<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=us-ascii">
<META content="MSHTML 6.00.6000.16735" name=GENERATOR></HEAD>
<BODY>
<DIV dir=ltr align=left><FONT face=Arial color=#0000ff size=2><SPAN
class=453305216-09122008>Hi,</SPAN></FONT></DIV>
<DIV dir=ltr align=left><FONT face=Arial color=#0000ff size=2><SPAN
class=453305216-09122008> Can you send us the code for LaplaceSolver.exe
(or a test program that runs slow) ? How long do you have to wait for the
program to complete execution ? Are the machines connected using Ethernet
?</SPAN></FONT></DIV>
<DIV dir=ltr align=left><FONT face=Arial color=#0000ff size=2><SPAN
class=453305216-09122008></SPAN></FONT> </DIV>
<DIV dir=ltr align=left><FONT face=Arial color=#0000ff size=2><SPAN
class=453305216-09122008>Regards,</SPAN></FONT></DIV>
<DIV dir=ltr align=left><FONT face=Arial color=#0000ff size=2><SPAN
class=453305216-09122008>Jayesh</SPAN></FONT></DIV><BR>
<DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left>
<HR tabIndex=-1>
<FONT face=Tahoma size=2><B>From:</B> Seifer Lin [mailto:seiferlin@gmail.com]
<BR><B>Sent:</B> Monday, December 08, 2008 9:11 PM<BR><B>To:</B> Jayesh
Krishna<BR><B>Cc:</B> mpich-discuss@mcs.anl.gov<BR><B>Subject:</B> Re:
[mpich-discuss] MPICH2 1.0.6p1 & Windows HPC Server 2008
(badperformance)<BR></FONT><BR></DIV>
<DIV></DIV>Hi:<BR> <BR>After that I put all 4 nodes in the same domain
(originally they are in the same workgroup)<BR>For the MPI_Barrier(...)
function, version 1.0.6p1 and 1.0.8 both return at once.<BR>But for another
simple MPI program (Laplace equation solver), both 1.0.6p1 and 1.0.8 run very
SLOW!!!!<BR> <BR>I have noticed that if the command is like<BR>mpiexec
-hosts 2 <A href="http://192.168.1.1">192.168.1.1</A> <A
href="http://192.168.1.2">192.168.1.2</A> <A
href="file://192.168.1.1/shared/LaplaceSolver.exe">\\192.168.1.1\shared\LaplaceSolver.exe</A><BR>It
runs very SLOW!!! (the processes are located at different
machines)<BR> <BR>Another command is like<BR>mpiexec -hosts 2 <A
href="http://192.168.1.1">192.168.1.1</A> <A
href="http://192.168.1.1">192.168.1.1</A> <A
href="file://192.168.1.1/shared/LaplaceSolver.exe">\\192.168.1.1\shared\LaplaceSolver.exe</A><BR>It
runs at normal fast speed! (the processes are located at the SAME
machine)<BR> <BR> <BR>I think this may due to the strict policies on
network transfer of Windows HPC Server 2008 (and Vista also)<BR>Do you have any
solution to this? thank you!<BR> <BR>regards,<BR> <BR>Seifer
Lin<BR><BR>
<DIV class=gmail_quote>2008/12/8 Jayesh Krishna <SPAN dir=ltr><<A
href="mailto:jayesh@mcs.anl.gov">jayesh@mcs.anl.gov</A>></SPAN><BR>
<BLOCKQUOTE class=gmail_quote
style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">
<DIV>
<DIV dir=ltr align=left><FONT face=Arial color=#0000ff
size=2><SPAN>Hi,</SPAN></FONT></DIV>
<DIV dir=ltr align=left><FONT face=Arial color=#0000ff size=2><SPAN> Can
you try out the latest stable version (1.0.8) of MPICH2 (<A
