<?xml version="1.0" encoding="UTF-8"?>
<!-- Generated on Wed, 25 Nov 2009 12:08:57 -0800 -->
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <atom:link href="http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/feed" rel="self" type="application/rss+xml" />
    <title>Intel Software Network - <![CDATA[ Intel® Clusters and HPC Technology ]]> feed</title>
    <link>http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology</link>
    <description></description>
    <language>en-us</language>
    <item>
      <title>Misterious BLACS</title>
      <description><![CDATA[ Hello.<br />I use BLACS and ScaLapack to inverse dense matrix.<br />And I got some troubles.<br /><br />I set up nprow and npcol for block-cyclic distribution for example 2, 2.<br />and run this routine<br />
<pre name="code" class="cpp">nprow = 2;
npcol=2;
Cblacs_get(0,0,&amp;ictxt);
Cblacs_gridinit(&amp;ictxt,"R",nprow,npcol);
Cblacs_gridinfo(ictxt,&amp;nprow,&amp;npcol,&amp;myrow,&amp;mycol);
</pre>
<br />The stange thing is that after Cblacs_gridinfo(..) nprow is changed as 0.<br />Why is it changed?<br />Because it is changed to 0, numroc(..) called after that generates error.<br />Could you give me some ideas?<br />Thanks in advance.<br /><br />ps. Is there another method to inverse matrix without BLACS? ]]></description>
      <link>http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/70097/</link>
      <pubDate>Tue, 24 Nov 2009 00:15:29 -0800</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/70097/</guid>
      <category>Intel® Cluster Ready</category>
      <category>Tools</category>
    </item>
    <item>
      <title>Using new unified installer for Cluster Tools 3.2.2 - Slow </title>
      <description><![CDATA[ I downloaded the new cluster toolkit 3.2.2.<br /><br />Little 128 node cluster with master node.<br /><br />The install scipt installs the rpms onto the master node.<br />Then it is taking quite awhile to 'Analyze node configuration'<br />Watching it's progress: It's taken 46 minutes to process 23 nodes.<br />I'm guessing the installer is attempting to discern network topology, or if any compute nodes share the install directory.<br />Though, when prompted by the installer, I selected 'non-shared' install, i.e. /opt/intel is local on each compute node.<br /> ]]></description>
      <link>http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69954/</link>
      <pubDate>Tue, 17 Nov 2009 12:41:59 -0800</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69954/</guid>
      <category>Intel® Cluster Ready</category>
      <category>Tools</category>
    </item>
    <item>
      <title>Running mpi w/o mpiexec/mpirun</title>
      <description><![CDATA[ <span style="font-family: monospace; font-size: medium; white-space: pre;">
<pre>I can compile and run a mpi application using mpiexec without any difficulties. </pre>
<pre>What I want to do is to run a mpi application not in Terminal, I mean without using mpiexec or mpirun.</pre>
<pre>Is this doable? If it is, how can I setup mpd daemon and other necessary stuffs in source level?</pre>
<pre><span style="font-family: verdana, sans-serif;">Thanks. </span></pre>
<pre><br /></pre>
<pre>==========</pre>
<pre><span style="font-family: Verdana, Arial, Helvetica, sans-serif;">hello.c</span></pre>
<pre>==========</pre>
<pre>#<b><span style="color: #5f9ea0;">include</span></b> <b><span style="color: #bc8f8f;">&lt;stdio.h&gt;</span></b></pre>
<pre>#<b><span style="color: #5f9ea0;">include</span></b> <b><span style="color: #bc8f8f;">&lt;mpi.h&gt;</span></b></pre>
<pre><span style="font-family: Verdana, Arial, Helvetica, sans-serif;"><br /></span></pre>
<pre>static int run()</pre>
<pre>{
	<b><span style="color: #228b22;">int</span></b> node;</pre>
<pre>	MPI_Init(xxx, xxx); /* TODO */</pre>
<pre>	MPI_Comm_rank(MPI_COMM_WORLD, &amp;node);
	printf(<b><span style="color: #bc8f8f;">"Hello World from Node (%d)\n"</span></b>, node);
	MPI_Finalize();
	<b><span style="color: #a020f0;">return</span></b> 0;	
}
</pre>
