2)

test(namespace)
</pre></div>
</div>
An example showing how to use queues to feed tasks to a collection of worker
processes and collect the results:
<div class="highlight-python notranslate"><div class="highlight"><pre>#
# Simple example which uses a pool of workers to carry out some tasks.
#
# Notice that the results will probably not come out of the output
# queue in the same in the same order as the corresponding tasks were
# put on the input queue. If it is important to get the results back
# in the original order then consider using `Pool.map()` or
# `Pool.imap()` (which will save on the amount of code needed anyway).
#
# Copyright (c) 2006-2008, R Oudkerk
# All rights reserved.
#

import time
import random

#
# Function run by worker processes
#

def worker(input, output):
 for func, args in iter(input.get, &#39;STOP&#39;):
 result = calculate(func, args)
 output.put(result)

#
# Function used to calculate result
#

def calculate(func, args):
 result = func(*args)
 return &#39;%s says that %s%s = %s&#39; % \
 (current_process().name, func.__name__, args, result)

#
# Functions referenced by tasks
#

#
#
#

def test():
 NUMBER_OF_PROCESSES = 4
 TASKS1 = [(mul, (i, 7)) for i in range(20)]
 TASKS2 = [(plus, (i, 8)) for i in range(10)]

# Create queues
 task_queue = Queue()
 done_queue = Queue()

# Submit tasks
 for task in TASKS1:
 task_queue.put(task)

# Start worker processes
 for i in range(NUMBER_OF_PROCESSES):
 Process(target=worker, args=(task_queue, done_queue)).start()

# Get and print results
 print &#39;Unordered results:&#39;
 for i in range(len(TASKS1)):
 print &#39;\t&#39;, done_queue.get()

# Add more tasks using `put()`
 for task in TASKS2:
 task_queue.put(task)

# Get and print some more results
 for i in range(len(TASKS2)):
 print &#39;\t&#39;, done_queue.get()

# Tell child processes to stop
 for i in range(NUMBER_OF_PROCESSES):
 task_queue.put(&#39;STOP&#39;)

import os
import sys

from multiprocessing import Process, current_process, freeze_support
from BaseHTTPServer import HTTPServer
from SimpleHTTPServer import SimpleHTTPRequestHandler

def note(format, *args):
 sys.stderr.write(&#39;[%s]\t%s\n&#39; % (current_process().name, format%args))

class RequestHandler(SimpleHTTPRequestHandler):
 # we override log_message() to show which process is handling the request
 def log_message(self, format, *args):
 note(format, *args)

def serve_forever(server):
 note(&#39;starting server&#39;)
 try:
 server.serve_forever()
 except KeyboardInterrupt:
 pass

def runpool(address, number_of_processes):
 # create a single server object -- children will each inherit a copy
 server = HTTPServer(address, RequestHandler)

# create child processes to act as workers
 for i in range(number_of_processes-1):
 Process(target=serve_forever, args=(server,)).start()

# main process also acts as a worker
 serve_forever(server)

def test():
 DIR = os.path.join(os.path.dirname(__file__), &#39;..&#39;)
 ADDRESS = (&#39;localhost&#39;, 8000)
 NUMBER_OF_PROCESSES = 4

print &#39;Serving at http://%s:%d using %d worker processes&#39; % \
 (ADDRESS[0], ADDRESS[1], NUMBER_OF_PROCESSES)
 print &#39;To exit press Ctrl-&#39; + [&#39;C&#39;, &#39;Break&#39;][sys.platform==&#39;win32&#39;]

os.chdir(DIR)
 runpool(ADDRESS, NUMBER_OF_PROCESSES)

import time, sys, multiprocessing, threading, Queue, gc

if sys.platform == &#39;win32&#39;:
 _timer = time.clock
else:
 _timer = time.time

delta = 1

#### TEST_QUEUESPEED

def queuespeed_func(q, c, iterations):
 a = &#39;0&#39; * 256
 c.acquire()
 c.notify()
 c.release()

for i in xrange(iterations):
 q.put(a)

