core/org.eclipse.ptp.doc.user/html/03pRunConfig.html - ptp/org.eclipse.ptp - Git at Google

 <html>
 <head>
 <title>Running Parallel Programs</title>
 <link rel="stylesheet" type="text/css" href="help.css">
 <script type="text/javascript" src="thumb.js"> </script>
 </head>
 <body>

 <h1 id="top">Running Parallel Programs</h1>
 <p>This section describes how to run a parallel program. In PTP,
 this is called &quot;running a job&quot;.
 <p>Previous releases of PTP required setting up a <i>Resource Manager</i> for each job scheduler/runtime environment before
 a job could be launched.
 In PTP 6.0 and later this is not handled separately, and instead a
 <i>Target System Configuration</i>
 is specified on the Run Configuration page.  Further interaction to set up the parallel
 job is highly dependent on which TSC is selected.
 <p>PTP is
 able to launch parallel applications that use a variety of programming
 models (e.g. MPI, OpenMP, UPC, etc.) provided that the <i>Target System Configuration</i>
 supports the programming model. It is also possible to launch multiple
 copies of a sequential program so that they execute simultaneously.</p>

 <p>Eclipse uses a <i>run configuration</i> (also known as a <i>launch configuration</i>)
 to encapsulate all the
 settings and arguments necessary to launch a job normally or under the
 control of a debugger. A run configuration is created using the <b>Run
 &gt; Run Configurations...</b> menu.

 <p>This section only describes how to launch a parallel program. A
 description of how to monitor the application launch is provided in
 <a href="05monitoring.html">Monitoring Jobs and Systems</a>.
 A description of how to debug an application is provided in
 <a href="06parDebugging.html">Parallel Debugging</a>.</p>

 <p>Topics include:</p>
 <ol>
 	<li><a href="#create">Creating a <b>Parallel Application</b>
 	run configuration</a></li>
 	<li><a href="#resources">Completing the <b>Resources tab</b></a> - selecting the target system configuration
 	</li>
 	<li><a href="#application">Completing the <b>Application
 	tab</b></a></li>
 	<li><a href="#arguments">Completing the <b>Arguments tab</b></a></li>
 	<li><a href="#environment">Completing the <b>Environment
 	tab</b></a></li>
 	<li><a href="#synchronize">Completing the <b>Synchronize
 	tab</b></a></li>
 	<li><a href="#launch">Running the application </a></li>
 </ol>
 <h2 id="create">Step 1: Create a Parallel Application run
 configuration</h2>

 <ol>
 	<li>
 	<p>To set up a run configuration, open the <b>Run Configurations</b>
 	dialog using <b>Run &gt; Run Configurations...</b>
 	<p><img src="images/03runMenu.png">
 	<br>or click the run
 	icon's context menu on the tool bar (small triangle next to the icon) and select <b>Run Configurations...</b>.
 	<p><img src="images/03runMenu2.png">
 	</p>
 	</p>
 	<p><br>
 	</p>
 	</li>
 	<li>
 	<p>Create a new Parallel Application run configuration by
 	clicking <b>Parallel Application</b> and then clicking on the <b>New</b>
 	button icon.</p>
 	<p><br>
 	</p>
 	<p><img src="images/03launch1.png"></p>
 	<p><br>
 	</p>
 	</li>
 	<li>
 	<p>Enter a name for the configuration in the <b>Name:</b> field.
 	Here we use the same name as the project.  If the project was selected in the Project Explorer, it may be already filled in here.</p>
 	<p><br>
 	</p>
 	<p><img src="images/03launch1a.png"></p>
 	 </li>
 </ol>