<pre><span style="font-family: Verdana, Arial, Helvetica, sans-serif;"><br /></span></pre>
<pre>int main (int argc, char **argv)</pre>
<pre><span style="font-family: Verdana, Arial, Helvetica, sans-serif;">{</span></pre>
<pre>run();</pre>
<pre>return 0;</pre>
<pre>}</pre>
<div><span style="font-family: Verdana, Arial, Helvetica, sans-serif; font-size: x-small;"><span style="font-size: 10px;"><br /></span></span></div>
</span> ]]></description>
      <link>http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69784/</link>
      <pubDate>Mon, 09 Nov 2009 19:28:50 -0800</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69784/</guid>
      <category>Intel® Cluster Ready</category>
      <category>Tools</category>
    </item>
    <item>
      <title>HPCC benchmark failing with Intel MPI</title>
      <description><![CDATA[ Hello,<br /><br />I 've compiled the HPCC benchmark suite (http://icl.cs.utk.edu/hpcc/) with Intel MPI, but am facing the following run-time problem:<br /><br />[bart@head 2x8]$ /share/intel/impi/3.2.1.009/bin64/mpirun -f 1.nodelist -n 16 -r ssh ./hpcc<br />node002:27686:  reg_mr Cannot allocate memory<br />node001:2835:  reg_mr Cannot allocate memory<br />node002:27684:  reg_mr Cannot allocate memory<br />node002:27679:  reg_mr Cannot allocate memory<br />node001:2835:  reg_mr Cannot allocate memory<br />node002:27686:  reg_mr Cannot allocate memory<br />node001:2834:  reg_mr Cannot allocate memory<br />node002:27679:  reg_mr Cannot allocate memory<br />node001:2834:  reg_mr Cannot allocate memory<br />node002:27686:  reg_mr Cannot allocate memory<br />node001:2839:  reg_mr Cannot allocate memory<br />node001:2835:  reg_mr Cannot allocate memory<br />node002:27679:  reg_mr Cannot allocate memory<br />node002:27681:  reg_mr Cannot allocate memory<br />node001:2835:  reg_mr Cannot allocate memory<br />node002:27681:  reg_mr Cannot allocate memory<br />node002:27682:  reg_mr Cannot allocate memory<br />node002:27682:  reg_mr Cannot allocate memory<br />node002:27685:  reg_mr Cannot allocate memory<br />node002:27685:  reg_mr Cannot allocate memory<br />node002:27680:  reg_mr Cannot allocate memory<br />node001:2833:  reg_mr Cannot allocate memory<br />node001:2838:  reg_mr Cannot allocate memory<br />node002:27680:  reg_mr Cannot allocate memory<br />node001:2833:  reg_mr Cannot allocate memory<br />node001:2838:  reg_mr Cannot allocate memory<br />node002:27682:  reg_mr Cannot allocate memory<br />node001:2836:  reg_mr Cannot allocate memory<br />node002:27682:  reg_mr Cannot allocate memory<br />node001:2836:  reg_mr Cannot allocate memory<br />node002:27682:  reg_mr Cannot allocate memory<br />node001:2836:  reg_mr Cannot allocate memory<br />node002:27682:  reg_mr Cannot allocate memory<br />node001:2836:  reg_mr Cannot allocate memory<br />node002:27684:  reg_mr Cannot allocate memory<br />node001:2839:  reg_mr Cannot allocate memory<br />node001:2838:  reg_mr Cannot allocate memory<br />node001:2838:  reg_mr Cannot allocate memory<br />node001:2835:  reg_mr Cannot allocate memory<br />node002:27684:  reg_mr Cannot allocate memory<br />node001:2835:  reg_mr Cannot allocate memory<br />register failed 196608 [10] error(0x30000): OpenIB-cma: DAT_INSUFFICIENT_RESOURCES: <br /><br />node001:2835:  reg_mr Cannot allocate memory<br />[4:node002][rdma_iba.c:220] Intel MPI fatal error: DTO operation posted for [10:node001] completed with error. status=0x1. cookie=0x4000a<br />rank 10 in job 1  head_46465   caused collective abort of all ranks<br /> exit status of rank 10: return code 1 <br /><br /><br />The benchmark fails at the start of the HPL part of the benchmark. Any suggestions for fixes would be most appreciated.<br /><br />Thanks,<br />Bart<br /> ]]></description>
      <link>http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69750/</link>
      <pubDate>Sun, 08 Nov 2009 13:41:10 -0800</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69750/</guid>
      <category>Intel® Cluster Ready</category>
      <category>Tools</category>
