| |
| |
| |
| <!DOCTYPE html> |
| <html class="writer-html5" lang="en" > |
| <head> |
| <meta charset="utf-8"> |
| |
| <meta name="viewport" content="width=device-width, initial-scale=1.0"> |
| |
| <title>The Plasma In-Memory Object Store — Apache Arrow v2.0.0</title> |
| |
| |
| |
| <link rel="stylesheet" href="../_static/css/theme.css" type="text/css" /> |
| <link rel="stylesheet" href="../_static/pygments.css" type="text/css" /> |
| |
| |
| |
| |
| |
| |
| |
| <!--[if lt IE 9]> |
| <script src="../_static/js/html5shiv.min.js"></script> |
| <![endif]--> |
| |
| |
| <script type="text/javascript" id="documentation_options" data-url_root="../" src="../_static/documentation_options.js"></script> |
| <script src="../_static/jquery.js"></script> |
| <script src="../_static/underscore.js"></script> |
| <script src="../_static/doctools.js"></script> |
| <script src="../_static/language_data.js"></script> |
| |
| <script type="text/javascript" src="../_static/js/theme.js"></script> |
| |
| |
| <link rel="canonical" href="https://arrow.apache.org/docs/python/plasma.html" /> |
| <link rel="index" title="Index" href="../genindex.html" /> |
| <link rel="search" title="Search" href="../search.html" /> |
| <link rel="next" title="NumPy Integration" href="numpy.html" /> |
| <link rel="prev" title="pyarrow.HdfsFile" href="generated/pyarrow.HdfsFile.html" /> |
|
|
|
|
| <!-- Matomo -->
|
| <script>
|
| var _paq = window._paq = window._paq || [];
|
| /* tracker methods like "setCustomDimension" should be called before "trackPageView" */
|
| _paq.push(["setDoNotTrack", true]);
|
| _paq.push(["disableCookies"]);
|
| _paq.push(['trackPageView']);
|
| _paq.push(['enableLinkTracking']);
|
| (function() {
|
| var u="https://analytics.apache.org/";
|
| _paq.push(['setTrackerUrl', u+'matomo.php']);
|
| _paq.push(['setSiteId', '20']);
|
| var d=document, g=d.createElement('script'), s=d.getElementsByTagName('script')[0];
|
| g.async=true; g.src=u+'matomo.js'; s.parentNode.insertBefore(g,s);
|
| })();
|
| </script>
|
| <!-- End Matomo Code -->
|
|
|
| </head> |
| |
| <body class="wy-body-for-nav"> |
| |
| |
| <div class="wy-grid-for-nav"> |
| |
| <nav data-toggle="wy-nav-shift" class="wy-nav-side"> |
| <div class="wy-side-scroll"> |
| <div class="wy-side-nav-search" > |
| |
| |
| |
| <a href="../index.html" class="icon icon-home" alt="Documentation Home"> Apache Arrow |
| |
| |
| |
| </a> |
| |
| |
| |
| |
| <div class="version"> |
| 2.0.0 |
| </div> |
| |
| |
| |
| |
| <div role="search"> |
| <form id="rtd-search-form" class="wy-form" action="../search.html" method="get"> |
| <input type="text" name="q" placeholder="Search docs" /> |
| <input type="hidden" name="check_keywords" value="yes" /> |
| <input type="hidden" name="area" value="default" /> |
| </form> |
| </div> |
| |
| |
| </div> |
| |
| |
| <div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation"> |
| |
| |
| |
| |
| |
| |
| <p class="caption"><span class="caption-text">Specifications and Protocols</span></p> |
| <ul> |
| <li class="toctree-l1"><a class="reference internal" href="../format/Versioning.html">Format Versioning and Stability</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../format/Columnar.html">Arrow Columnar Format</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../format/Flight.html">Arrow Flight RPC</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../format/Integration.html">Integration Testing</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../format/CDataInterface.html">The Arrow C data interface</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../format/CStreamInterface.html">The Arrow C stream interface</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../format/Other.html">Other Data Structures</a></li> |
| </ul> |
| <p class="caption"><span class="caption-text">Libraries</span></p> |
| <ul class="current"> |
| <li class="toctree-l1"><a class="reference internal" href="../status.html">Implementation Status</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://arrow.apache.org/docs/c_glib/">C/GLib</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../cpp/index.html">C++</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://github.com/apache/arrow/blob/master/csharp/README.md">C#</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://godoc.org/github.com/apache/arrow/go/arrow">Go</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../java/index.html">Java</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://arrow.apache.org/docs/js/">JavaScript</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://github.com/apache/arrow/blob/master/matlab/README.md">MATLAB</a></li> |
| <li class="toctree-l1 current"><a class="reference internal" href="index.html">Python</a><ul class="current"> |
| <li class="toctree-l2"><a class="reference internal" href="install.html">Installing PyArrow</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="memory.html">Memory and IO Interfaces</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="data.html">Data Types and In-Memory Data Model</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="compute.html">Compute Functions</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="ipc.html">Streaming, Serialization, and IPC</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="filesystems.html">Filesystem Interface</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="filesystems_deprecated.html">Filesystem Interface (legacy)</a></li> |
| <li class="toctree-l2 current"><a class="current reference internal" href="#">The Plasma In-Memory Object Store</a><ul> |
| <li class="toctree-l3"><a class="reference internal" href="#the-plasma-api">The Plasma API</a><ul> |
| <li class="toctree-l4"><a class="reference internal" href="#starting-the-plasma-store">Starting the Plasma store</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#creating-a-plasma-client">Creating a Plasma client</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#object-ids">Object IDs</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#putting-and-getting-python-objects">Putting and Getting Python Objects</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#creating-an-object-buffer">Creating an Object Buffer</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#getting-an-object-buffer">Getting an Object Buffer</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#listing-objects-in-the-store">Listing objects in the store</a></li> |
| </ul> |
| </li> |
| <li class="toctree-l3"><a class="reference internal" href="#using-arrow-and-pandas-with-plasma">Using Arrow and Pandas with Plasma</a><ul> |
| <li class="toctree-l4"><a class="reference internal" href="#storing-arrow-objects-in-plasma">Storing Arrow Objects in Plasma</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#getting-arrow-objects-from-plasma">Getting Arrow Objects from Plasma</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#storing-pandas-dataframes-in-plasma">Storing Pandas DataFrames in Plasma</a></li> |