 <h2 id="resources">Step 2: Complete the Resources tab</h2>
 <p>The <b>Resources tab</b> is used to specify the resources
 required for the job execution. After selecting the <b>target system configuration</b>, the contents of this tab will vary for
 different target configurations, and your installation may have added others.
 <h3 id="ptp-tscs">PTP supported Target System Configurations</h3>
 <p>
 PTP includes a number of target configurations
 for different systems that support various parallel runtime systems and job schedulers. The following table lists the common configurations along with
 links to additional help documentation.
 <p>
 Note that the associated runtime or job scheduler needs to be available on the target system prior to using PTP:
 <blockquote>
 <table border="1">
    <tr>
    <th scope="col">Target Configuration</th>
    <th scope="col">Description</th>
    </tr>
    <tr id="gridengine">
     <td><a href="rms/gridengine.html">Grid Engine-Generic-Batch</a></td>
     <td>Batch execution using Grid Engine</td>
   </tr>
   <tr id="ibm-ll">
     <td><a href="rms/ibm-ll.html">IBM LoadLeveler</a></td>
 	<td>Generic IBM LoadLeveler batch systems</td>
   </tr>
   <tr id="ibm-ll-bg">
     <td><a href="rms/ibm-ll.html">IBM LoadLeveler (Blue Gene)</a></td>
 	<td>IBM LoadLeveler on Blue Gene systems</td>
   </tr>
   <tr id="ibm-pe">
     <td><a href="rms/ibm-pe.html">IBM Parallel Environment</a></td>
     <td>Interactive execution using IBM Parallel Environment</td>
   </tr>
     <tr id="ibm-lsf">
     <td><a href="PLUGINS_ROOT/org.eclipse.ptp.rm.ibm.platform.lsf.doc.user/html/lsf.html">IBM Platform LSF</a></td>
  <td>Batch execution on IBM Platform LSF systems</td>
   </tr>
   <tr id="ibm-platform-mpi">
     <td><a href="rms/ibm-pmpi.html">IBM Platform MPI</a></td>
     <td>IBM Platform MPI</td>
   </tr>
       <tr id="mpich2">
     <td>MPICH2-Generic-Interactive</td>
     <td>Interactive execution on MPICH2 systems</td>
   </tr>
   <tr id="openMPI">
     <td><a href="rms/openMPI.html">Open MPI-Generic-Interactive</a></td>
     <td>Interactive execution on Open MPI systems</td>
   </tr>
   <tr id="pbs-int">
     <td><a href="rms/pbs.html">PBS-Generic-Interactive</a></td>
     <td>Interactive job execution using PBS batch systems</td>
   </tr>
   <tr id="pbs-bgp-batch">
     <td><a href="rms/pbs.html">PBS-BGP-Batch</a></td>
     <td>Blue Gene/P systems using </td>
   </tr>
   <tr id="pbs-bgq-batch">
     <td><a href="rms/pbs.html">PBS-BGQ-Batch</a></td>
     <td>Blue Gene/Q systems using PBS</td>
   </tr>

   <tr id="slurm">
     <td><a href="rms/slurm.html">SLURM-Generic-Batch</a></td>
     <td>Generic SLURM batch systems</td>
   </tr>
   <tr id="slurm-alps">
     <td><a href="rms/slurmalps.html">SLURM-ALPS-Generic-Batch</a></td>
     <td>Combination of SLURM and ALPS as batch system</td>
   </tr>
   <tr id="slurm-bgp">
     <td><a href="rms/slurm.html">SLURM-BGP-Batch</a></td>
     <td>Blue Gene/P systems using SLURM</td>
   </tr>
   <tr id="slurm-bgq-batch">
     <td><a href="rms/slurm.html">SLURM-BGQ-Batch</a></td>
     <td>Blue Gene/Q systems using SLURM</td>
   </tr>
   <tr id="pbs-batch">
     <td><a href="rms/pbs.html">PBS-Generic-Batch</a></td>
     <td>Generic PBS batch systems</td>
   </tr>
     <tr id="remote-generic-interactive">
     <td><a href="rms/remote-generic-interactive.html">Remote-Generic-Interactive</a></td>
     <td>Simple (non-parallel) interactive execution on remote systems</td>
   </tr>

   <tr id="torque-batch">
     <td><a href="rms/torque.html">Torque-Generic-Batch</a></td>
     <td>Generic TORQUE batch systems</td>
   </tr>
   <tr id="torque-int">
     <td><a href="rms/torque.html">Torque-Generic-Interactive</a></td>
     <td>Interactive job execution using Torque batch systems</td>
   </tr>
   <tr id="juelich-judge">
     <td><a href="rms/de.fz-juelich.judge.html">de.fz-juelich.judge.torque.batch</a></td>
     <td>Running batch jobs on the JSC system JUDGE</td>
   </tr>
   <tr id="juelich-juropa">
     <td><a href="rms/de.fz-juelich.juropa.html">de.fz-juelich.juropa.torque.batch</a></td>
     <td>Running batch jobs on the JSC system JUROPA</td>
   </tr>
   <tr id="juelich-juqueen">
     <td><a href="rms/de.fz-juelich.juqueen.html">de.fz-juelich.juqueen.ll_bg</a></td>
     <td>Running batch jobs on the JSC system JUQUEEN</td>
   </tr>
 </table>
 </blockquote>
 </p>
 <ol>
 	<li>
 	<p>Select a <b>Target System Configuration</b>
 	you wish to use from the drop-down.