q.put(&#39;STOP&#39;)

def test_queuespeed(Process, q, c):
 elapsed = 0
 iterations = 1

while elapsed &lt; delta:
 iterations *= 2

p = Process(target=queuespeed_func, args=(q, c, iterations))
 c.acquire()
 p.start()
 c.wait()
 c.release()

result = None
 t = _timer()

while result != &#39;STOP&#39;:
 result = q.get()

elapsed = _timer() - t

p.join()

print iterations, &#39;objects passed through the queue in&#39;, elapsed, &#39;seconds&#39;
 print &#39;average number/sec:&#39;, iterations/elapsed

#### TEST_PIPESPEED

def pipe_func(c, cond, iterations):
 a = &#39;0&#39; * 256
 cond.acquire()
 cond.notify()
 cond.release()

for i in xrange(iterations):
 c.send(a)

c.send(&#39;STOP&#39;)

def test_pipespeed():
 c, d = multiprocessing.Pipe()
 cond = multiprocessing.Condition()
 elapsed = 0
 iterations = 1

p = multiprocessing.Process(target=pipe_func,
 args=(d, cond, iterations))
 cond.acquire()
 p.start()
 cond.wait()
 cond.release()

result = None
 t = _timer()

while result != &#39;STOP&#39;:
 result = c.recv()

elapsed = _timer() - t
 p.join()

print iterations, &#39;objects passed through connection in&#39;,elapsed,&#39;seconds&#39;
 print &#39;average number/sec:&#39;, iterations/elapsed

#### TEST_SEQSPEED

def test_seqspeed(seq):
 elapsed = 0
 iterations = 1

t = _timer()

for i in xrange(iterations):
 a = seq[5]

elapsed = _timer()-t

print iterations, &#39;iterations in&#39;, elapsed, &#39;seconds&#39;
 print &#39;average number/sec:&#39;, iterations/elapsed

#### TEST_LOCK

def test_lockspeed(l):
 elapsed = 0
 iterations = 1

t = _timer()

for i in xrange(iterations):
 l.acquire()
 l.release()

elapsed = _timer()-t

#### TEST_CONDITION

def conditionspeed_func(c, N):
 c.acquire()
 c.notify()

for i in xrange(N):
 c.wait()
 c.notify()

c.release()

def test_conditionspeed(Process, c):
 elapsed = 0
 iterations = 1

c.acquire()
 p = Process(target=conditionspeed_func, args=(c, iterations))
 p.start()

c.wait()

t = _timer()

for i in xrange(iterations):
 c.notify()
 c.wait()

elapsed = _timer()-t

c.release()
 p.join()

print iterations * 2, &#39;waits in&#39;, elapsed, &#39;seconds&#39;
 print &#39;average number/sec:&#39;, iterations * 2 / elapsed

####

def test():
 manager = multiprocessing.Manager()

gc.disable()

print &#39;\n\t######## testing Queue.Queue\n&#39;
 test_queuespeed(threading.Thread, Queue.Queue(),
 threading.Condition())
 print &#39;\n\t######## testing multiprocessing.Queue\n&#39;
 test_queuespeed(multiprocessing.Process, multiprocessing.Queue(),
 multiprocessing.Condition())
 print &#39;\n\t######## testing Queue managed by server process\n&#39;
 test_queuespeed(multiprocessing.Process, manager.Queue(),
 manager.Condition())
 print &#39;\n\t######## testing multiprocessing.Pipe\n&#39;
 test_pipespeed()

print

print &#39;\n\t######## testing list\n&#39;
 test_seqspeed(range(10))
 print &#39;\n\t######## testing list managed by server process\n&#39;
 test_seqspeed(manager.list(range(10)))
 print &#39;\n\t######## testing Array(&quot;i&quot;, ..., lock=False)\n&#39;
 test_seqspeed(multiprocessing.Array(&#39;i&#39;, range(10), lock=False))
 print &#39;\n\t######## testing Array(&quot;i&quot;, ..., lock=True)\n&#39;
 test_seqspeed(multiprocessing.Array(&#39;i&#39;, range(10), lock=True))

