[MPICH] mpich2 program hangs

[Watford, Christopher A (GE Infra, Energy)]

When I have had MPI apps hang, it typically was network card related (my
instance: TCP Offloading).


--
Christopher
9106755743

-----Original Message-----
From: owner-mpich-discuss at mcs.anl.gov
[mailto:owner-mpich-discuss at mcs.anl.gov] On Behalf Of Rajeev Thakur
Sent: Thursday, August 02, 2007 8:55 AM
To: 'si ceng'; mpich-discuss at mcs.anl.gov
Subject: RE: [MPICH] mpich2 program hangs

This doesn't usually happen. It could even be something wrong with the
network card/driver on one of the machines. Can you run it on a single
machine? Can you run it on another pair of machines?

Rajeev

> -----Original Message-----
> From: owner-mpich-discuss at mcs.anl.gov 
> [mailto:owner-mpich-discuss at mcs.anl.gov] On Behalf Of si ceng
> Sent: Thursday, August 02, 2007 4:23 AM
> To: mpich-discuss at mcs.anl.gov
> Subject: [MPICH] mpich2 program hangs
> 
> Dear all:
> I installed mpich2-1.0.5p4 on two machines and tried the example cpi.
> the two machines are machine1 and machine2.
> I can run ssh/rsh without password on both machines.
> 
> when I only use one machine for computation, it returned good result.
> 
> when I use two machines,
> 
> my mpd.hosts is:
> machine1
> machine2
> 
> I ran mpdboot -n 2 -f mpd.hosts
> 
> I ran mpdtrace, the output is :
> machine1
> machine2
> 
> when I ran mpiexec -n 5 examples/cpi:
> the output is
> Process 0 of 5 is on machine1
> Process 2 of 5 is on machine2
> Process 4 of 5 is on machine1
> Process 3 of 5 is on machine2
> Process 1 of 5 is on machine2
> 
> then the program stop there and no output then.
> Can anyone help?
> 
> thanks
> 
> _________________________________________________________________
> FREE pop-up blocking with the new MSN Toolbar - get it now! 
> http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/
> 
>