 	</p>
 	<p>&nbsp;
 	<p><img src="images/03resources_configs.png">
 	</p>
 	<p>&nbsp;

 	<p>Note that if you type a lone character (or a few) into the selection box, the list of target system configurations
 	will be limited to those that start with the character.
 	<p><img src="images/03resources_configs_cull.png">

 	</li>
 	<li>
 	<!-- !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! -->
 	<p>Select the connection to use for the target system.
 	If you have previously configured a synchronized or remote project, then a connection should be
 	available, and may be already be selected.
 	<br>(To launch on the local machine, select <b>Local</b> for the connection type.)
 	<br>You can create a new connection here by clicking the <b>New...</b> button.
 	 </p>
 	<p>
 	<p>&nbsp;</p>
 	<li>
 	<p>
 	Once you select the connection, you may be prompted to open the connection if it is not already open. The contents of the <b>Resources</b>
 	tab will then populate with controls that allow you to configure the necessary launch information for the target system.

 	This area usually (depending on the target system configuration) includes further tabs to
     segment the other information you can provide.
     Fill in the information required by the target system configuration.
 	<br>
 	</p>
 	</li>


 	<!-- unused image<img src="images/03launchResTab.png"> -->

 </ol>

 <h2 id="application">Step 3: Complete the <b>Application</b> tab</h2>
 <p>The <b>Application tab</b> is used to specify the application
 (executable) to be launched.</p>
 <ol>
 	<li>
 	<p>Ensure that the correct <b>Parallel Project</b> is selected.
 	This is the project containing the source and binaries for the
 	application.</p>
 	<p><br>
 	</p>
 	</li>

 	<li>
 	<p>Select the <b>Application Program</b> (executable) you wish to
 	launch. If the target system connection points to a remote target, then
 	this means the <i>path of the executable on the remote target</i>.
 	Otherwise, for a connection pointing to the local machine, it should be
 	the path to the executable within the project directory.
 	<p>The <b>Browse...</b> button can be used to select the executable file.
 	</p>
 	<p><br>
 	</p>
 	</li>

 	<li>
 	<p>If you are running the application on a remote machine and the
 	project was <i>built</i> locally, then you will have to copy the
 	executable to the remote machine first. For this purpose, you may
 	enable <b>Copy executable from local filesystem</b> and fill <b>Path
 	to the local file</b> with the path to the executable within the project
 	directory. This configures the launcher to copy the executable
 	automatically to the remote target before starting the run/debug
 	session. If not enabled, then you will need to copy the executable
 	manually to the remote target.</p>

 	<p><br>
 	</p>
 	</li>

 	<p><img src="images/03launch6.png"></p>
 	<p><br>
 	</p>

 	<li>
 	<p>If you wish to have the output from all processes in a single
 	output console, ensure that the <b>Display output from all
 	processes in a console view</b> option is checked.</p>
 	<p><br>
 	</p>
 	</li>

 </ol>

 <h2 id="arguments">Step 4: Complete the Arguments tab</h2>
 <p>The <b>Arguments tab</b> is used to specify any command-line
 arguments require by the parallel application and to modify the working
 directory.</p>
 <ol>
 	<li>
 	<p>Enter any arguments required by the application in the <b>Program arguments</b> field.
 	This is a free form text field that will be
 	passed to the application verbatim.</p>
 	<p><br>
 	</p>
 	</li>
 	<li>
 	<p>Select a different working directory if desired. The working
 	directory is set just prior to the application launch. Note that for a
 	remote launch, this will be a directory on the remote machine.</p>
 	</li>
 	<p><img src="images/03launch3.png"></p>
 	<p><br>
 	</p>
 </ol>

