Full memory Problem in Liggghts.

Submitted by terces907 on Tue, 11/06/2012 - 07:16

Hi,
I have a problem with memory. My memory is about 16 GB but not enough for my simulation. It can not finish because It stuck an error seems like memory is full. How to figure out this problem?
P.S. When i implement code with cohesion it will use so much memory. If don't use cohesion it quite OK can run till finished

##Technical Info:##
error messages:

ERROR on proc 0: Failed to allocate 783451440 bytes for array neigh:binhead (memory.cpp:45)
--------------------------------------------------------------------------
MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD
with errorcode 1.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------
--------------------------------------------------------------------------
mpirun has exited due to process rank 0 with PID 30823 on
node speacer exiting without calling "finalize". This may
have caused other processes in the application to be
terminated by signals sent by mpirun (as reported here).
--------------------------------------------------------------------------

CPU: Intel® Core™ i7-3770 CPU @ 3.40GHz × 8
Memory: 15.8 GiB
Compilation: openmpi (parrallel processing with option -np 8)

Forums:

Philippe's picture

Philippe | Tue, 11/06/2012 - 08:55

Hello terces,

I don't know the memory demand of the cohesion model, but the error comes from binning during neighbor list build, so if you increase the binsize, there will be less memory needed for neigh:binhead and you might get away with it.

Have you monitored the memory consumption of your case? If it is very close to filling your RAM then you maybe should consider decreasing the amount of particles.

best
Philippe

moritzhoefert | Fri, 11/09/2012 - 15:47

Hi,

I sometimes get memory problems if I don't use
boundary f f f
(or p alternatively) which fixes the box size.

If your box size can vary you can run into problems if you have got for example a leaky stl wall. Then particles might travel infinitely far away in the direction of gravity.

Best,
Moritz