Re: [moose-devel] Error in running recurrentIntFire.py

[Subhasis R. <ray...@gm...>]

Dear Misha,

On Sat, May 3, 2014 at 9:15 PM, Misha Ahmadian <mis...@tt...>wrote:

> results, but the problem is the recurrentIntFire.py file. This file in the
> async13/tests/python/mpi could be running, but without any termination and
> real MPI working. we run this part as bellow:
>
> but still we cannot get any result from this model. Would you please let me
> know, if we can have the correct recurrentIntFire.py, to produce our
> results. It would be highly appreciated if you help us as soon as possible.
>

This seems to be system dependent issue and I am also encountering it while
others are not. From my notes I see that I had personally tested the mpi
version with recurrentIntFire.py on 24-01-2014 (that should be when
test/mpi/benchmark.py was also committed). Since something may have changed
in between, can you try checking out a revision from around that date and
see if that works? By the way, the benchmark is done when you see a
"t={some time}" message - that is the time it took to run the simulation.
If it is stuck after that, that is the command prompt of MOOSE expecting
user input - I cannot tell right away why it is acting like that, but we'll
try to address that ASAP. In the mean time also note that benchmark.py is
supposed to benchmark any pymoose script on a multicore machine and you can
imitate it to get a multi-node benchmark.

Best,
  Subhasis