 <h2 id="environment">Step 5: Complete the Environment tab</h2>
 <p>The <b>Environment tab</b> is used to specify any environment
 variables require by the parallel application or in order to run the
 application. The tab also provides the option to append the environment
 variables to the native environment, or to replace the native
 environment with the variables specified in the tab.</p>
 <ol>
 	<li>
 	<p>Use the <b>New...</b> button to create a new environment
 	variable.</p>
 	<p><br>
 	</p>
 	</li>
 	<li>
 	<p>Use the <b>Select...</b> button to import environment variables.</p>
 	<p><br>
 	</p>
 	</li>
 	<li>
 	<p>Use the <b>Edit...</b> button to edit an environment variables.</p>
 	<p><br>
 	</p>
 	</li>
 	<li>
 	<p>Use the <b>Remove...</b> button to remove an environment
 	variables.</p>
 	<p><br>
 	</p>
 	</li>
 	<p><img src="images/03launch4.png"></p>
 	<p><br>
 	</p>
 </ol>

 <h2 id="synchronize">Step 6: Complete the <b>Synchronize</b> tab</h2>
 <p>The <b>Synchronize tab</b> is used in scenarios where the
 connection refers to a remote target. The tab specifies
 rules that describe files or directories that need to be copied (<i>uploaded</i>)
 to the remote target before starting the parallel application execution.
 Typically, this feature is used to upload dependencies (like dynamic
 libraries) or input data for test cases.</p>
 <p>Note that this may not be necessary for <a href="sync.html">Synchronized Projects</a>,
 in which files are copied back and forth to the remote system independently of the
 launching mechanism.

 <p>This tab also allows to specify rules describing files on the
 remote host to be retrieved back (<i>downloaded</i>) to the local host
 after the parallel application finishes execution, as output data
 produced by the parallel application execution.</p>

 <ol>
 	<li>
 	<p>If you wish to enable file transfer to the remote target, enable
 	the <b>Upload rules enabled</b> or <b>Download rules enabled</b>
 	according which direction of file transfer is required.</p>
 	<p><br>
 	</p>
 	<p><img src="images/03launch7.png"></p>
 	<p><br>
 	</p>
 	</li>

 	<li>
 	<p>In order to specify files to be copied from the local system to
 	the remote target (<i>uploaded</i>), click on <b>New upload rule</b>.
 	An upload rule lists a collection of files or directories that shall be
 	copied into a single directory.</p>
 	<p><br>
 	</p>
 	<p>If <b>Use directory from launch configuration</b> is enabled,
 	then the listed files or directories are copied to the working
 	directory specified on the launch configuration. At the moment, this is
 	equivalent to same directory that contains the executable for the
 	parallel application. If not enabled, they you must specify on <b>Remote
 	directory</b> a path where the files or directories will be placed. The
 	path may be absolute or relative (to the working directory).</p>
 	<p><br>
 	</p>
 	<p>In the list underneath, <b>Selected file(s)</b>, add the files
 	or directories that shall be copied. Click on <b>File(s)</b> to select
 	one or more individual files from anywhere in the local file system.
 	The <b>Directory</b> button works similarly to select a directory from
 	the local file system. Directories are copied recursively. The <b>Workspace</b>
 	button allows easily choosing files from the current workspace.</p>
 	<p><br>
 	</p>
 	<p>On the bottom of the dialog, you may wish to set attributes to
 	be applied to the copied files on the remote target. The options <b>Readonly</b>
 	or <b>Executable</b> sets all copied files to read-only or executable
 	on the remote host. If <b>Preserve time attributes</b> is enabled, then
 	the remote copy will assume the same time stamp then the local file,
 	otherwise, the time stamp will be the time the file was copied. Also,
 	if the file already exists on the remote target, you may choose how to
 	react in the <b>If file already exists</b> dropdown.</p>
 	<p><br>
 	</p>
 	<p><img src="images/03launch8.png"></p>
 	<p><br>
 	</p>
 	<p><br>
 	</p>
 	</li>
 	<li>
 	<p>In order to specify files to be copied from the remote target to
 	the local system (<i>downloaded</i>) after the parallel application
 	finishes, click on <b>New download rule</b>.</p>
 	<p><br>
 	</p>
 	<p>An download rule lists a collection of files or directories that
 	shall be copied into a single local directory. Its options are very
 	similar to the upload rule.</p>
 	<p><br>
 	</p>
 	<p><img src="images/03launch9.png"></p>
 	<p><br>
 	</p>
 	</li>
 </ol>