| <li class="toctree-l4"><a class="reference internal" href="#getting-pandas-dataframes-from-plasma">Getting Pandas DataFrames from Plasma</a></li> |
| </ul> |
| </li> |
| <li class="toctree-l3"><a class="reference internal" href="#using-plasma-with-huge-pages">Using Plasma with Huge Pages</a></li> |
| </ul> |
| </li> |
| <li class="toctree-l2"><a class="reference internal" href="numpy.html">NumPy Integration</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="pandas.html">Pandas Integration</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="timestamps.html">Timestamps</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="csv.html">Reading CSV files</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="feather.html">Feather File Format</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="json.html">Reading JSON files</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="parquet.html">Reading and Writing the Apache Parquet Format</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="dataset.html">Tabular Datasets</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="cuda.html">CUDA Integration</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="extending_types.html">Extending pyarrow</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="extending.html">Using pyarrow from C++ and Cython Code</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="api.html">API Reference</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="getting_involved.html">Getting Involved</a></li> |
| <li class="toctree-l2"><a class="reference internal" href="benchmarks.html">Benchmarks</a></li> |
| </ul> |
| </li> |
| <li class="toctree-l1"><a class="reference external" href="https://arrow.apache.org/docs/r/">R</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://github.com/apache/arrow/blob/master/ruby/README.md">Ruby</a></li> |
| <li class="toctree-l1"><a class="reference external" href="https://docs.rs/crate/arrow/">Rust</a></li> |
| </ul> |
| <p class="caption"><span class="caption-text">Development</span></p> |
| <ul> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/contributing.html">Contributing to Apache Arrow</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/cpp/index.html">C++ Development</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/python.html">Python Development</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/archery.html">Daily Development using Archery</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/crossbow.html">Packaging and Testing with Crossbow</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/docker.html">Running Docker Builds</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/benchmarks.html">Benchmarks</a></li> |
| <li class="toctree-l1"><a class="reference internal" href="../developers/documentation.html">Building the Documentation</a></li> |
| </ul> |
| |
| |
| |
| </div> |
| |
| </div> |
| </nav> |
| |
| <section data-toggle="wy-nav-shift" class="wy-nav-content-wrap"> |
| |
| |
| <nav class="wy-nav-top" aria-label="top navigation"> |
| |
| <i data-toggle="wy-nav-top" class="fa fa-bars"></i> |
| <a href="../index.html">Apache Arrow</a> |
| |
| </nav> |
| |
| |
| <div class="wy-nav-content"> |
| |
| <div class="rst-content"> |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| <div role="navigation" aria-label="breadcrumbs navigation"> |
| |
| <ul class="wy-breadcrumbs"> |
| |
| <li><a href="../index.html" class="icon icon-home"></a> »</li> |
| |
| <li><a href="index.html">Python bindings</a> »</li> |
| |
| <li>The Plasma In-Memory Object Store</li> |
| |
| |
| <li class="wy-breadcrumbs-aside"> |
| |
| |
| <a href="../_sources/python/plasma.rst.txt" rel="nofollow"> View page source</a> |
| |
| |
| </li> |
| |
| </ul> |
| |
| |
| <hr/> |
| </div> |
| <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article"> |
| <div itemprop="articleBody"> |
| |
| <div class="section" id="the-plasma-in-memory-object-store"> |
| <span id="plasma"></span><h1><a class="toc-backref" href="#id1">The Plasma In-Memory Object Store</a><a class="headerlink" href="#the-plasma-in-memory-object-store" title="Permalink to this headline">¶</a></h1> |
| <div class="contents topic" id="contents"> |
| <p class="topic-title">Contents</p> |
| <ul class="simple"> |
| <li><p><a class="reference internal" href="#the-plasma-in-memory-object-store" id="id1">The Plasma In-Memory Object Store</a></p> |
| <ul> |
| <li><p><a class="reference internal" href="#the-plasma-api" id="id2">The Plasma API</a></p> |
| <ul> |
| <li><p><a class="reference internal" href="#starting-the-plasma-store" id="id3">Starting the Plasma store</a></p></li> |
| <li><p><a class="reference internal" href="#creating-a-plasma-client" id="id4">Creating a Plasma client</a></p></li> |
| <li><p><a class="reference internal" href="#object-ids" id="id5">Object IDs</a></p></li> |
| <li><p><a class="reference internal" href="#putting-and-getting-python-objects" id="id6">Putting and Getting Python Objects</a></p></li> |
| <li><p><a class="reference internal" href="#creating-an-object-buffer" id="id7">Creating an Object Buffer</a></p></li> |
| <li><p><a class="reference internal" href="#getting-an-object-buffer" id="id8">Getting an Object Buffer</a></p></li> |
| <li><p><a class="reference internal" href="#listing-objects-in-the-store" id="id9">Listing objects in the store</a></p></li> |
| </ul> |
| </li> |
| <li><p><a class="reference internal" href="#using-arrow-and-pandas-with-plasma" id="id10">Using Arrow and Pandas with Plasma</a></p> |
| <ul> |
| <li><p><a class="reference internal" href="#storing-arrow-objects-in-plasma" id="id11">Storing Arrow Objects in Plasma</a></p></li> |
| <li><p><a class="reference internal" href="#getting-arrow-objects-from-plasma" id="id12">Getting Arrow Objects from Plasma</a></p></li> |
| <li><p><a class="reference internal" href="#storing-pandas-dataframes-in-plasma" id="id13">Storing Pandas DataFrames in Plasma</a></p></li> |
| <li><p><a class="reference internal" href="#getting-pandas-dataframes-from-plasma" id="id14">Getting Pandas DataFrames from Plasma</a></p></li> |
| </ul> |
| </li> |
| <li><p><a class="reference internal" href="#using-plasma-with-huge-pages" id="id15">Using Plasma with Huge Pages</a></p></li> |
| </ul> |
| </li> |
| </ul> |
| </div> |
| <div class="admonition note"> |
| <p class="admonition-title">Note</p> |
| <p>As present, Plasma is only supported for use on Linux and macOS.</p> |
| </div> |
| <div class="section" id="the-plasma-api"> |
| <h2><a class="toc-backref" href="#id2">The Plasma API</a><a class="headerlink" href="#the-plasma-api" title="Permalink to this headline">¶</a></h2> |