print

print &#39;\n\t######## testing threading.Lock\n&#39;
 test_lockspeed(threading.Lock())
 print &#39;\n\t######## testing threading.RLock\n&#39;
 test_lockspeed(threading.RLock())
 print &#39;\n\t######## testing multiprocessing.Lock\n&#39;
 test_lockspeed(multiprocessing.Lock())
 print &#39;\n\t######## testing multiprocessing.RLock\n&#39;
 test_lockspeed(multiprocessing.RLock())
 print &#39;\n\t######## testing lock managed by server process\n&#39;
 test_lockspeed(manager.Lock())
 print &#39;\n\t######## testing rlock managed by server process\n&#39;
 test_lockspeed(manager.RLock())

print

print &#39;\n\t######## testing threading.Condition\n&#39;
 test_conditionspeed(threading.Thread, threading.Condition())
 print &#39;\n\t######## testing multiprocessing.Condition\n&#39;
 test_conditionspeed(multiprocessing.Process, multiprocessing.Condition())
 print &#39;\n\t######## testing condition managed by a server process\n&#39;
 test_conditionspeed(multiprocessing.Process, manager.Condition())

gc.enable()

</div>
 </div>
 </div>
 <div class="sphinxsidebar" role="navigation" aria-label="main navigation">
 <div class="sphinxsidebarwrapper">
 <h3><a href="../contents.html">Table Of Contents</a></h3>
 <ul>
<li><a class="reference internal" href="#">16.6. <code class="docutils literal notranslate">multiprocessing</code> — Process-based “threading” interface</a><ul>
<li><a class="reference internal" href="#introduction">16.6.1. Introduction</a><ul>
<li><a class="reference internal" href="#the-process-class">16.6.1.1. The <code class="docutils literal notranslate">Process</code> class</a></li>
<li><a class="reference internal" href="#exchanging-objects-between-processes">16.6.1.2. Exchanging objects between processes</a></li>
<li><a class="reference internal" href="#synchronization-between-processes">16.6.1.3. Synchronization between processes</a></li>
<li><a class="reference internal" href="#sharing-state-between-processes">16.6.1.4. Sharing state between processes</a></li>
<li><a class="reference internal" href="#using-a-pool-of-workers">16.6.1.5. Using a pool of workers</a></li>
</ul>
</li>
<li><a class="reference internal" href="#reference">16.6.2. Reference</a><ul>
<li><a class="reference internal" href="#process-and-exceptions">16.6.2.1. <code class="docutils literal notranslate">Process</code> and exceptions</a></li>
<li><a class="reference internal" href="#pipes-and-queues">16.6.2.2. Pipes and Queues</a></li>
<li><a class="reference internal" href="#miscellaneous">16.6.2.3. Miscellaneous</a></li>
<li><a class="reference internal" href="#connection-objects">16.6.2.4. Connection Objects</a></li>
<li><a class="reference internal" href="#synchronization-primitives">16.6.2.5. Synchronization primitives</a></li>
<li><a class="reference internal" href="#shared-ctypes-objects">16.6.2.6. Shared <code class="docutils literal notranslate">ctypes</code> Objects</a><ul>
<li><a class="reference internal" href="#module-multiprocessing.sharedctypes">16.6.2.6.1. The <code class="docutils literal notranslate">multiprocessing.sharedctypes</code> module</a></li>
</ul>
</li>
<li><a class="reference internal" href="#managers">16.6.2.7. Managers</a><ul>
<li><a class="reference internal" href="#customized-managers">16.6.2.7.1. Customized managers</a></li>
<li><a class="reference internal" href="#using-a-remote-manager">16.6.2.7.2. Using a remote manager</a></li>
</ul>
</li>
<li><a class="reference internal" href="#proxy-objects">16.6.2.8. Proxy Objects</a><ul>
<li><a class="reference internal" href="#cleanup">16.6.2.8.1. Cleanup</a></li>
</ul>
</li>
<li><a class="reference internal" href="#module-multiprocessing.pool">16.6.2.9. Process Pools</a></li>
<li><a class="reference internal" href="#module-multiprocessing.connection">16.6.2.10. Listeners and Clients</a><ul>
<li><a class="reference internal" href="#address-formats">16.6.2.10.1. Address Formats</a></li>
</ul>
</li>
<li><a class="reference internal" href="#authentication-keys">16.6.2.11. Authentication keys</a></li>
<li><a class="reference internal" href="#logging">16.6.2.12. Logging</a></li>
<li><a class="reference internal" href="#module-multiprocessing.dummy">16.6.2.13. The <code class="docutils literal notranslate">multiprocessing.dummy</code> module</a></li>
</ul>
</li>
<li><a class="reference internal" href="#programming-guidelines">16.6.3. Programming guidelines</a><ul>
<li><a class="reference internal" href="#all-platforms">16.6.3.1. All platforms</a></li>
<li><a class="reference internal" href="#windows">16.6.3.2. Windows</a></li>
</ul>
</li>
<li><a class="reference internal" href="#examples">16.6.4. Examples</a></li>
</ul>
</li>
</ul>

