CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   OF mpirun and parallel problem (http://www.cfd-online.com/Forums/openfoam-solving/123149-mpirun-parallel-problem.html)

heksel8i September 5, 2013 09:35

OF mpirun and parallel problem
 
Hey!

I had plenty of problems to get the parallel run working. I overcame many of them like installing scotch, setting PINC in settings.sh etc.

At the moment I get this kind of error report while running

>mpirun --mca btl ^openib --hostfile machines -np 2 simpleFoam -parallel

or

>mpirun --hostfile machines -np 2 simpleFoam -parallel

librdmacm: couldn't read ABI version.
librdmacm: assuming: 4
CMA: unable to get RDMA device list
--------------------------------------------------------------------------
WARNING: Failed to open "OpenIB-cma" [DAT_INTERNAL_ERROR:].
This may be a real error or it may be an invalid entry in the uDAPL
Registry which is contained in the dat.conf file. Contact your local
System Administrator to confirm the availability of the interfaces in
the dat.conf file.
--------------------------------------------------------------------------
librdmacm: couldn't read ABI version.
librdmacm: assuming: 4
CMA: unable to get RDMA device list
[hostname:32634] 1 more process has sent help message help-mpi-btl-udapl.txt / dat_ia_open fail
[hostname:32634] Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages

-------It stays here as long as I kill it------------

^Cmpirun: killing job...

--------------------------------------------------------------------------
mpirun noticed that process rank 0 with PID 32635 on node 'hostname' exited on signal 0 (Unknown signal 0).
--------------------------------------------------------------------------
2 total processes killed (some possibly by mpirun during cleanup)
mpirun: clean termination accomplished


What dat -file the warning means? Any idea?

My system is Linux SuSe 11.2 , OF 2.1.x, Third party package is installed. OF was built from the sources.

heksel8i September 10, 2013 10:10

Still having the same problem...:(

heksel8i September 11, 2013 05:33

Problem was about wrong command:

openib tag refers to the IB cards what I'm not apparently having, so in my case the working command is:

>mpirun --hostfile machines -np 2 --mca btl sm,self simpleFoam -parallel

Hopefully this helps someone in the future...


All times are GMT -4. The time now is 15:18.