| <div class="section" id="starting-the-plasma-store"> |
| <h3><a class="toc-backref" href="#id3">Starting the Plasma store</a><a class="headerlink" href="#starting-the-plasma-store" title="Permalink to this headline">¶</a></h3> |
| <p>You can start the Plasma store by issuing a terminal command similar to the |
| following:</p> |
| <div class="highlight-bash notranslate"><div class="highlight"><pre><span></span>plasma_store -m <span class="m">1000000000</span> -s /tmp/plasma |
| </pre></div> |
| </div> |
| <p>The <code class="docutils literal notranslate"><span class="pre">-m</span></code> flag specifies the size of the store in bytes, and the <code class="docutils literal notranslate"><span class="pre">-s</span></code> flag |
| specifies the socket that the store will listen at. Thus, the above command |
| allows the Plasma store to use up to 1GB of memory, and sets the socket to |
| <code class="docutils literal notranslate"><span class="pre">/tmp/plasma</span></code>.</p> |
| <p>Leaving the current terminal window open as long as Plasma store should keep |
| running. Messages, concerning such as disconnecting clients, may occasionally be |
| printed to the screen. To stop running the Plasma store, you can press |
| <code class="docutils literal notranslate"><span class="pre">Ctrl-C</span></code> in the terminal.</p> |
| </div> |
| <div class="section" id="creating-a-plasma-client"> |
| <h3><a class="toc-backref" href="#id4">Creating a Plasma client</a><a class="headerlink" href="#creating-a-plasma-client" title="Permalink to this headline">¶</a></h3> |
| <p>To start a Plasma client from Python, call <code class="docutils literal notranslate"><span class="pre">plasma.connect</span></code> using the same |
| socket name:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">pyarrow.plasma</span> <span class="kn">as</span> <span class="nn">plasma</span> |
| <span class="n">client</span> <span class="o">=</span> <span class="n">plasma</span><span class="o">.</span><span class="n">connect</span><span class="p">(</span><span class="s2">"/tmp/plasma"</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>If the following error occurs from running the above Python code, that |
| means that either the socket given is incorrect, or the <code class="docutils literal notranslate"><span class="pre">./plasma_store</span></code> is |
| not currently running. Check to see if the Plasma store is still running.</p> |
| <div class="highlight-shell notranslate"><div class="highlight"><pre><span></span>>>> <span class="nv">client</span> <span class="o">=</span> plasma.connect<span class="o">(</span><span class="s2">"/tmp/plasma"</span><span class="o">)</span> |
| Connection to socket failed <span class="k">for</span> pathname /tmp/plasma |
| Could not connect to socket /tmp/plasma |
| </pre></div> |
| </div> |
| </div> |
| <div class="section" id="object-ids"> |
| <h3><a class="toc-backref" href="#id5">Object IDs</a><a class="headerlink" href="#object-ids" title="Permalink to this headline">¶</a></h3> |
| <p>Each object in the Plasma store should be associated with a unique ID. The |
| Object ID then serves as a key that any client can use to retrieve that object |
| from the Plasma store. You can form an <code class="docutils literal notranslate"><span class="pre">ObjectID</span></code> object from a byte string of |
| length 20.</p> |
| <div class="highlight-shell notranslate"><div class="highlight"><pre><span></span><span class="c1"># Create an ObjectID.</span> |
| >>> <span class="nv">id</span> <span class="o">=</span> plasma.ObjectID<span class="o">(</span><span class="m">20</span> * b<span class="s2">"a"</span><span class="o">)</span> |
| |
| <span class="c1"># The character "a" is encoded as 61 in hex.</span> |
| >>> id |
| ObjectID<span class="o">(</span><span class="m">6161616161616161616161616161616161616161</span><span class="o">)</span> |
| </pre></div> |
| </div> |
| <p>The random generation of Object IDs is often good enough to ensure unique IDs. |
| You can easily create a helper function that randomly generates object IDs as |
| follows:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">numpy</span> <span class="kn">as</span> <span class="nn">np</span> |
| |
| <span class="k">def</span> <span class="nf">random_object_id</span><span class="p">():</span> |
| <span class="k">return</span> <span class="n">plasma</span><span class="o">.</span><span class="n">ObjectID</span><span class="p">(</span><span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">bytes</span><span class="p">(</span><span class="mi">20</span><span class="p">))</span> |
| </pre></div> |
| </div> |
| </div> |
| <div class="section" id="putting-and-getting-python-objects"> |
| <h3><a class="toc-backref" href="#id6">Putting and Getting Python Objects</a><a class="headerlink" href="#putting-and-getting-python-objects" title="Permalink to this headline">¶</a></h3> |
| <p>Plasma supports two APIs for creating and accessing objects: A high level |
| API that allows storing and retrieving Python objects and a low level |
| API that allows creating, writing and sealing buffers and operating on |
| the binary data directly. In this section we describe the high level API.</p> |
| <p>This is how you can put and get a Python object:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Create a python object.</span> |
| <span class="n">object_id</span> <span class="o">=</span> <span class="n">client</span><span class="o">.</span><span class="n">put</span><span class="p">(</span><span class="s2">"hello, world"</span><span class="p">)</span> |
| |
| <span class="c1"># Get the object.</span> |
| <span class="n">client</span><span class="o">.</span><span class="n">get</span><span class="p">(</span><span class="n">object_id</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>This works with all Python objects supported by the Arrow Python object |
| serialization.</p> |
| <p>You can also get multiple objects at the same time (which can be more |
| efficient since it avoids IPC round trips):</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Create multiple python objects.</span> |
| <span class="n">object_id1</span> <span class="o">=</span> <span class="n">client</span><span class="o">.</span><span class="n">put</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span> |
| <span class="n">object_id2</span> <span class="o">=</span> <span class="n">client</span><span class="o">.</span><span class="n">put</span><span class="p">(</span><span class="mi">2</span><span class="p">)</span> |
| <span class="n">object_id3</span> <span class="o">=</span> <span class="n">client</span><span class="o">.</span><span class="n">put</span><span class="p">(</span><span class="mi">3</span><span class="p">)</span> |
| |
| <span class="c1"># Get the objects.</span> |
