We run pnetcdf in a hybrid environment in cesm, we have not used the particular combination of intel/openmp but have not noted any problems on the systems we do use.<br><br>Jim<br><br><div class="gmail_quote">On Mon, Oct 10, 2011 at 10:04 AM, Wei-keng Liao <span dir="ltr"><<a href="mailto:wkliao@ece.northwestern.edu">wkliao@ece.northwestern.edu</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">Hi, John,<br>
<br>
We have not yet tested pnetcdf under the MPI-OpenMP hybrid programming environment.<br>
Could you please describe more in detailed on how your program makes pnetcdf calls?<br>
Or provide us a sample program that can re-produce the hanging?<br>
Did you see the same hanging happened when using a different file system?<br>
Have you tried run one OpenMP thread per MPI process and seen the same problem?<br>
<font color="#888888"><br>
Wei-keng<br>
</font><div><div></div><div class="h5"><br>
<br>
On Oct 10, 2011, at 7:17 AM, John Michalakes wrote:<br>
<br>
> I'm running into a case where WRF reads and writes through pNetCDF okay when compiled and running in MPI-only mode, but hangs when running hybrid MPI/OpenMP mode. This is on a Lustre file system. Application executable is compiled and linked using Intel compilers and OpenMPI. The system is NREL's RedMesa Sun Blade Linux cluster (a part of Sandia's Red Sky: <a href="http://www.top500.org/system/10188" target="_blank">http://www.top500.org/system/10188</a> ). Are there reports of problems combining pNetCDF on Lustre with threaded MPI applications? Thanks, -John<br>
><br>
> --<br>
> John Michalakes<br>
> National Renewable Energy Laboratory<br>
> 1617 Cole Blvd.<br>
> Golden, Colorado 80401<br>
> Phone: <a href="tel:303-275-4297" value="+13032754297">303-275-4297</a><br>
> Fax: <a href="tel:303-275-4091" value="+13032754091">303-275-4091</a><br>
> <a href="http://www.nrel.gov/csc/staff_michalakes.html" target="_blank">http://www.nrel.gov/csc/staff_michalakes.html</a><br>
><br>
<br>
</div></div></blockquote></div><br><br clear="all"><br>-- <br>Jim Edwards<br><br><font>CESM Software Engineering Group<br>National Center for Atmospheric Research<br>Boulder, CO <br>303-497-1842<br></font><br>