search

anuga-community / anuga_core

ANUGA for the simulation of the shallow water equation
https://anuga.anu.edu.au
Other
32 stars 23 forks source link

Runtime issue: for finer-grained mesh resolution #14

Closed samcom12 closed 2 years ago

samcom12 commented 2 years ago

Hi, We can run Anuga with coarser-grained mesh resolution well. But, while using finer-grained mesh, we get a below error.

ValueError: send_recv_via_dicts: Number of recv communication buffers > 10

Which is getting triggered from utilities/parallel_abstraction.py

It could create and distribute domains and fails in EVOLVE phase.

stoiver commented 2 years ago

@samcom12 I need a bit more information.

Python version OS Number of processes

This is a strange error as usually send_recv_via_dicts is called to communicate the conserved quantities, stage, xmomentum and ymomentum, ie 3 quantities. So is less than 10.

Are you using domain.update_ghosts() somewhere?

Can you provide more of the error message.

stoiver commented 2 years ago

@samcom12 I just added a little more info on the error message. Could pull the new code from anuga-community.

samcom12 commented 2 years ago

@samcom12 I need a bit more information.

Python version OS Number of processes

This is a strange error as usually send_recv_via_dicts is called to communicate the conserved quantities, stage, xmomentum and ymomentum, ie 3 quantities. So is less than 10.

Are you using domain.update_ghosts() somewhere?

Can you provide more of the error message.

I have used python-3.9.0, GCC-11.2.0 on Linux Centos7 cascade lake with MPI processes 368 and 384.

stoiver commented 2 years ago

@samcom12, on looking at the code it seems that the restriction on the number of buffers is not necessary now we are using mpi4py. So I have remove the restriction, and so your code should now run (after pulling the latest version from anuga-community).

Let me know if it works for your problem.

samcom12 commented 2 years ago

Hi @stoiver, Now it works.

Thanks !!

cheers, Samir