| <span class="n">client</span><span class="o">.</span><span class="n">get</span><span class="p">([</span><span class="n">object_id1</span><span class="p">,</span> <span class="n">object_id2</span><span class="p">,</span> <span class="n">object_id3</span><span class="p">])</span> |
| </pre></div> |
| </div> |
| <p>Furthermore, it is possible to provide a timeout for the get call. If the |
| object is not available within the timeout, the special object |
| <cite>pyarrow.ObjectNotAvailable</cite> will be returned.</p> |
| </div> |
| <div class="section" id="creating-an-object-buffer"> |
| <h3><a class="toc-backref" href="#id7">Creating an Object Buffer</a><a class="headerlink" href="#creating-an-object-buffer" title="Permalink to this headline">¶</a></h3> |
| <p>Objects are created in Plasma in two stages. First, they are <strong>created</strong>, which |
| allocates a buffer for the object. At this point, the client can write to the |
| buffer and construct the object within the allocated buffer.</p> |
| <p>To create an object for Plasma, you need to create an object ID, as well as |
| give the object’s maximum size in bytes.</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Create an object buffer.</span> |
| <span class="n">object_id</span> <span class="o">=</span> <span class="n">plasma</span><span class="o">.</span><span class="n">ObjectID</span><span class="p">(</span><span class="mi">20</span> <span class="o">*</span> <span class="sa">b</span><span class="s2">"a"</span><span class="p">)</span> |
| <span class="n">object_size</span> <span class="o">=</span> <span class="mi">1000</span> |
| <span class="nb">buffer</span> <span class="o">=</span> <span class="n">memoryview</span><span class="p">(</span><span class="n">client</span><span class="o">.</span><span class="n">create</span><span class="p">(</span><span class="n">object_id</span><span class="p">,</span> <span class="n">object_size</span><span class="p">))</span> |
| |
| <span class="c1"># Write to the buffer.</span> |
| <span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">1000</span><span class="p">):</span> |
| <span class="nb">buffer</span><span class="p">[</span><span class="n">i</span><span class="p">]</span> <span class="o">=</span> <span class="n">i</span> <span class="o">%</span> <span class="mi">128</span> |
| </pre></div> |
| </div> |
| <p>When the client is done, the client <strong>seals</strong> the buffer, making the object |
| immutable, and making it available to other Plasma clients.</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Seal the object. This makes the object immutable and available to other clients.</span> |
| <span class="n">client</span><span class="o">.</span><span class="n">seal</span><span class="p">(</span><span class="n">object_id</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| </div> |
| <div class="section" id="getting-an-object-buffer"> |
| <h3><a class="toc-backref" href="#id8">Getting an Object Buffer</a><a class="headerlink" href="#getting-an-object-buffer" title="Permalink to this headline">¶</a></h3> |
| <p>After an object has been sealed, any client who knows the object ID can get |
| the object buffer.</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Create a different client. Note that this second client could be</span> |
| <span class="c1"># created in the same or in a separate, concurrent Python session.</span> |
| <span class="n">client2</span> <span class="o">=</span> <span class="n">plasma</span><span class="o">.</span><span class="n">connect</span><span class="p">(</span><span class="s2">"/tmp/plasma"</span><span class="p">)</span> |
| |
| <span class="c1"># Get the object in the second client. This blocks until the object has been sealed.</span> |
| <span class="n">object_id2</span> <span class="o">=</span> <span class="n">plasma</span><span class="o">.</span><span class="n">ObjectID</span><span class="p">(</span><span class="mi">20</span> <span class="o">*</span> <span class="sa">b</span><span class="s2">"a"</span><span class="p">)</span> |
| <span class="p">[</span><span class="n">buffer2</span><span class="p">]</span> <span class="o">=</span> <span class="n">client2</span><span class="o">.</span><span class="n">get_buffers</span><span class="p">([</span><span class="n">object_id</span><span class="p">])</span> |
| </pre></div> |
| </div> |
| <p>If the object has not been sealed yet, then the call to client.get_buffers will |
| block until the object has been sealed by the client constructing the object. |
| Using the <code class="docutils literal notranslate"><span class="pre">timeout_ms</span></code> argument to get, you can specify a timeout for this (in |
| milliseconds). After the timeout, the interpreter will yield control back.</p> |
| <div class="highlight-shell notranslate"><div class="highlight"><pre><span></span>>>> buffer |
| <memory at 0x7fdbdc96e708> |
| >>> buffer<span class="o">[</span><span class="m">1</span><span class="o">]</span> |
| <span class="m">1</span> |
| >>> buffer2 |
| <plasma.plasma.PlasmaBuffer object at 0x7fdbf2770e88> |
| >>> <span class="nv">view2</span> <span class="o">=</span> memoryview<span class="o">(</span>buffer2<span class="o">)</span> |
| >>> view2<span class="o">[</span><span class="m">1</span><span class="o">]</span> |
| <span class="m">1</span> |
| >>> view2<span class="o">[</span><span class="m">129</span><span class="o">]</span> |
| <span class="m">1</span> |
| >>> bytes<span class="o">(</span>buffer<span class="o">[</span><span class="m">1</span>:4<span class="o">])</span> |
| b<span class="s1">'\x01\x02\x03'</span> |
| >>> bytes<span class="o">(</span>view2<span class="o">[</span><span class="m">1</span>:4<span class="o">])</span> |
| b<span class="s1">'\x01\x02\x03'</span> |
| </pre></div> |
| </div> |
| </div> |
| <div class="section" id="listing-objects-in-the-store"> |
| <h3><a class="toc-backref" href="#id9">Listing objects in the store</a><a class="headerlink" href="#listing-objects-in-the-store" title="Permalink to this headline">¶</a></h3> |
| <p>The objects in the store can be listed in the following way (note that |
| this functionality is currently experimental and the concrete representation |
| of the object info might change in the future):</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">pyarrow.plasma</span> <span class="kn">as</span> <span class="nn">plasma</span> |
| <span class="kn">import</span> <span class="nn">time</span> |
| |
| <span class="n">client</span> <span class="o">=</span> <span class="n">plasma</span><span class="o">.</span><span class="n">connect</span><span class="p">(</span><span class="s2">"/tmp/plasma"</span><span class="p">)</span> |
| |
| <span class="n">client</span><span class="o">.</span><span class="n">put</span><span class="p">(</span><span class="s2">"hello, world"</span><span class="p">)</span> |