 <h2 id="launch">Step 7: Launching the application</h2>
 <p>The launch configuration is now complete.
 Select the <b>Run</b> button (in the bottom right corner of the <b>Run Configurations</b> dialog to launch the job.
 Switch to the <b>System Monitoring</b> perspective
 to view the status of the job and inspect the program output.
 <p><a href="#top">Back to Top</a> | <a href="toc.html">Back to Table of Contents</a>
 </body>
 </html>
	<html>
	<head>
	<title>Running Parallel Programs</title>
	<link rel="stylesheet" type="text/css" href="help.css">
	<script type="text/javascript" src="thumb.js"> </script>
	</head>
	<body>

	<h1 id="top">Running Parallel Programs</h1>
	<p>This section describes how to run a parallel program. In PTP,
	this is called "running a job".
	<p>Previous releases of PTP required setting up a <i>Resource Manager</i> for each job scheduler/runtime environment before
	a job could be launched.
	In PTP 6.0 and later this is not handled separately, and instead a
	<i>Target System Configuration</i>
	is specified on the Run Configuration page. Further interaction to set up the parallel
	job is highly dependent on which TSC is selected.
	<p>PTP is
	able to launch parallel applications that use a variety of programming
	models (e.g. MPI, OpenMP, UPC, etc.) provided that the <i>Target System Configuration</i>
	supports the programming model. It is also possible to launch multiple
	copies of a sequential program so that they execute simultaneously.</p>

	<p>Eclipse uses a <i>run configuration</i> (also known as a <i>launch configuration</i>)
	to encapsulate all the
	settings and arguments necessary to launch a job normally or under the
	control of a debugger. A run configuration is created using the <b>Run
	> Run Configurations...</b> menu.

	<p>This section only describes how to launch a parallel program. A
	description of how to monitor the application launch is provided in
	<a href="05monitoring.html">Monitoring Jobs and Systems</a>.
	A description of how to debug an application is provided in
	<a href="06parDebugging.html">Parallel Debugging</a>.</p>

	<p>Topics include:</p>
	<ol>
	<li><a href="#create">Creating a <b>Parallel Application</b>
	run configuration</a></li>
	<li><a href="#resources">Completing the <b>Resources tab</b></a> - selecting the target system configuration
	</li>
	<li><a href="#application">Completing the <b>Application
	tab</b></a></li>
	<li><a href="#arguments">Completing the <b>Arguments tab</b></a></li>
	<li><a href="#environment">Completing the <b>Environment
	tab</b></a></li>
	<li><a href="#synchronize">Completing the <b>Synchronize
	tab</b></a></li>
	<li><a href="#launch">Running the application </a></li>
	</ol>
	<h2 id="create">Step 1: Create a Parallel Application run
	configuration</h2>

	<ol>
	<li>
	<p>To set up a run configuration, open the <b>Run Configurations</b>
	dialog using <b>Run > Run Configurations...</b>
	<p><img src="images/03runMenu.png">
	<br>or click the run
	icon's context menu on the tool bar (small triangle next to the icon) and select <b>Run Configurations...</b>.
	<p><img src="images/03runMenu2.png">
	</p>
	</p>
	<p><br>
	</p>
	</li>
	<li>
	<p>Create a new Parallel Application run configuration by
	clicking <b>Parallel Application</b> and then clicking on the <b>New</b>
	button icon.</p>
	<p><br>
	</p>
	<p><img src="images/03launch1.png"></p>
	<p><br>
	</p>
	</li>
	<li>
	<p>Enter a name for the configuration in the <b>Name:</b> field.
	Here we use the same name as the project. If the project was selected in the Project Explorer, it may be already filled in here.</p>
	<p><br>
	</p>
	<p><img src="images/03launch1a.png"></p>
	</li>
	</ol>

