<html>
<body>
Dear Rajeev,<br><br>
thanks for the advice; ch3:nemesis is what I'm using, as I read about it
in a previous message on the forum.<br><br>
I'll keep you informed about my tests.<br><br>
Regards.<br><br>
Gaetano<br><br>
At 03.07 16/01/2008, you wrote:<br>
<blockquote type=cite class=cite cite=""><font size=2 color="#0000FF">
Gaetano,<br>
I would recommend using the ch3:nemesis device. Configure with
--with-device=ch3:nemesis. It will use shared memory within a node and
TCP (or other network) between nodes. You can probably ignore those
timeout errors for now.<br>
</font> <br>
<font size=2 color="#0000FF">Rajeev<br>
</font><br>
<dl><hr>
<dd><font face="Tahoma" size=2>From:</b> Gaetano Bellanca
[<a href="mailto:gaetano.bellanca@unife.it" eudora="autourl">
mailto:gaetano.bellanca@unife.it</a>] <br>
<dd>Sent:</b> Tuesday, January 15, 2008 5:55 PM<br>
<dd>To:</b> Rajeev Thakur<br>
<dd>Subject:</b> RE: [MPICH] Confused about compiling/using mpich2<br>
</font><br>
<dd>Dear Rajeev,<br><br>
<dd>finally I was able to reach some good results in compiling/using
mpich2.<br>
<dd>The focused that the problem was the use of mpd. I was using and mpd
started by root, and this was not consistent with the correspondent
mpich2 compilation. I was using the one started at boot time, which was
an mpd coming from a previous version of mpich2 (and compiled with
probably also a different communication device).<br><br>
<dd>Now, I'm running successfully all the tests making make testing on my
home desktop (amd 5600+, single processor, dual core). My code run
using mpiexec -n 2 regularly.<br><br>
<dd>I test mpich2 in a different configuration, with a dual
processor / quad core linux box (Intel Xeon). In this case, running
make testing, I had some errors of this kind . All the other tests run
successfully (I can send you configure.log if you think it could be
useful)<br><br>
<dd><font color="#0000FF">Unexpected output in bcast3: mpiexec_campi17
(handle_sig_occurred 1123): job ending due to env var
MPIEXEC_TIMEOUT=180<br><br>
</font>
<dd>My code run successfully also on this machine, but I still have to
check the performances in terms of scalability.<br><br>
<dd>As in this case the computer has two network cards (but only one is
active in the moment), do you have any idea/suggestion in how to
configure/use mpich2 to have the best use of the 8 different
cores?<br><br>
<dd>Thanks very much for your help and assistance.<br><br>
<dd>Gaetano<br><br><blockquote type=cite class=cite cite=""><hr>
</blockquote><hr>
<dd><tt>Gaetano Bellanca - Department of Engineering - University of
Ferrara <br>
<dd>Via Saragat, 1 - 44100 - Ferrara -
ITALY
<br>
<dd>Voice (VoIP): +39 0532 974809
Fax: +39 0532 974870<br>
<dd><a href="mailto:gaetano.bellanca@unife.it" eudora="autourl">
mailto:gaetano.bellanca@unife.it</a> <br>
<hr>
</tt></blockquote>
<x-sigsep><p></x-sigsep>
</dl><hr>
<tt>Gaetano Bellanca - Department of Engineering - University of
Ferrara <br>
Via Saragat, 1 - 44100 - Ferrara -
ITALY
<br>
Voice (VoIP): +39 0532 974809 Fax:
+39 0532 974870<br>
<a href="mailto:gaetano.bellanca@unife.it" eudora="autourl">
mailto:gaetano.bellanca@unife.it</a> <br>
<hr>
</body>
</html>