| <span class="c1"># Sleep a little so we get different creation times</span> |
| <span class="n">time</span><span class="o">.</span><span class="n">sleep</span><span class="p">(</span><span class="mi">2</span><span class="p">)</span> |
| <span class="n">client</span><span class="o">.</span><span class="n">put</span><span class="p">(</span><span class="s2">"another object"</span><span class="p">)</span> |
| <span class="c1"># Create an object that is not sealed yet</span> |
| <span class="n">object_id</span> <span class="o">=</span> <span class="n">plasma</span><span class="o">.</span><span class="n">ObjectID</span><span class="o">.</span><span class="n">from_random</span><span class="p">()</span> |
| <span class="n">client</span><span class="o">.</span><span class="n">create</span><span class="p">(</span><span class="n">object_id</span><span class="p">,</span> <span class="mi">100</span><span class="p">)</span> |
| <span class="k">print</span><span class="p">(</span><span class="n">client</span><span class="o">.</span><span class="n">list</span><span class="p">())</span> |
| |
| <span class="o">>>></span> <span class="p">{</span><span class="n">ObjectID</span><span class="p">(</span><span class="mi">4</span><span class="n">cba8f80c54c6d265b46c2cdfcee6e32348b12be</span><span class="p">):</span> <span class="p">{</span><span class="s1">'construct_duration'</span><span class="p">:</span> <span class="mi">0</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'create_time'</span><span class="p">:</span> <span class="mi">1535223642</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'data_size'</span><span class="p">:</span> <span class="mi">460</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'metadata_size'</span><span class="p">:</span> <span class="mi">0</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'ref_count'</span><span class="p">:</span> <span class="mi">0</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'state'</span><span class="p">:</span> <span class="s1">'sealed'</span><span class="p">},</span> |
| <span class="o">>>></span> <span class="n">ObjectID</span><span class="p">(</span><span class="n">a7598230b0c26464c9d9c99ae14773ee81485428</span><span class="p">):</span> <span class="p">{</span><span class="s1">'construct_duration'</span><span class="p">:</span> <span class="mi">0</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'create_time'</span><span class="p">:</span> <span class="mi">1535223644</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'data_size'</span><span class="p">:</span> <span class="mi">460</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'metadata_size'</span><span class="p">:</span> <span class="mi">0</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'ref_count'</span><span class="p">:</span> <span class="mi">0</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'state'</span><span class="p">:</span> <span class="s1">'sealed'</span><span class="p">},</span> |
| <span class="o">>>></span> <span class="n">ObjectID</span><span class="p">(</span><span class="n">e603ab0c92098ebf08f90bfcea33ff98f6476870</span><span class="p">):</span> <span class="p">{</span><span class="s1">'construct_duration'</span><span class="p">:</span> <span class="o">-</span><span class="mi">1</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'create_time'</span><span class="p">:</span> <span class="mi">1535223644</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'data_size'</span><span class="p">:</span> <span class="mi">100</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'metadata_size'</span><span class="p">:</span> <span class="mi">0</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'ref_count'</span><span class="p">:</span> <span class="mi">1</span><span class="p">,</span> |
| <span class="o">>>></span> <span class="s1">'state'</span><span class="p">:</span> <span class="s1">'created'</span><span class="p">}}</span> |
| </pre></div> |
| </div> |
| </div> |
| </div> |
| <div class="section" id="using-arrow-and-pandas-with-plasma"> |
| <h2><a class="toc-backref" href="#id10">Using Arrow and Pandas with Plasma</a><a class="headerlink" href="#using-arrow-and-pandas-with-plasma" title="Permalink to this headline">¶</a></h2> |
| <div class="section" id="storing-arrow-objects-in-plasma"> |
| <h3><a class="toc-backref" href="#id11">Storing Arrow Objects in Plasma</a><a class="headerlink" href="#storing-arrow-objects-in-plasma" title="Permalink to this headline">¶</a></h3> |
| <p>To store an Arrow object in Plasma, we must first <strong>create</strong> the object and then |
| <strong>seal</strong> it. However, Arrow objects such as <code class="docutils literal notranslate"><span class="pre">Tensors</span></code> may be more complicated |
| to write than simple binary data.</p> |
| <p>To create the object in Plasma, you still need an <code class="docutils literal notranslate"><span class="pre">ObjectID</span></code> and a size to |
| pass in. To find out the size of your Arrow object, you can use pyarrow |
| API such as <code class="docutils literal notranslate"><span class="pre">pyarrow.ipc.get_tensor_size</span></code>.</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">numpy</span> <span class="kn">as</span> <span class="nn">np</span> |
| <span class="kn">import</span> <span class="nn">pyarrow</span> <span class="kn">as</span> <span class="nn">pa</span> |
| |
| <span class="c1"># Create a pyarrow.Tensor object from a numpy random 2-dimensional array</span> |
| <span class="n">data</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">10</span><span class="p">,</span> <span class="mi">4</span><span class="p">)</span> |
| <span class="n">tensor</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">Tensor</span><span class="o">.</span><span class="n">from_numpy</span><span class="p">(</span><span class="n">data</span><span class="p">)</span> |
| |
| <span class="c1"># Create the object in Plasma</span> |
| <span class="n">object_id</span> <span class="o">=</span> <span class="n">plasma</span><span class="o">.</span><span class="n">ObjectID</span><span class="p">(</span><span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">bytes</span><span class="p">(</span><span class="mi">20</span><span class="p">))</span> |
| <span class="n">data_size</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">ipc</span><span class="o">.</span><span class="n">get_tensor_size</span><span class="p">(</span><span class="n">tensor</span><span class="p">)</span> |
| <span class="n">buf</span> <span class="o">=</span> <span class="n">client</span><span class="o">.</span><span class="n">create</span><span class="p">(</span><span class="n">object_id</span><span class="p">,</span> <span class="n">data_size</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>To write the Arrow <code class="docutils literal notranslate"><span class="pre">Tensor</span></code> object into the buffer, you can use Plasma to |
