fabric_en.html



<!DOCTYPE html>
<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
<head>
  <meta charset="utf-8">
  
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  
  <title>Fabric &mdash; PaddlePaddle  documentation</title>
  

    <link rel="stylesheet" href="../../../_static/css/theme.css" type="text/css" />
  

        <link rel="index" title="Index"
              href="../../../genindex.html"/>
        <link rel="search" title="Search" href="../../../search.html"/>
    <link rel="top" title="PaddlePaddle  documentation" href="../../../index.html"/>
        <link rel="up" title="Use different clusters" href="index_en.html"/>
        <link rel="next" title="OpenMPI" href="openmpi_en.html"/>
        <link rel="prev" title="Use different clusters" href="index_en.html"/> 

  
  <script src="../../../_static/js/modernizr.min.js"></script>

</head>

<body class="wy-body-for-nav" role="document">

  <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_en.html" class="icon icon-home"> PaddlePaddle
          

          </a>

          
<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">
          
            
                <ul class="current">
<li class="toctree-l1"><a class="reference internal" href="../../../getstarted/index_en.html">GET STARTED</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../../build_and_install/index_en.html">Install and Build</a></li>
<li class="toctree-l1 current"><a class="reference internal" href="../../index_en.html">HOW TO</a><ul class="current">
<li class="toctree-l2"><a class="reference internal" href="../../cmd_parameter/index_en.html">Set Command-line Parameters</a></li>
<li class="toctree-l2 current"><a class="reference internal" href="../index_en.html">Distributed Training</a><ul class="current">
<li class="toctree-l3"><a class="reference internal" href="../preparations_en.html">Preparations</a></li>
<li class="toctree-l3"><a class="reference internal" href="../cmd_argument_en.html">Command-line arguments</a></li>
<li class="toctree-l3 current"><a class="reference internal" href="index_en.html">Use different clusters</a><ul class="current">
<li class="toctree-l4 current"><a class="current reference internal" href="#">Fabric</a></li>
<li class="toctree-l4"><a class="reference internal" href="openmpi_en.html">OpenMPI</a></li>
<li class="toctree-l4"><a class="reference internal" href="k8s_en.html">Kubernetes</a></li>
<li class="toctree-l4"><a class="reference internal" href="k8s_aws_en.html">Kubernetes on AWS</a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../rnn/index_en.html">RNN Models</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../../dev/index_en.html">Development</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../../faq/index_en.html">FAQ</a></li>
</ul>

            
        </div>
      </div>
    </nav>

    <section data-toggle="wy-nav-shift" class="wy-nav-content-wrap">

      
      <nav class="wy-nav-top" role="navigation" aria-label="top navigation">
        <i data-toggle="wy-nav-top" class="fa fa-bars"></i>
        <a href="../../../index_en.html">PaddlePaddle</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_en.html">Docs</a> &raquo;</li>
      
