On Thu, 1 Sep 2005, Pierre Valiron wrote:
> Well, I finally found the problem was related to the behaviour of
> MPI_INIT. The code snippet below is buggy when started ever many
> nodes and procs:
I could not reproduce it here with LAM/MPI 7.0.3. I'm able to lamboot
on 80 nodes and then 'mpirun C program' works fine (all on one shell
line such that pauses caused by my typing are avoided). I'm using real
rsh (no queueing system) and most of the nodes are already running
some other jobs, so if the wrong behaviour is caused by some race, I'm
probably not able to see it in these conditions.
--
Bogdan Costescu
IWR - Interdisziplinaeres Zentrum fuer Wissenschaftliches Rechnen
Universitaet Heidelberg, INF 368, D-69120 Heidelberg, GERMANY
Telephone: +49 6221 54 8869, Telefax: +49 6221 54 8868
E-mail: Bogdan.Costescu_at_[hidden]
|
