LAM/MPI logo

LAM/MPI General User's Mailing List Archives

  |   Home   |   Download   |   Documentation   |   FAQ   |   all just in this list

From: Yaakoub Y El-Khamra (yye00_at_[hidden])
Date: 2004-03-24 03:58:11


Mohammad, where are you? which university?
Sorry I have not been keeping up with your problems but I will start
doing that once I clear cactuscode beta 13 release this friday. glad to
see a Mohammad on the mailing list. Thought I was the only one!

Regards
Yaakoub

On Wed, 2004-03-24 at 10:39, ELKANOUNI Mohammed wrote:
> The only way to detect if program is running is checking the running process
> by system command ps.
> You can use this command on your c program by using the c command int
> system(const char *string)
> For script program i have no idea.
> I hope this would be helpful for you
>
> <html><DIV>
> <DIV>&nbsp;</DIV></DIV></html>
>
>
>
>
>
> &gt;From: Edwin Vollebregt &lt;edwin.vollebregt_at_[hidden]&gt;
> &gt;Reply-To: General LAM/MPI mailing list &lt;lam_at_[hidden]&gt;
> &gt;To: lam_at_[hidden]
> &gt;Subject: LAM: Detecting whether MPI is running
> &gt;Date: Wed, 24 Mar 2004 14:18:31 +0100
> &gt;
> &gt;
> &gt;Hello,
> &gt;
> &gt;Does anyone have suggestions for the following situation:
> &gt;
> &gt;We have parallelized a program with PVM in such a way that the same
> &gt;executable can be used for stand-alone sequential runs as well as for
> &gt;master-slave parallel runs. Now we are porting to MPI.
> &gt;
> &gt;The application is started by a shell script with option -npart
> &lt;xx&gt;: number
> &gt;of parts to be used in the decomposition.
> &gt;
> &gt; if [ $npart -le 1 ]; then
> &gt; program.exe
> &gt; else
> &gt; mpiexec -machinefile ... -n 1 master.exe &lt; config
> &gt; # --&gt; master starts the appropriate number of instances of
> program.exe
> &gt; fi
> &gt;
> &gt;The problem is that calling MPI_Init in a sequential run kills the
> program
> &gt;because LAM has not been booted.
> &gt;
> &gt;How can we detect whether MPI is running within the program ?
> &gt;
> &gt;This problem is similar to &quot;detecting mpirun&quot; on the archive
> of the
> &gt;mailinglist. It is a slightly different scenario than the ones I found
> &gt;there, so maybe there are different options.
> &gt;
> &gt;Any thoughts ?
> &gt;
> &gt;Thanks,
> &gt; Edwin
> &gt;--
> &gt;
> _______________________________________________________________________
> &gt;|
> |
> &gt;| VORtech Computing
> |
> &gt;|_______________________________________________________________________|
> &gt;| experts in industrial computing
> |
> &gt;| %&quot;&quot;
> |
> &gt;| %% %%%% dr.ir. E.A.H. Vollebregt
> |
> &gt;| %% &quot; %% phone
> +31(0)15-2850125 |
> &gt;| %% % P.O. Box 260 fax +31(0)15-2850126
> |
> &gt;| %% %% 2600 AG Delft web www.vortech.nl
> |
> &gt;| %% %% the Netherlands edwin.vollebregt_at_[hidden]
> |
> &gt;| %%%%%%
> |
> &gt;| &quot;&quot; visit-address: Torenhove building, M.Nijhoflaan
> 2, Delft |
> &gt;|_______________________________________________________________________|
> &gt;
> &gt;_______________________________________________
> &gt;This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>
> _________________________________________________________________
> MSN Search, le moteur de recherche qui pense comme vous !
> http://search.msn.fr/worldwide.asp
>
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/