| convert the <code class="docutils literal notranslate"><span class="pre">memoryview</span></code> buffer into a <code class="docutils literal notranslate"><span class="pre">pyarrow.FixedSizeBufferWriter</span></code> |
| object. A <code class="docutils literal notranslate"><span class="pre">pyarrow.FixedSizeBufferWriter</span></code> is a format suitable for Arrow’s |
| <code class="docutils literal notranslate"><span class="pre">pyarrow.ipc.write_tensor</span></code>:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Write the tensor into the Plasma-allocated buffer</span> |
| <span class="n">stream</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">FixedSizeBufferWriter</span><span class="p">(</span><span class="n">buf</span><span class="p">)</span> |
| <span class="n">pa</span><span class="o">.</span><span class="n">ipc</span><span class="o">.</span><span class="n">write_tensor</span><span class="p">(</span><span class="n">tensor</span><span class="p">,</span> <span class="n">stream</span><span class="p">)</span> <span class="c1"># Writes tensor's 552 bytes to Plasma stream</span> |
| </pre></div> |
| </div> |
| <p>To finish storing the Arrow object in Plasma, call <code class="docutils literal notranslate"><span class="pre">seal</span></code>:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Seal the Plasma object</span> |
| <span class="n">client</span><span class="o">.</span><span class="n">seal</span><span class="p">(</span><span class="n">object_id</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| </div> |
| <div class="section" id="getting-arrow-objects-from-plasma"> |
| <h3><a class="toc-backref" href="#id12">Getting Arrow Objects from Plasma</a><a class="headerlink" href="#getting-arrow-objects-from-plasma" title="Permalink to this headline">¶</a></h3> |
| <p>To read the object, first retrieve it as a <code class="docutils literal notranslate"><span class="pre">PlasmaBuffer</span></code> using its object ID.</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Get the arrow object by ObjectID.</span> |
| <span class="p">[</span><span class="n">buf2</span><span class="p">]</span> <span class="o">=</span> <span class="n">client</span><span class="o">.</span><span class="n">get_buffers</span><span class="p">([</span><span class="n">object_id</span><span class="p">])</span> |
| </pre></div> |
| </div> |
| <p>To convert the <code class="docutils literal notranslate"><span class="pre">PlasmaBuffer</span></code> back into an Arrow <code class="docutils literal notranslate"><span class="pre">Tensor</span></code>, first create a |
| pyarrow <code class="docutils literal notranslate"><span class="pre">BufferReader</span></code> object from it. You can then pass the <code class="docutils literal notranslate"><span class="pre">BufferReader</span></code> |
| into <code class="docutils literal notranslate"><span class="pre">pyarrow.ipc.read_tensor</span></code> to reconstruct the Arrow <code class="docutils literal notranslate"><span class="pre">Tensor</span></code> object:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Reconstruct the Arrow tensor object.</span> |
| <span class="n">reader</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">BufferReader</span><span class="p">(</span><span class="n">buf2</span><span class="p">)</span> |
| <span class="n">tensor2</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">ipc</span><span class="o">.</span><span class="n">read_tensor</span><span class="p">(</span><span class="n">reader</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>Finally, you can use <code class="docutils literal notranslate"><span class="pre">pyarrow.ipc.read_tensor</span></code> to convert the Arrow object |
| back into numpy data:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Convert back to numpy</span> |
| <span class="n">array</span> <span class="o">=</span> <span class="n">tensor2</span><span class="o">.</span><span class="n">to_numpy</span><span class="p">()</span> |
| </pre></div> |
| </div> |
| </div> |
| <div class="section" id="storing-pandas-dataframes-in-plasma"> |
| <h3><a class="toc-backref" href="#id13">Storing Pandas DataFrames in Plasma</a><a class="headerlink" href="#storing-pandas-dataframes-in-plasma" title="Permalink to this headline">¶</a></h3> |
| <p>Storing a Pandas <code class="docutils literal notranslate"><span class="pre">DataFrame</span></code> still follows the <strong>create</strong> then <strong>seal</strong> |
| process of storing an object in the Plasma store, however one cannot directly |
| write the <code class="docutils literal notranslate"><span class="pre">DataFrame</span></code> to Plasma with Pandas alone. Plasma also needs to know |
| the size of the <code class="docutils literal notranslate"><span class="pre">DataFrame</span></code> to allocate a buffer for.</p> |
| <p>See <a class="reference internal" href="pandas.html#pandas-interop"><span class="std std-ref">Pandas Integration</span></a> for more information on using Arrow with Pandas.</p> |
| <p>You can create the pyarrow equivalent of a Pandas <code class="docutils literal notranslate"><span class="pre">DataFrame</span></code> by using |
| <code class="docutils literal notranslate"><span class="pre">pyarrow.from_pandas</span></code> to convert it to a <code class="docutils literal notranslate"><span class="pre">RecordBatch</span></code>.</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">pyarrow</span> <span class="kn">as</span> <span class="nn">pa</span> |
| <span class="kn">import</span> <span class="nn">pandas</span> <span class="kn">as</span> <span class="nn">pd</span> |
| |
| <span class="c1"># Create a Pandas DataFrame</span> |
| <span class="n">d</span> <span class="o">=</span> <span class="p">{</span><span class="s1">'one'</span> <span class="p">:</span> <span class="n">pd</span><span class="o">.</span><span class="n">Series</span><span class="p">([</span><span class="mf">1.</span><span class="p">,</span> <span class="mf">2.</span><span class="p">,</span> <span class="mf">3.</span><span class="p">],</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'a'</span><span class="p">,</span> <span class="s1">'b'</span><span class="p">,</span> <span class="s1">'c'</span><span class="p">]),</span> |
| <span class="s1">'two'</span> <span class="p">:</span> <span class="n">pd</span><span class="o">.</span><span class="n">Series</span><span class="p">([</span><span class="mf">1.</span><span class="p">,</span> <span class="mf">2.</span><span class="p">,</span> <span class="mf">3.</span><span class="p">,</span> <span class="mf">4.</span><span class="p">],</span> <span class="n">index</span><span class="o">=</span><span class="p">[</span><span class="s1">'a'</span><span class="p">,</span> <span class="s1">'b'</span><span class="p">,</span> <span class="s1">'c'</span><span class="p">,</span> <span class="s1">'d'</span><span class="p">])}</span> |
