<span style="color: rgb(51, 102, 255);"> <span style="color: rgb(51, 102, 255);">Hi all,</span></span><br style="color: rgb(51, 102, 255);"><br style="color: rgb(51, 102, 255);"><span style="color: rgb(51, 102, 255);"> I'm doing the Tight MPICH2 (not MPICH) Integration with SGE on a cluster with, dual core dual AMD64 opteron processor.
</span><br style="color: rgb(51, 102, 255);"><span style="color: rgb(51, 102, 255);"> </span><br style="color: rgb(51, 102, 255);"><span style="color: rgb(51, 102, 255);"> Followed the sun document located at:</span><br style="color: rgb(51, 102, 255);">
<br style="color: rgb(51, 51, 255);"><span style="color: rgb(51, 51, 255);"> <a href="http://gridengine.sunsource.net/howto/mpich2-integration/mpich2-integration.html">http://gridengine.sunsource.net/howto/mpich2-integration/mpich2-integration.html
</a></span><br style="color: rgb(102, 102, 204);"><br style="color: rgb(51, 102, 255);"><span style="color: rgb(51, 102, 255);"> The document explains following three kinds of TI: </span>
<p class="MsoListParagraph" style="margin-left: 22.5pt; text-indent: -0.25in; color: rgb(102, 51, 255);"><span style=""><span style=""> <span style="font-family: "Times New Roman"; font-style: normal; font-variant: normal; font-weight: normal; font-size: 7pt; line-height: normal; font-size-adjust: none; font-stretch: normal;">
</span></span></span>Tight
Integration(TI) using Process Manager(PM): gforker<br><span style=""><span style=""><span style="font-family: "Times New Roman"; font-style: normal; font-variant: normal; font-weight: normal; font-size: 7pt; line-height: normal; font-size-adjust: none; font-stretch: normal;">
</span></span></span>TI
using PM: SMPD – Daemonless<br><span style=""><span style=""><span style="font-family: "Times New Roman"; font-style: normal; font-variant: normal; font-weight: normal; font-size: 7pt; line-height: normal; font-size-adjust: none; font-stretch: normal;">
</span></span></span>TI
using PM: SMPD – Daemonbased</p>
<p style="color: rgb(102, 51, 255);" class="MsoNormal">I did the TI with gforker and tested it successfully.</p><p class="MsoNormal"><br><span style="color: rgb(51, 102, 255);">But failed to do TI with daemonless-SMPD. </span>
<br style="color: rgb(51, 102, 255);"></p><p style="color: rgb(51, 102, 255);" class="MsoNormal">Let me explain what I did.</p><p style="color: rgb(51, 102, 255);" class="MsoNormal">Installed the MPICH2 with smpd configuration.
</p><p style="color: rgb(51, 102, 255);" class="MsoNormal">The sge is installed at: /opt/gridengine</p><p style="color: rgb(51, 102, 255);" class="MsoNormal">And created MPICH2-SM folder in /opt/gridengine/mpi by referring the following lines from the document
</p><pre style="font-family: arial,sans-serif; color: rgb(153, 0, 0);">start_proc_args /usr/sge/mpich2_smpd_rsh/startmpich2.sh -catch_rsh $pe_hostfile<br>stop_proc_args /usr/sge/mpich2_smpd_rsh/stopmpich2.sh</pre><p style="color: rgb(51, 102, 255);" class="MsoNormal">
Copied the startmpi.sh, stopmpi.sh from /opt/gridengine/mpi to /opt/gridengine/mpi/MPICH2-SM dir, because nothing has given in the doc what to include in these scripts.</p><span style="color: rgb(51, 102, 255);">Using qmon, created MPICH2-GF pe
</span>.<br><br><span style="color: rgb(255, 102, 102);"># qconf -sp MPICH2-SM</span><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">pe_name MPICH2-SM</span><br style="color: rgb(255, 102, 102);">
<span style="color: rgb(255, 102, 102);">slots 999</span><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">user_lists rootuserset</span><br style="color: rgb(255, 102, 102);">
<span style="color: rgb(255, 102, 102);">xuser_lists NONE</span><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">start_proc_args /opt/gridengine/mpi/MPICH2-SM/startmpich2sm.sh</span>
<br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">stop_proc_args /opt/gridengine/mpi/MPICH2-SM/stopmpich2sm.sh</span><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">
allocation_rule $round_robin</span><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">control_slaves FALSE</span><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">
job_is_first_task TRUE</span><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">urgency_slots min</span><br><br><span style="color: rgb(51, 102, 255);">Added this PE to default queue all.q
.</span><br style="color: rgb(51, 102, 255);"><br style="color: rgb(51, 102, 255);"><span style="color: rgb(51, 102, 255);">Then submitted the job with following script</span>:<br><br><span style="color: rgb(255, 102, 102);">
# cat sgeSM.sh</span><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">#!/bin/sh</span><br style="color: rgb(255, 102, 102);"><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">
#$ -cwd</span><br style="color: rgb(255, 102, 102);"><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">#$ -pe MPICH2-SM 4</span><br style="color: rgb(255, 102, 102);"><br style="color: rgb(255, 102, 102);">
<span style="color: rgb(255, 102, 102);">#$ -e msge2.Err</span><br style="color: rgb(255, 102, 102);"><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">#$ -o msge2.out</span><br style="color: rgb(255, 102, 102);">
<br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">#$ -v MPI_HOME=/opt/MPI_LIBS/MPICH2-GNU/MPICH2-SM/bin</span><br style="color: rgb(255, 102, 102);"><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">
#$ -v MEME_DIRECTORY=/opt/MEME-MAX</span><br style="color: rgb(255, 102, 102);"><br style="color: rgb(255, 102, 102);"><span style="color: rgb(255, 102, 102);">$MPI_HOME/mpiexec -np 4 -machinefile /root/MFM /opt/MEME-MAX/bin/meme_p /opt/MEME-MAX/NCCS/samevivo_sample.txt -dna -mod tcm -nmotifs 10 -nsites 100 -minw 5 -maxw 50 -revcomp -text -maxsize 200500
</span><br><br><span style="color: rgb(51, 102, 255);">It gave following error:</span><br><br># cat msge2.Err<br><br><span style="color: rgb(255, 0, 0);">startmpich2sm.sh: got wrong number of arguments</span><br style="color: rgb(255, 0, 0);">
<span style="color: rgb(255, 0, 0);">rm: cannot remove `/tmp/92.1.all.q/machines': No such file or directory</span><br style="color: rgb(255, 0, 0);"><span style="color: rgb(255, 0, 0);">rm: cannot remove `/tmp/92.1.all.q/rsh': No such file or directory
</span><br><br><span style="color: rgb(51, 102, 255);">I guess the problem might be with the scripts startmpich2sm.sh and
stopmpich2sm.sh.</span> <br><br style="color: rgb(51, 102, 255);"><span style="color: rgb(51, 102, 255);">Can any one guide me to resolve this issue..<br><br><font style="font-family: trebuchet ms,sans-serif;" size="2">Thanks & Regards,
</font><br><span style="font-family: trebuchet ms,sans-serif;">Sangamesh</span><br style="font-family: trebuchet ms,sans-serif;"><span style="font-family: trebuchet ms,sans-serif;">HPC Engineer</span><br></span>
<p class="MsoNormal"> </p>
<p class="MsoNormal"><br></p>