I added it to the path of each node hence when i executed the command from
the node it shows the path /paracel/lam/[webstructure_at_compute-0-0 ~]$ which
mpirun
/paracel/lam/bin/mpirun
Later i found that the path /usr/bin on all compute nodes had a soft link of
mpirun,mpiCC,mpicc etc.. i removed all the soft links and added an
environment file in .ssh/. After doing this the outputs where as follows:
[webstructure_at_hpc-icrisat bin]$ ssh compute-0-0.local which mpirun
/paracel/lam/bin/mpirun
[webstructure_at_hpc-icrisat bin]$ ssh compute-0-2.local which mpirun
/paracel/lam/bin/mpirun
[webstructure_at_hpc-icrisat bin]$ ssh compute-0-3.local which mpirun
/paracel/lam/bin/mpirun
[webstructure_at_hpc-icrisat bin]$ ssh compute-0-4.local which mpirun
/paracel/lam/bin/mpirun
but still i get the same error message when i run my application
the error is
0 - MPI_SEND : Invalid rank 1
[0] Aborting program !
[0] Aborting program!
-----------------------------------------------------------------------------
It seems that [at least] one of the processes that was started with
mpirun did not invoke MPI_INIT before quitting (it is possible that
more than one process did not invoke MPI_INIT -- mpirun was only
notified of the first one, which was on node n0).
mpirun can *only* be used with MPI programs (i.e., programs that
invoke MPI_INIT and MPI_FINALIZE). You can use the "lamexec" program
to run non-MPI programs over the lambooted nodes.
-----------------------------------------------------------------------------
On Thu, Feb 26, 2009 at 8:22 PM, Abhirup Chakraborty
<abhirupc91_at_[hidden]>wrote:
> Try adding the directory (/paracel/lam/bin/) to PATH environment variable.
> Hope it will work.
>
>
> On Thu, Feb 26, 2009 at 4:56 AM, Bhanu Prakash <sneham02i_at_[hidden]>wrote:
>
>> Hi..
>>
>> i'm working with lam 7.1.2
>>
>> lam is installed @ /paracel/lam
>>
>> this is the error message i receive when i run my application.
>>
>> 0 - MPI_SEND : Invalid rank 1
>> [0] Aborting program !
>> [0] Aborting program!
>>
>> -----------------------------------------------------------------------------
>> It seems that [at least] one of the processes that was started with
>> mpirun did not invoke MPI_INIT before quitting (it is possible that
>> more than one process did not invoke MPI_INIT -- mpirun was only
>> notified of the first one, which was on node n0).
>>
>> mpirun can *only* be used with MPI programs (i.e., programs that
>> invoke MPI_INIT and MPI_FINALIZE). You can use the "lamexec" program
>> to run non-MPI programs over the lambooted nodes.
>>
>> -----------------------------------------------------------------------------
>>
>> when i look for which mpirun i get the folllowing messages:
>>
>> on headnode
>> [webstructure_at_hpc~]$ which mpirun
>> /paracel/lam/bin/mpirun
>>
>> from headnode:
>> [webstructure_at_hpc~]$ ssh compute-0-0.local which mpirun
>> /usr/bin/mpirun
>>
>> on compute node:
>> [webstructure_at_hpc~]$ ssh compute-0-0.local
>> Last login: Thu Feb 26 14:31:19 2009 from hpc
>> Rocks Compute Node
>> Rocks 5.0 (V)
>> Profile built 17:45 03-Sep-2008
>>
>> Kickstarted 17:58 03-Sep-2008
>> [webstructure_at_compute-0-0 ~]$ which mpirun
>> /paracel/lam/bin/mpirun
>>
>>
>> i feel this is the reason for the mpirun not working
>>
>> plz do suggest me how can i solve this issue.
>>
>> Thanks in advance.
>>
>> --
>> Don't go where the path may lead, go
>> instead where there is no path and leave
>> a trail.
>>
>> _______________________________________________
>> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>>
>
>
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>
--
Don't go where the path may lead, go
instead where there is no path and leave
a trail.
|