| <span class="n">df</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">DataFrame</span><span class="p">(</span><span class="n">d</span><span class="p">)</span> |
| |
| <span class="c1"># Convert the Pandas DataFrame into a PyArrow RecordBatch</span> |
| <span class="n">record_batch</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">RecordBatch</span><span class="o">.</span><span class="n">from_pandas</span><span class="p">(</span><span class="n">df</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>Creating the Plasma object requires an <code class="docutils literal notranslate"><span class="pre">ObjectID</span></code> and the size of the |
| data. Now that we have converted the Pandas <code class="docutils literal notranslate"><span class="pre">DataFrame</span></code> into a PyArrow |
| <code class="docutils literal notranslate"><span class="pre">RecordBatch</span></code>, use the <code class="docutils literal notranslate"><span class="pre">MockOutputStream</span></code> to determine the |
| size of the Plasma object.</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Create the Plasma object from the PyArrow RecordBatch. Most of the work here</span> |
| <span class="c1"># is done to determine the size of buffer to request from the object store.</span> |
| <span class="n">object_id</span> <span class="o">=</span> <span class="n">plasma</span><span class="o">.</span><span class="n">ObjectID</span><span class="p">(</span><span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">bytes</span><span class="p">(</span><span class="mi">20</span><span class="p">))</span> |
| <span class="n">mock_sink</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">MockOutputStream</span><span class="p">()</span> |
| <span class="n">stream_writer</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">RecordBatchStreamWriter</span><span class="p">(</span><span class="n">mock_sink</span><span class="p">,</span> <span class="n">record_batch</span><span class="o">.</span><span class="n">schema</span><span class="p">)</span> |
| <span class="n">stream_writer</span><span class="o">.</span><span class="n">write_batch</span><span class="p">(</span><span class="n">record_batch</span><span class="p">)</span> |
| <span class="n">stream_writer</span><span class="o">.</span><span class="n">close</span><span class="p">()</span> |
| <span class="n">data_size</span> <span class="o">=</span> <span class="n">mock_sink</span><span class="o">.</span><span class="n">size</span><span class="p">()</span> |
| <span class="n">buf</span> <span class="o">=</span> <span class="n">client</span><span class="o">.</span><span class="n">create</span><span class="p">(</span><span class="n">object_id</span><span class="p">,</span> <span class="n">data_size</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>The DataFrame can now be written to the buffer as follows.</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Write the PyArrow RecordBatch to Plasma</span> |
| <span class="n">stream</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">FixedSizeBufferWriter</span><span class="p">(</span><span class="n">buf</span><span class="p">)</span> |
| <span class="n">stream_writer</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">RecordBatchStreamWriter</span><span class="p">(</span><span class="n">stream</span><span class="p">,</span> <span class="n">record_batch</span><span class="o">.</span><span class="n">schema</span><span class="p">)</span> |
| <span class="n">stream_writer</span><span class="o">.</span><span class="n">write_batch</span><span class="p">(</span><span class="n">record_batch</span><span class="p">)</span> |
| <span class="n">stream_writer</span><span class="o">.</span><span class="n">close</span><span class="p">()</span> |
| </pre></div> |
| </div> |
| <p>Finally, seal the finished object for use by all clients:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Seal the Plasma object</span> |
| <span class="n">client</span><span class="o">.</span><span class="n">seal</span><span class="p">(</span><span class="n">object_id</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| </div> |
| <div class="section" id="getting-pandas-dataframes-from-plasma"> |
| <h3><a class="toc-backref" href="#id14">Getting Pandas DataFrames from Plasma</a><a class="headerlink" href="#getting-pandas-dataframes-from-plasma" title="Permalink to this headline">¶</a></h3> |
| <p>Since we store the Pandas DataFrame as a PyArrow <code class="docutils literal notranslate"><span class="pre">RecordBatch</span></code> object, |
| to get the object back from the Plasma store, we follow similar steps |
| to those specified in <a class="reference internal" href="#getting-arrow-objects-from-plasma">Getting Arrow Objects from Plasma</a>.</p> |
| <p>We first have to convert the <code class="docutils literal notranslate"><span class="pre">PlasmaBuffer</span></code> returned from |
| <code class="docutils literal notranslate"><span class="pre">client.get_buffers</span></code> into an Arrow <code class="docutils literal notranslate"><span class="pre">BufferReader</span></code> object.</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Fetch the Plasma object</span> |
| <span class="p">[</span><span class="n">data</span><span class="p">]</span> <span class="o">=</span> <span class="n">client</span><span class="o">.</span><span class="n">get_buffers</span><span class="p">([</span><span class="n">object_id</span><span class="p">])</span> <span class="c1"># Get PlasmaBuffer from ObjectID</span> |
| <span class="nb">buffer</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">BufferReader</span><span class="p">(</span><span class="n">data</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| <p>From the <code class="docutils literal notranslate"><span class="pre">BufferReader</span></code>, we can create a specific <code class="docutils literal notranslate"><span class="pre">RecordBatchStreamReader</span></code> |
| in Arrow to reconstruct the stored PyArrow <code class="docutils literal notranslate"><span class="pre">RecordBatch</span></code> object.</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Convert object back into an Arrow RecordBatch</span> |
| <span class="n">reader</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">RecordBatchStreamReader</span><span class="p">(</span><span class="nb">buffer</span><span class="p">)</span> |
| <span class="n">record_batch</span> <span class="o">=</span> <span class="n">reader</span><span class="o">.</span><span class="n">read_next_batch</span><span class="p">()</span> |