	<h2 id="resources">Step 2: Complete the Resources tab</h2>
	<p>The <b>Resources tab</b> is used to specify the resources
	required for the job execution. After selecting the <b>target system configuration</b>, the contents of this tab will vary for
	different target configurations, and your installation may have added others.
	<h3 id="ptp-tscs">PTP supported Target System Configurations</h3>
	<p>
	PTP includes a number of target configurations
	for different systems that support various parallel runtime systems and job schedulers. The following table lists the common configurations along with
	links to additional help documentation.
	<p>
	Note that the associated runtime or job scheduler needs to be available on the target system prior to using PTP:
	<blockquote>
	<table border="1">
	<tr>
	<th scope="col">Target Configuration</th>
	<th scope="col">Description</th>
	</tr>
	<tr id="gridengine">
	<td><a href="rms/gridengine.html">Grid Engine-Generic-Batch</a></td>
	<td>Batch execution using Grid Engine</td>
	</tr>
	<tr id="ibm-ll">
	<td><a href="rms/ibm-ll.html">IBM LoadLeveler</a></td>
	<td>Generic IBM LoadLeveler batch systems</td>
	</tr>
	<tr id="ibm-ll-bg">
	<td><a href="rms/ibm-ll.html">IBM LoadLeveler (Blue Gene)</a></td>
	<td>IBM LoadLeveler on Blue Gene systems</td>
	</tr>
	<tr id="ibm-pe">
	<td><a href="rms/ibm-pe.html">IBM Parallel Environment</a></td>
	<td>Interactive execution using IBM Parallel Environment</td>
	</tr>
	<tr id="ibm-lsf">
	<td><a href="PLUGINS_ROOT/org.eclipse.ptp.rm.ibm.platform.lsf.doc.user/html/lsf.html">IBM Platform LSF</a></td>
	<td>Batch execution on IBM Platform LSF systems</td>
	</tr>
	<tr id="ibm-platform-mpi">
	<td><a href="rms/ibm-pmpi.html">IBM Platform MPI</a></td>
	<td>IBM Platform MPI</td>
	</tr>
	<tr id="mpich2">
	<td>MPICH2-Generic-Interactive</td>
	<td>Interactive execution on MPICH2 systems</td>
	</tr>
	<tr id="openMPI">
	<td><a href="rms/openMPI.html">Open MPI-Generic-Interactive</a></td>
	<td>Interactive execution on Open MPI systems</td>
	</tr>
	<tr id="pbs-int">
	<td><a href="rms/pbs.html">PBS-Generic-Interactive</a></td>
	<td>Interactive job execution using PBS batch systems</td>
	</tr>
	<tr id="pbs-bgp-batch">
	<td><a href="rms/pbs.html">PBS-BGP-Batch</a></td>
	<td>Blue Gene/P systems using </td>
	</tr>
	<tr id="pbs-bgq-batch">
	<td><a href="rms/pbs.html">PBS-BGQ-Batch</a></td>
	<td>Blue Gene/Q systems using PBS</td>
	</tr>

	<tr id="slurm">
	<td><a href="rms/slurm.html">SLURM-Generic-Batch</a></td>
	<td>Generic SLURM batch systems</td>
	</tr>
	<tr id="slurm-alps">
	<td><a href="rms/slurmalps.html">SLURM-ALPS-Generic-Batch</a></td>
	<td>Combination of SLURM and ALPS as batch system</td>
	</tr>
	<tr id="slurm-bgp">
	<td><a href="rms/slurm.html">SLURM-BGP-Batch</a></td>
	<td>Blue Gene/P systems using SLURM</td>
	</tr>
	<tr id="slurm-bgq-batch">
	<td><a href="rms/slurm.html">SLURM-BGQ-Batch</a></td>
	<td>Blue Gene/Q systems using SLURM</td>
	</tr>
	<tr id="pbs-batch">
	<td><a href="rms/pbs.html">PBS-Generic-Batch</a></td>
	<td>Generic PBS batch systems</td>
	</tr>
	<tr id="remote-generic-interactive">
	<td><a href="rms/remote-generic-interactive.html">Remote-Generic-Interactive</a></td>
	<td>Simple (non-parallel) interactive execution on remote systems</td>
	</tr>

	<tr id="torque-batch">
	<td><a href="rms/torque.html">Torque-Generic-Batch</a></td>
	<td>Generic TORQUE batch systems</td>
	</tr>
	<tr id="torque-int">
	<td><a href="rms/torque.html">Torque-Generic-Interactive</a></td>
	<td>Interactive job execution using Torque batch systems</td>
	</tr>
	<tr id="juelich-judge">
	<td><a href="rms/de.fz-juelich.judge.html">de.fz-juelich.judge.torque.batch</a></td>
	<td>Running batch jobs on the JSC system JUDGE</td>
	</tr>
	<tr id="juelich-juropa">
	<td><a href="rms/de.fz-juelich.juropa.html">de.fz-juelich.juropa.torque.batch</a></td>
	<td>Running batch jobs on the JSC system JUROPA</td>
	</tr>
	<tr id="juelich-juqueen">
	<td><a href="rms/de.fz-juelich.juqueen.html">de.fz-juelich.juqueen.ll_bg</a></td>
	<td>Running batch jobs on the JSC system JUQUEEN</td>
	</tr>
	</table>
	</blockquote>
	</p>
	<ol>
	<li>
	<p>Select a <b>Target System Configuration</b>
	you wish to use from the drop-down.