    </item>
    <item>
      <title>Solaris Compiler</title>
      <description><![CDATA[ Is there an Intel compiler for Solaris (ultimatly supporting an MPI distributed system)? ]]></description>
      <link>http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69640/</link>
      <pubDate>Wed, 04 Nov 2009 05:16:04 -0800</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69640/</guid>
      <category>Intel® Cluster Ready</category>
      <category>Tools</category>
    </item>
    <item>
      <title>Problems installing cluster toolkit under windows server 2008</title>
      <description><![CDATA[ Hi all,
<div><span style="font-family: Verdana, Arial, Helvetica, sans-serif;"><br /></span></div>
<div>during the installation of Cluster Toolkit for Windows HPC Server 2008 I had some "noncritical errors":</div>
<div><br /></div>
<div style="padding-left: 30px;">An error occured in Merge Module SUBST.</div>
<div style="padding-left: 30px;">Installation will be continued but NOT all of the following files will be configured appropriately:</div>
<div style="padding-left: 30px;"><br /></div>
<div style="padding-left: 30px;">C:\Program Files (x86)\Intel\ICTCE\3.2.1.015\Compiler\bin\ifortvars.bat</div>
<div style="padding-left: 30px;"><br /></div>
<div style="padding-left: 30px;"><br /></div>
<div>etc.</div>
<div><br /></div>
<div>After installing Cluster Tookit I have checked build environment with the following output:</div>
<div><br /></div>
<div>
<div style="padding-left: 30px;">The system cannot find the file specified.</div>
<div style="padding-left: 30px;">The system cannot find the file specified.</div>
<div style="padding-left: 30px;">The system cannot find the file specified.</div>
<div style="padding-left: 30px;">The system cannot find the file specified.</div>
<div style="padding-left: 30px;">The system cannot find the file specified.</div>
<div style="padding-left: 30px;">The system cannot find the file specified.</div>
<div style="padding-left: 30px;"><br /></div>
<div style="padding-left: 30px;">The system cannot find the file specified.</div>
<div style="padding-left: 30px;">Copyright (C) 2007-2009 Intel Corporation. All rights reserved.</div>
<div style="padding-left: 30px;"><br /></div>
<div style="padding-left: 30px;">The filename, directory name, or volume label syntax is incorrect.</div>
<div style="padding-left: 30px;">The filename, directory name, or volume label syntax is incorrect.</div>
<div style="padding-left: 30px;"><br /></div>
<div style="padding-left: 30px;">The system cannot find the file specified.</div>
<div style="padding-left: 30px;">Copyright (C) 2009 Intel Corporation. All rights reserved.</div>
</div>
<div style="padding-left: 30px;"><br /></div>
<div style="padding-left: 30px;"><br /></div>
<div>After checking .bat files I have seen the following variable:</div>
<div><br /></div>
<div>&lt;INSTALLDIR&gt;</div>
<div><br /></div>
<div>the meaning of which is not clear for me.</div>
<div><br /></div>
<div><br /></div>
<div>I am triying to install this version of Toolkit:</div>
<div><br /></div>
<div>
<div style="padding-left: 30px;">Package ID: w_ict_p_3.2.1.015</div>
<div style="padding-left: 30px;">Package Contents: Intel(R) Cluster Toolkit Compiler Edition 3.2.1 for Windows*<br /><br /></div>
Windows version:<br /><br />Windows Server 2008 HPC Edition with SP2<br /><br />
<div><br />Any ideas?</div>
<div><br /></div>
<div>Regards,</div>
<div>Igor</div>
</div> ]]></description>
      <link>http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69582/</link>
      <pubDate>Mon, 02 Nov 2009 06:38:04 -0800</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69582/</guid>
      <category>Intel® Cluster Ready</category>
      <category>Tools</category>
    </item>
    <item>
      <title>.nsfxxxx files generated at execution with mpi</title>