| </pre></div> |
| </div> |
| <p>The last step is to convert the PyArrow <code class="docutils literal notranslate"><span class="pre">RecordBatch</span></code> object back into |
| the original Pandas <code class="docutils literal notranslate"><span class="pre">DataFrame</span></code> structure.</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="c1"># Convert back into Pandas</span> |
| <span class="n">result</span> <span class="o">=</span> <span class="n">record_batch</span><span class="o">.</span><span class="n">to_pandas</span><span class="p">()</span> |
| </pre></div> |
| </div> |
| </div> |
| </div> |
| <div class="section" id="using-plasma-with-huge-pages"> |
| <h2><a class="toc-backref" href="#id15">Using Plasma with Huge Pages</a><a class="headerlink" href="#using-plasma-with-huge-pages" title="Permalink to this headline">¶</a></h2> |
| <p>On Linux it is possible to use the Plasma store with huge pages for increased |
| throughput. You first need to create a file system and activate huge pages with</p> |
| <div class="highlight-shell notranslate"><div class="highlight"><pre><span></span>sudo mkdir -p /mnt/hugepages |
| <span class="nv">gid</span><span class="o">=</span><span class="sb">`</span>id -g<span class="sb">`</span> |
| <span class="nv">uid</span><span class="o">=</span><span class="sb">`</span>id -u<span class="sb">`</span> |
| sudo mount -t hugetlbfs -o <span class="nv">uid</span><span class="o">=</span><span class="nv">$uid</span> -o <span class="nv">gid</span><span class="o">=</span><span class="nv">$gid</span> none /mnt/hugepages |
| sudo bash -c <span class="s2">"echo </span><span class="nv">$gid</span><span class="s2"> > /proc/sys/vm/hugetlb_shm_group"</span> |
| sudo bash -c <span class="s2">"echo 20000 > /proc/sys/vm/nr_hugepages"</span> |
| </pre></div> |
| </div> |
| <p>Note that you only need root access to create the file system, not for |
| running the object store. You can then start the Plasma store with the <code class="docutils literal notranslate"><span class="pre">-d</span></code> |
| flag for the mount point of the huge page file system and the <code class="docutils literal notranslate"><span class="pre">-h</span></code> flag |
| which indicates that huge pages are activated:</p> |
| <div class="highlight-shell notranslate"><div class="highlight"><pre><span></span>plasma_store -s /tmp/plasma -m <span class="m">10000000000</span> -d /mnt/hugepages -h |
| </pre></div> |
| </div> |
| <p>You can test this with the following script:</p> |
| <div class="highlight-python notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">numpy</span> <span class="kn">as</span> <span class="nn">np</span> |
| <span class="kn">import</span> <span class="nn">pyarrow</span> <span class="kn">as</span> <span class="nn">pa</span> |
| <span class="kn">import</span> <span class="nn">pyarrow.plasma</span> <span class="kn">as</span> <span class="nn">plasma</span> |
| <span class="kn">import</span> <span class="nn">time</span> |
| |
| <span class="n">client</span> <span class="o">=</span> <span class="n">plasma</span><span class="o">.</span><span class="n">connect</span><span class="p">(</span><span class="s2">"/tmp/plasma"</span><span class="p">)</span> |
| |
| <span class="n">data</span> <span class="o">=</span> <span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">100000000</span><span class="p">)</span> |
| <span class="n">tensor</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">Tensor</span><span class="o">.</span><span class="n">from_numpy</span><span class="p">(</span><span class="n">data</span><span class="p">)</span> |
| |
| <span class="n">object_id</span> <span class="o">=</span> <span class="n">plasma</span><span class="o">.</span><span class="n">ObjectID</span><span class="p">(</span><span class="n">np</span><span class="o">.</span><span class="n">random</span><span class="o">.</span><span class="n">bytes</span><span class="p">(</span><span class="mi">20</span><span class="p">))</span> |
| <span class="n">buf</span> <span class="o">=</span> <span class="n">client</span><span class="o">.</span><span class="n">create</span><span class="p">(</span><span class="n">object_id</span><span class="p">,</span> <span class="n">pa</span><span class="o">.</span><span class="n">ipc</span><span class="o">.</span><span class="n">get_tensor_size</span><span class="p">(</span><span class="n">tensor</span><span class="p">))</span> |
| |
| <span class="n">stream</span> <span class="o">=</span> <span class="n">pa</span><span class="o">.</span><span class="n">FixedSizeBufferWriter</span><span class="p">(</span><span class="n">buf</span><span class="p">)</span> |
| <span class="n">stream</span><span class="o">.</span><span class="n">set_memcopy_threads</span><span class="p">(</span><span class="mi">4</span><span class="p">)</span> |
| <span class="n">a</span> <span class="o">=</span> <span class="n">time</span><span class="o">.</span><span class="n">time</span><span class="p">()</span> |
| <span class="n">pa</span><span class="o">.</span><span class="n">ipc</span><span class="o">.</span><span class="n">write_tensor</span><span class="p">(</span><span class="n">tensor</span><span class="p">,</span> <span class="n">stream</span><span class="p">)</span> |
| <span class="k">print</span><span class="p">(</span><span class="s2">"Writing took "</span><span class="p">,</span> <span class="n">time</span><span class="o">.</span><span class="n">time</span><span class="p">()</span> <span class="o">-</span> <span class="n">a</span><span class="p">)</span> |
| </pre></div> |
| </div> |
| </div> |
| </div> |
| |
| |
| </div> |
| |
| </div> |
| <footer> |
| |
| <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation"> |
| |
| <a href="numpy.html" class="btn btn-neutral float-right" title="NumPy Integration" accesskey="n" rel="next">Next <span class="fa fa-arrow-circle-right"></span></a> |
| |
| |
| <a href="generated/pyarrow.HdfsFile.html" class="btn btn-neutral float-left" title="pyarrow.HdfsFile" accesskey="p" rel="prev"><span class="fa fa-arrow-circle-left"></span> Previous</a> |
| |
| </div> |
| |
| |
| <hr/> |
| |
| <div role="contentinfo"> |
| <p> |
| |
| © Copyright 2016-2019 Apache Software Foundation |
| |
| </p> |
| </div> |
| |
| |
| |
| Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a |
| |
| <a href="https://github.com/rtfd/sphinx_rtd_theme">theme</a> |
| |
| provided by <a href="https://readthedocs.org">Read the Docs</a>. |
| |
| </footer> |
| |
| </div> |
| </div> |
| |
| </section> |
| |
| </div> |
| |
| |
| <script type="text/javascript"> |
| jQuery(function () { |
| SphinxRtdTheme.Navigation.enable(true); |
| }); |
| </script> |
| |
| |
| |
| |
| |
| |
| |
| <script type="text/javascript" src="/docs/_static/versionwarning.js"></script></body> |
| </html> |