          <li><a href="../../index_en.html">HOW TO</a> &raquo;</li>
      
          <li><a href="../index_en.html">Distributed Training</a> &raquo;</li>
      
          <li><a href="index_en.html">Use different clusters</a> &raquo;</li>
      
    <li>Fabric</li>
      <li class="wy-breadcrumbs-aside">
        
          
            <a href="../../../_sources/howto/cluster/multi_cluster/fabric_en.md.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="fabric">
<span id="fabric"></span><h1>Fabric<a class="headerlink" href="#fabric" title="Permalink to this headline">¶</a></h1>
<div class="section" id="prepare-a-linux-cluster">
<span id="prepare-a-linux-cluster"></span><h2>Prepare a Linux cluster<a class="headerlink" href="#prepare-a-linux-cluster" title="Permalink to this headline">¶</a></h2>
<p>Run <code class="docutils literal"><span class="pre">kubectl</span> <span class="pre">-f</span> <span class="pre">ssh_servers.yaml</span></code> under the directory:  <code class="docutils literal"><span class="pre">paddle/scripts/cluster_train_v2/fabric/docker_cluster</span></code> will launch a demo cluster. Run <code class="docutils literal"><span class="pre">kubectl</span> <span class="pre">get</span> <span class="pre">po</span> <span class="pre">-o</span> <span class="pre">wide</span></code> to get IP addresses of these nodes.</p>
</div>
<div class="section" id="launching-cluster-job">
<span id="launching-cluster-job"></span><h2>Launching Cluster Job<a class="headerlink" href="#launching-cluster-job" title="Permalink to this headline">¶</a></h2>
<p><code class="docutils literal"><span class="pre">paddle.py</span></code> provides automatical scripts to start all PaddlePaddle cluster processes in different nodes. By default, all command line options can be set as <code class="docutils literal"><span class="pre">paddle.py</span></code> command options and <code class="docutils literal"><span class="pre">paddle.py</span></code> will transparently and automatically set these options to PaddlePaddle lower level processes.</p>
<p><code class="docutils literal"><span class="pre">paddle.py</span></code>provides two distinguished command option for easy job launching.</p>
<ul class="simple">
<li><code class="docutils literal"><span class="pre">job_dispatch_package</span></code> set it with local <code class="docutils literal"><span class="pre">workspace</span></code> directory, it will be dispatched to all nodes which is set in <code class="docutils literal"><span class="pre">conf.py</span></code>. It could be helpful for frequently manipulating workspace files. otherwise, frequent multi-nodes workspace deployment is very annoying.</li>
<li><code class="docutils literal"><span class="pre">job_workspace</span></code>  set it with already deployed workspace directory, <code class="docutils literal"><span class="pre">paddle.py</span></code> will skip dispatch stage to directly launch cluster job with all nodes. It could help to reduce heavy
dispatch latency.</li>
</ul>
<p><code class="docutils literal"><span class="pre">cluster_train/run.sh</span></code> provides command line sample to run <code class="docutils literal"><span class="pre">demo/recommendation</span></code> cluster job, just modify <code class="docutils literal"><span class="pre">job_dispatch_package</span></code> and <code class="docutils literal"><span class="pre">job_workspace</span></code> with your defined directory, then:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">sh</span> <span class="n">run</span><span class="o">.</span><span class="n">sh</span>
</pre></div>
</div>
<p>The cluster Job will start in several seconds.</p>
</div>
<div class="section" id="kill-cluster-job">
<span id="kill-cluster-job"></span><h2>Kill Cluster Job<a class="headerlink" href="#kill-cluster-job" title="Permalink to this headline">¶</a></h2>
<p><code class="docutils literal"><span class="pre">paddle.py</span></code> can capture <code class="docutils literal"><span class="pre">Ctrl</span> <span class="pre">+</span> <span class="pre">C</span></code> SIGINT signal to automatically kill all processes launched by it. So just stop <code class="docutils literal"><span class="pre">paddle.py</span></code> to kill cluster job. You should manually kill the job if the program crashed.</p>
</div>
<div class="section" id="check-cluster-training-result">
<span id="check-cluster-training-result"></span><h2>Check Cluster Training Result<a class="headerlink" href="#check-cluster-training-result" title="Permalink to this headline">¶</a></h2>
<p>Check log in $workspace/log for details, each node owns same log structure.</p>
<p><code class="docutils literal"><span class="pre">paddle_trainer.INFO</span></code>
It provides almost all internal output log for training,  same as local training. Check runtime model convergence here.</p>
<p><code class="docutils literal"><span class="pre">paddle_pserver2.INFO</span></code>
It provides parameter server running log, which could help to diagnose distributed error.</p>
<p><code class="docutils literal"><span class="pre">server.log</span></code>
It provides stderr and stdout of parameter server process. Check error log if training crashes.</p>
<p><code class="docutils literal"><span class="pre">train.log</span></code>
It provides stderr and stdout of trainer process. Check error log if training crashes.</p>
</div>
<div class="section" id="check-model-output">
<span id="check-model-output"></span><h2>Check Model Output<a class="headerlink" href="#check-model-output" title="Permalink to this headline">¶</a></h2>
<p>After one pass finished, model files will be written in <code class="docutils literal"><span class="pre">output</span></code> directory in node 0.
<code class="docutils literal"><span class="pre">nodefile</span></code> in workspace indicates the node id of current cluster job.</p>
</div>
</div>


           </div>
          </div>
          <footer>
  
    <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
      
        <a href="openmpi_en.html" class="btn btn-neutral float-right" title="OpenMPI" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
      
      
        <a href="index_en.html" class="btn btn-neutral" title="Use different clusters" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
      
    </div>
  

  <hr/>

  <div role="contentinfo">
    <p>
        &copy; Copyright 2016, PaddlePaddle developers.

    </p>
  </div>
  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/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">
        var DOCUMENTATION_OPTIONS = {
            URL_ROOT:'../../../',
            VERSION:'',
            COLLAPSE_INDEX:false,
            FILE_SUFFIX:'.html',
            HAS_SOURCE:  true
        };
    </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="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>

  
    <script type="text/javascript" src="../../../_static/js/theme.js"></script>
  

  <script type="text/javascript">
      jQuery(function () {
          SphinxRtdTheme.StickyNav.enable();
      });
  </script>
   

</body>
</html>