Hi Pavan,<div><br></div><div>I tried mpiexec.hydra.</div><div><br></div><div>Even "mpiexec.hydra -np 1 examples/cpi" doesn't work. The program just hangs there with no output.</div><div><br></div><div>I tried "mpdcheck -v". It doesn't seem to give error info.</div>
<div><br></div><div>It could be a serious networking problem as you said. Because this is a single SMP machine, there may be some networking stuff that was not set up properly. Any suggestion on how to check the networking problem, so that I can contact our admin. Thanks!</div>
<div><br></div><div>--</div><div>Eric</div><div><br><div class="gmail_quote">On Sun, Dec 13, 2009 at 1:24 PM, Pavan Balaji <span dir="ltr"><<a href="mailto:balaji@mcs.anl.gov">balaji@mcs.anl.gov</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div class="im"><br>
On 12/13/2009 11:44 AM, Eric Sun wrote:<br>
> FYI, I needed to edit src/openpa/src/Makefile to change "AR = ar" to "AR<br>
> = ar -X64". So, the AR="ar -X64" didn't apply to this directory. Also,<br>
> somehow I need to edit the main Makefile to add full path for "INSTALL =<br>
> confdb/install-sh". Everything else is fine.<br>
<br>
</div>What errors were you seeing without these changes? This should work out<br>
of the box, AFAICT.<br>
<div class="im"><br>
> I'm now doing a first test using "mpiexec -l -n 4 hostname". For n=1, it<br>
> always works. n=2 occasionally fails. n=3 occasionally successes. n=4<br>
> always gives the following message:<br>
</div>> <a href="http://mpiexec_pls01.ccni.rpi.edu" target="_blank">mpiexec_pls01.ccni.rpi.edu</a> <<a href="http://mpiexec_pls01.ccni.rpi.edu" target="_blank">http://mpiexec_pls01.ccni.rpi.edu</a>> (mpiexec<br>
<div class="im">> 392): no msg recvd from mpd when expecting ack of request<br>
<br>
</div>This usually points to some problem with the networking setup. Can you<br>
use the mpdcheck utility to see if it can find any? Do you have any<br>
firewall for some of the ports?<br>
<br>
Alternatively, you can try using the Hydra process manager<br>
(mpiexec.hydra instead of mpiexec). Here's some documentation:<br>
<a href="http://wiki.mcs.anl.gov/mpich2/index.php/Using_the_Hydra_Process_Manager" target="_blank">http://wiki.mcs.anl.gov/mpich2/index.php/Using_the_Hydra_Process_Manager</a><br>
<br>
If there's a serious networking setup problem, both mpd and Hydra will<br>
give you errors, of course. So, you'll need to make sure your setup is<br>
fine in either case.<br>
<br>
-- Pavan<br>
<font color="#888888"><br>
--<br>
</font><div><div></div><div class="h5">Pavan Balaji<br>
<a href="http://www.mcs.anl.gov/~balaji" target="_blank">http://www.mcs.anl.gov/~balaji</a><br>
_______________________________________________<br>
mpich-discuss mailing list<br>
<a href="mailto:mpich-discuss@mcs.anl.gov">mpich-discuss@mcs.anl.gov</a><br>
<a href="https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss" target="_blank">https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss</a><br>
</div></div></blockquote></div><br><br clear="all"><br>-- <br>Yiyang Sun<br>Department of Physics, Applied Physics, and Astronomy<br>Rensselaer Polytechnic Institute<br>Troy, NY 12180, USA<br>
</div>