href="http://www.mcs.anl.gov/research/projects/mpich2/downloads/index.php?s=downloads"
target=_blank>http://www.mcs.anl.gov/research/projects/mpich2/downloads/index.php?s=downloads</A>)?</SPAN></FONT></DIV>
<DIV dir=ltr align=left><FONT face=Arial color=#0000ff
size=2><SPAN></SPAN></FONT> </DIV>
<DIV dir=ltr align=left><FONT face=Arial color=#0000ff
size=2><SPAN>Regards,</SPAN></FONT></DIV>
<DIV dir=ltr align=left><FONT face=Arial color=#0000ff
size=2><SPAN>Jayesh</SPAN></FONT></DIV><BR>
<DIV lang=en-us dir=ltr align=left>
<HR>
<FONT face=Tahoma size=2><B>From:</B> <A
href="mailto:mpich-discuss-bounces@mcs.anl.gov"
target=_blank>mpich-discuss-bounces@mcs.anl.gov</A> [mailto:<A
href="mailto:mpich-discuss-bounces@mcs.anl.gov"
target=_blank>mpich-discuss-bounces@mcs.anl.gov</A>] <B>On Behalf Of
</B>Seifer Lin<BR><B>Sent:</B> Monday, December 08, 2008 12:47
AM<BR><B>To:</B> <A href="mailto:mpich-discuss@mcs.anl.gov"
target=_blank>mpich-discuss@mcs.anl.gov</A><BR><B>Subject:</B> [mpich-discuss]
MPICH2 1.0.6p1 & Windows HPC Server 2008
(badperformance)<BR></FONT><BR></DIV>
<DIV>
<DIV></DIV>
<DIV class=Wj3C7c>
<DIV></DIV>
<DIV>Hi everyone:</DIV>
<DIV> </DIV>
<DIV>I have a cluster of 4 nodes, all of them are with Windows HPC server 2008
installed.</DIV>
<DIV>I make all of the 4 nodes in the same workgroup. I use MPICH2 1.0.6p1
from Argonne Lab.</DIV>
<DIV>And then</DIV>
<DIV>1. firewall of all 4 nodes are turned off</DIV>
<DIV>2. UAC (User Account Control) of all 4 nodes are turned off</DIV>
<DIV>3. I start smpd.exe (1.0.6p1 x64) in all the 4 nodes</DIV>
<DIV> </DIV>
<DIV>And I run a very simple MPI program (test_mpich2.exe)</DIV>
<DIV> </DIV>
<DIV>#include "mpi.h"<BR>#include <iostream></DIV>
<DIV>int main(int argc, char **argv)<BR>{<BR> int cpuid,
ncpu;<BR> MPI_Init(&argc,
&argv);<BR> MPI_Comm_size(MPI_COMM_WORLD,
&ncpu);<BR> MPI_Comm_rank(MPI_COMM_WORLD,
&cpuid);</DIV>
<DIV> printf("NCPU:%d, CPUID:%d\n", ncpu,
cpuid);<BR> fflush(stdout);</DIV>
<DIV> printf("start barrier\n");
fflush(stdout);<BR>
MPI_Barrier(MPI_COMM_WORLD);<BR> printf("end barrier\n");
fflush(stdout);</DIV>
<DIV> MPI_Finalize();</DIV>
<DIV> return 0;</DIV>
<DIV>}</DIV>
<DIV> </DIV>
<DIV>The command is </DIV>
<DIV>mpiexec -hosts 2 <A href="http://192.168.1.1/"
target=_blank>192.168.1.1</A> <A href="http://192.168.1.2/"
target=_blank>192.168.1.2</A>
<A>\\192.168.1.1\shared\test_mpich2.exe</A></DIV>
<DIV> </DIV>
<DIV>And the MPI_Barrier(...) function costs 10 seconds to return !!!!!</DIV>
<DIV> </DIV>
<DIV>If the same code is running on a Windows XP cluster, MPI_Barrier(...)
returns at once!</DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV>Does anyone know how to solve this problem on Windows HPC Server 2008 ?
(Windows Vista has the same problem, too)</DIV>
<DIV> </DIV>
<DIV>regards,</DIV>
<DIV> </DIV>
<DIV>Seifer Lin</DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV> </DIV></DIV></DIV></DIV></BLOCKQUOTE></DIV><BR></BODY></HTML>