Hi Dominik,

Yes, that is what I fixed for the Open MPI test on OSCAR 5.1.
http://svn.oscar.openclustergroup.org/trac/oscar/changeset/6724

How could I forget about this!! Thanks for your post and contribution.

Regards,

- DongInn


Dominik Schips wrote:
> Hi,
> 
> just for the protocol. I got everything working now.
> 
> I read a little bit more about OpenMPI and found in the OpenMPI FAQ that
> the handling about the nodefile has changed between 1.1 and 1.2.
> 
> http://www.open-mpi.org/faq/?category=tm#tm-modify-pbs-nodefile
> 
> Then I took the testing script (rm_script.openmpi) from
> 
> http://svn.oscar.openclustergroup.org/oscar/branches/branch-5-1/packages/openmpi/testing/rm_script.openmpi
> 
> because in OSCAR 5.1 OpenMPI 1.2.4 is used.
> 
> After this everything at the clustertest works fine.
> 
> Thank you for everybody who help me.
> 
> 
> Am Mittwoch, den 23.01.2008, 20:31 +0100 schrieb Dominik Schips:
>> Hi Bernard,
>>
>>
>> Am Mittwoch, den 23.01.2008, 11:09 -0800 schrieb Bernard Li:
>>> Hi Dominik:
>>>
>>> On 1/23/08, Dominik Schips <[EMAIL PROTECTED]> wrote:
>>>
>>>> sles10oscar:/home/oscartst/openmpi # cat openmpitest.err
>>>> [oscarnode2:04413] pls:tm: failed to poll for a spawned proc, return
>>>> status = 17002
>>>> [oscarnode2:04413] [0,0,0] ORTE_ERROR_LOG: In errno in file rmgr_urm.c
>>>> at line 462
>>>> [oscarnode2:04413] mpiexec: spawn failed with errno=-11
>>>> sles10oscar:/home/oscartst/openmpi # cat openmpitest.out
>>>> Running Open MPI test
>>>> Open MPI appears to have TM suppport.  Yippee!
>>>>
>>>> --> MPI C bindings test:
>>>>
>>>> TEST FAILED!
>>>> Commands: cp cpi.c /tmp/openmpi-test && cd /tmp/openmpi-test && mpicc
>>>> cpi.c -o openmpi-cpi && cp openmpi-cpi /home/oscartst/openmpi &&
>>>> cd /home/oscartst/openmpi && mpiexec
>>>> -machinefile /var/spool/pbs/aux//131.sles10oscar -n 2 openmpi-cpi
>>> Have you:
>>>
>>> 1) Tried compiling the code (see above) and see if that works -- if it
>>> compiles fine, then the issue is with launching the MPI job.  Try
>>> launching and see if you get any additional error messages
>> Yes but it was only a try. I am not very familar with MPI and I didn't
>> know how I can compile this correct. As I tried the mpicc (I think) show
>> me that I don't have to compile as root. So I switched to oscartest but
>> didn't get it work. Further tests about this are on my todo for
>> tomorrow. 
>>
>>> 2) Tried googling for the error message you see above?
>> Yes but didn't found something that help me much. I tought here on the
>> mailinglist are some OSCAR people maybe that have seen the same error.
>>
>>> 3) Asked at the Open MPI mailing-list?  If this is a simple setup
>>> issue, they can usually spot the problems much quicker than we do
>> I found some mails on their mailinglist. The discussion was that it
>> could be a bug but someone said it could be a feature ;)
>>
>> A ask on their mailinglist tomorrow. Thank you for your help.
>>

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Oscar-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/oscar-devel

Reply via email to