      <description><![CDATA[ Hello<br /><br />When I use my software running with Intel MPI 2.0 (sorry can't update MPI for the moment), I can see numerous temporary files named .nfsXXXXXXXX which worried my customers . The process running with MPI needs 4 files in input and generates 4 files in output, all these files are read and written on a remote disk via the network and the I use 4 cpus. <br />Can you tell me more about these files, is there a relationship between the number of .nsfxxx files and the files in input/output? ]]></description>
      <link>http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69573/</link>
      <pubDate>Mon, 02 Nov 2009 00:02:13 -0800</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69573/</guid>
      <category>Intel® Cluster Ready</category>
      <category>Tools</category>
    </item>
    <item>
      <title>Intel Fortran debugger and MPICH</title>
      <description><![CDATA[ Hi,<br />I want to debug this parallel Fortran program that I'm trying to make run on a Linux-type cluster using the Intel Fortran compiler (v.9.1.045) and debugger (v.9.1-28) and MPICH2.1.2. While building the executable with the -g option is straightforward, but when trying to invoke the compiler it crashes with this message:<br />$ idb -parallel mpiexec -machinefile machines -n 4 ./stagyympi <br />Intel(R) Debugger for applications running on IA-32, Version 9.1-28, Build 20070305<br />execve failed: No such file or directory<br />Error: could not start debuggee<br />Could not start process for mpiexec<br />No image loaded ... Recovering ... <br /><br />Any ideas what may be wrong here?<br />Thomas ]]></description>
      <link>http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69512/</link>
      <pubDate>Thu, 29 Oct 2009 12:47:49 -0700</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69512/</guid>
      <category>Intel® Cluster Ready</category>
      <category>Tools</category>
    </item>
    <item>
      <title>how to distribute data to different computing nodes, using MPI</title>
      <description><![CDATA[ Dear all, I am started to using MPI for a simple data decomposition of a 2-D domain. Assuming that I am using 2 computing nodes, each having 8 processors, I want to make message pass only between the two nodes, while inside each node, all processors can access their shared memory. <br /> After calling MPI_rank and receiving 0~15 for processor rank,  how can I know to which node a processor belongs? Do processors with rank 0 to 7 belong to computing node1 and 8 to 15 belong to computing node 2?<br />By the way, the machine is using a Windows compute cluster pack, MS_MPI library. Are there some options to know the rank of each processor on each node, if multiple nodes are used?<br /><br />Thanks and regards<br />skiff<br /> ]]></description>
      <link>http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69469/</link>
      <pubDate>Wed, 28 Oct 2009 07:53:07 -0700</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69469/</guid>
      <category>Intel® Cluster Ready</category>
      <category>Tools</category>
    </item>
    <item>
      <title>How does Intel MPI handle network failures</title>
      <description><![CDATA[ Hi all, <br /><br />I am new to the forum and have a question regarding network failures and MPI applications (specifically using the Intel MPI binding).<br /><br />What happens if I have a a number of processes running on a cluster, and someone unplugs a network cable? As far as I have read, the MPI processes gets terminated immediately. How can I circumvent this, say by using some sort of a WAIT or TIMEOUT command if a network fault is detected, so that they can see if maybe they can again recover after a number of (set) seconds?<br /><br />Any help would be very much appreciated! ]]></description>
      <link>http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69428/</link>
      <pubDate>Tue, 27 Oct 2009 03:53:40 -0700</pubDate>
      <guid isPermaLink="true">http://software.intel.com/en-us/forums/intel-clusters-and-hpc-technology/topic/69428/</guid>
      <category>Intel® Cluster Ready</category>
      <category>Tools</category>
    </item>
  </channel></rss>