	</p>
	<p>
	<p><img src="images/03resources_configs.png">
	</p>
	<p>

	<p>Note that if you type a lone character (or a few) into the selection box, the list of target system configurations
	will be limited to those that start with the character.
	<p><img src="images/03resources_configs_cull.png">

	</li>
	<li>
	<!-- !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! -->
	<p>Select the connection to use for the target system.
	If you have previously configured a synchronized or remote project, then a connection should be
	available, and may be already be selected.
	<br>(To launch on the local machine, select <b>Local</b> for the connection type.)
	<br>You can create a new connection here by clicking the <b>New...</b> button.
	</p>
	<p>
	<p> </p>
	<li>
	<p>
	Once you select the connection, you may be prompted to open the connection if it is not already open. The contents of the <b>Resources</b>
	tab will then populate with controls that allow you to configure the necessary launch information for the target system.

	This area usually (depending on the target system configuration) includes further tabs to
	segment the other information you can provide.
	Fill in the information required by the target system configuration.
	<br>
	</p>
	</li>


	<!-- unused image<img src="images/03launchResTab.png"> -->

	</ol>

	<h2 id="application">Step 3: Complete the <b>Application</b> tab</h2>
	<p>The <b>Application tab</b> is used to specify the application
	(executable) to be launched.</p>
	<ol>
	<li>
	<p>Ensure that the correct <b>Parallel Project</b> is selected.
	This is the project containing the source and binaries for the
	application.</p>
	<p><br>
	</p>
	</li>

	<li>
	<p>Select the <b>Application Program</b> (executable) you wish to
	launch. If the target system connection points to a remote target, then
	this means the <i>path of the executable on the remote target</i>.
	Otherwise, for a connection pointing to the local machine, it should be
	the path to the executable within the project directory.
	<p>The <b>Browse...</b> button can be used to select the executable file.
	</p>
	<p><br>
	</p>
	</li>

	<li>
	<p>If you are running the application on a remote machine and the
	project was <i>built</i> locally, then you will have to copy the
	executable to the remote machine first. For this purpose, you may
	enable <b>Copy executable from local filesystem</b> and fill <b>Path
	to the local file</b> with the path to the executable within the project
	directory. This configures the launcher to copy the executable
	automatically to the remote target before starting the run/debug
	session. If not enabled, then you will need to copy the executable
	manually to the remote target.</p>

	<p><br>
	</p>
	</li>

	<p><img src="images/03launch6.png"></p>
	<p><br>
	</p>

	<li>
	<p>If you wish to have the output from all processes in a single
	output console, ensure that the <b>Display output from all
	processes in a console view</b> option is checked.</p>
	<p><br>
	</p>
	</li>

	</ol>

	<h2 id="arguments">Step 4: Complete the Arguments tab</h2>
	<p>The <b>Arguments tab</b> is used to specify any command-line
	arguments require by the parallel application and to modify the working
	directory.</p>
	<ol>
	<li>
	<p>Enter any arguments required by the application in the <b>Program arguments</b> field.
	This is a free form text field that will be
	passed to the application verbatim.</p>
	<p><br>
	</p>
	</li>
	<li>
	<p>Select a different working directory if desired. The working
	directory is set just prior to the application launch. Note that for a
	remote launch, this will be a directory on the remote machine.</p>
	</li>
	<p><img src="images/03launch3.png"></p>
	<p><br>
	</p>
	</ol>