<h4>Previous topic</h4>
 <a href="dummy_thread.html"
 title="previous chapter">16.5. <code class="docutils literal notranslate">dummy_thread</code> — Drop-in replacement for the <code class="docutils literal notranslate">thread</code> module</a>
 <h4>Next topic</h4>
 <a href="mmap.html"
 title="next chapter">16.7. <code class="docutils literal notranslate">mmap</code> — Memory-mapped file support</a>
 <div role="note" aria-label="source link">
 <h3>This Page</h3>
 <ul class="this-page-menu">
 <li><a href="../_sources/library/multiprocessing.rst.txt"
 rel="nofollow">Show Source</a></li>
 </ul>
 </div>
<div id="searchbox" style="display: none" role="search">
 <h3>Quick search</h3>
 <div class="searchformwrapper">
 <form class="search" action="../search.html" method="get">
 <input type="text" name="q" />
 <input type="submit" value="Go" />
 <input type="hidden" name="check_keywords" value="yes" />
 <input type="hidden" name="area" value="default" />
 </form>
 </div>
</div>
<script type="text/javascript">$('#searchbox').show(0);</script>
 </div>
 </div>
 <div class="clearer"></div>
 </div> 
 <div class="related" role="navigation" aria-label="related navigation">
 <h3>Navigation</h3>
 <ul>
 <li class="right" style="margin-right: 10px">
 <a href="../genindex.html" title="General Index"
 >index</a></li>
 <li class="right" >
 <a href="../py-modindex.html" title="Python Module Index"
 >modules</a> |</li>
 <li class="right" >
 <a href="mmap.html" title="16.7. mmap — Memory-mapped file support"
 >next</a> |</li>
 <li class="right" >
 <a href="dummy_thread.html" title="16.5. dummy_thread — Drop-in replacement for the thread module"
 >previous</a> |</li>
 <li><img src="../_static/py.png" alt=""
 style="vertical-align: middle; margin-top: -1px"/></li>
 <li><a href="https://www.python.org/">Python</a> &#187;</li>
 <li>
 <a href="../index.html">Python 2.7.16 documentation</a> &#187;
 </li>

<li class="nav-item nav-item-1"><a href="index.html" >The Python Standard Library</a> &#187;</li>
 <li class="nav-item nav-item-2"><a href="someos.html" >16. Optional Operating System Services</a> &#187;</li> 
 </ul>
 </div> 
 <div class="footer">
 &copy; <a href="../copyright.html">Copyright</a> 1990-2019, Python Software Foundation.
 
 The Python Software Foundation is a non-profit corporation.
 <a href="https://www.python.org/psf/donations/">Please donate.</a>
 
 Last updated on Mar 27, 2019.
 <a href="../bugs.html">Found a bug</a>?
 
 Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.7.6.
 </div>

</body>
</html>

MR SKB