	<h2 id="environment">Step 5: Complete the Environment tab</h2>
	<p>The <b>Environment tab</b> is used to specify any environment
	variables require by the parallel application or in order to run the
	application. The tab also provides the option to append the environment
	variables to the native environment, or to replace the native
	environment with the variables specified in the tab.</p>
	<ol>
	<li>
	<p>Use the <b>New...</b> button to create a new environment
	variable.</p>
	<p><br>
	</p>
	</li>
	<li>
	<p>Use the <b>Select...</b> button to import environment variables.</p>
	<p><br>
	</p>
	</li>
	<li>
	<p>Use the <b>Edit...</b> button to edit an environment variables.</p>
	<p><br>
	</p>
	</li>
	<li>
	<p>Use the <b>Remove...</b> button to remove an environment
	variables.</p>
	<p><br>
	</p>
	</li>
	<p><img src="images/03launch4.png"></p>
	<p><br>
	</p>
	</ol>

	<h2 id="synchronize">Step 6: Complete the <b>Synchronize</b> tab</h2>
	<p>The <b>Synchronize tab</b> is used in scenarios where the
	connection refers to a remote target. The tab specifies
	rules that describe files or directories that need to be copied (<i>uploaded</i>)
	to the remote target before starting the parallel application execution.
	Typically, this feature is used to upload dependencies (like dynamic
	libraries) or input data for test cases.</p>
	<p>Note that this may not be necessary for <a href="sync.html">Synchronized Projects</a>,
	in which files are copied back and forth to the remote system independently of the
	launching mechanism.

	<p>This tab also allows to specify rules describing files on the
	remote host to be retrieved back (<i>downloaded</i>) to the local host
	after the parallel application finishes execution, as output data
	produced by the parallel application execution.</p>

	<ol>
	<li>
	<p>If you wish to enable file transfer to the remote target, enable
	the <b>Upload rules enabled</b> or <b>Download rules enabled</b>
	according which direction of file transfer is required.</p>
	<p><br>
	</p>
	<p><img src="images/03launch7.png"></p>
	<p><br>
	</p>
	</li>

	<li>
	<p>In order to specify files to be copied from the local system to
	the remote target (<i>uploaded</i>), click on <b>New upload rule</b>.
	An upload rule lists a collection of files or directories that shall be
	copied into a single directory.</p>
	<p><br>
	</p>
	<p>If <b>Use directory from launch configuration</b> is enabled,
	then the listed files or directories are copied to the working
	directory specified on the launch configuration. At the moment, this is
	equivalent to same directory that contains the executable for the
	parallel application. If not enabled, they you must specify on <b>Remote
	directory</b> a path where the files or directories will be placed. The
	path may be absolute or relative (to the working directory).</p>
	<p><br>
	</p>
	<p>In the list underneath, <b>Selected file(s)</b>, add the files
	or directories that shall be copied. Click on <b>File(s)</b> to select
	one or more individual files from anywhere in the local file system.
	The <b>Directory</b> button works similarly to select a directory from
	the local file system. Directories are copied recursively. The <b>Workspace</b>
	button allows easily choosing files from the current workspace.</p>
	<p><br>
	</p>
	<p>On the bottom of the dialog, you may wish to set attributes to
	be applied to the copied files on the remote target. The options <b>Readonly</b>
	or <b>Executable</b> sets all copied files to read-only or executable
	on the remote host. If <b>Preserve time attributes</b> is enabled, then
	the remote copy will assume the same time stamp then the local file,
	otherwise, the time stamp will be the time the file was copied. Also,
	if the file already exists on the remote target, you may choose how to
	react in the <b>If file already exists</b> dropdown.</p>
	<p><br>
	</p>
	<p><img src="images/03launch8.png"></p>
	<p><br>
	</p>
	<p><br>
	</p>
	</li>
	<li>
	<p>In order to specify files to be copied from the remote target to
	the local system (<i>downloaded</i>) after the parallel application
	finishes, click on <b>New download rule</b>.</p>
	<p><br>
	</p>
	<p>An download rule lists a collection of files or directories that
	shall be copied into a single local directory. Its options are very
	similar to the upload rule.</p>
	<p><br>
	</p>
	<p><img src="images/03launch9.png"></p>
	<p><br>
	</p>
	</li>
	</ol>

	<h2 id="launch">Step 7: Launching the application</h2>
	<p>The launch configuration is now complete.
	Select the <b>Run</b> button (in the bottom right corner of the <b>Run Configurations</b> dialog to launch the job.
	Switch to the <b>System Monitoring</b> perspective
	to view the status of the job and inspect the program output.
	<p><a href="#top">Back to Top</a> \| <a href="toc.html">Back to Table of Contents</a>
	</body>
	</html>