SimpleFoam run in Parallel issue!

tareqkh · October 13, 2016, 11:02

Hello Foamers,

I have been running a special case using simpleFoam and came to a question that some of you might have experienced. I am trying to monitor some results including the residual using simpleFoam > log.simpleFoam command, but when I switch to mpirun -np 12 simpleFoam -parallel > log.simpleFoam, some of the output that the simpleFoam provides become 0 instead of the real values. Has anyone had this issue before could tell us the reason of not getting the results in parallel?

Best,

Saideep · October 17, 2016, 13:41

Hi,

I generally use "mpiexec -n 12 solverName -parallel > log&".

I am not sure why do you need log.simpleFoam. But I never had a problem with my statement either run on parallel or on a single core(by removing the mpi related part).

Saideep

tareqkh · October 17, 2016, 14:45

Quote:

Originally Posted by Saideep

Hi,

I generally use "mpiexec -n 12 solverName -parallel > log&".

I am not sure why do you need log.simpleFoam. But I never had a problem with my statement either run on parallel or on a single core(by removing the mpi related part).

Saideep

Hello Saideep,

Thank you for the answer. However, I am using a modified simpleFoam i.e. something relevant to my project. So now, the modified simpleFoam doesn't provide an output when I run in parallel. Any idea, where to look in the source code to fix the issue about not getting the results in parallel.

Best,

tareqkh · October 19, 2016, 11:00

Hello Foamers,

I have been running a special case using a modified simpleFoam and came to a question that some of you might have experienced. I am trying to monitor some results including the residual using simpleFoam > log.simpleFoam command, but when I switch to mpirun -np 12 simpleFoam -parallel > log.simpleFoam, some of the output that the simpleFoam provides become 0 instead of the real values. Has anyone had this issue before could tell us the reason of not getting the results in parallel? Any idea, where to look in the source code to fix the issue about not getting the results in parallel.

Best,

Saideep · October 21, 2016, 09:28

Hi,

It is a bit strange. Few things here. From what I understand is that,
1. your modified simpleFOAM is working when not using mpi. Can you check your results. It shall be spurious. I guess the problem comes from the modification part.

2. if results behave as you expect then the culprit is mpi, but I never encountered a problem with mpi on modifications.

Saideep

Traction · October 21, 2016, 13:05

Hey,

just to be sure: Do you use the specific mpi reduce operations for generating the output in your modification ?

wyldckat · January 1, 2018, 18:07

Quick answer: I'm very late to this discussion, but from the description, it seems that this was a bug. But only by knowing the OpenFOAM version and the function objects used, would it be possible to at least know if this is a known bug or not.

October 13, 2016, 11:02	SimpleFoam run in Parallel issue!	#1
tareqkh Senior Member CFD Join Date: Nov 2010 Location: United States Posts: 243 Rep Power: 16	Hello Foamers, I have been running a special case using simpleFoam and came to a question that some of you might have experienced. I am trying to monitor some results including the residual using simpleFoam > log.simpleFoam command, but when I switch to mpirun -np 12 simpleFoam -parallel > log.simpleFoam, some of the output that the simpleFoam provides become 0 instead of the real values. Has anyone had this issue before could tell us the reason of not getting the results in parallel? Best,

October 19, 2016, 11:00	SimpleFoam in Parallel issue!	#4
tareqkh Senior Member CFD Join Date: Nov 2010 Location: United States Posts: 243 Rep Power: 16	Hello Foamers, I have been running a special case using a modified simpleFoam and came to a question that some of you might have experienced. I am trying to monitor some results including the residual using simpleFoam > log.simpleFoam command, but when I switch to mpirun -np 12 simpleFoam -parallel > log.simpleFoam, some of the output that the simpleFoam provides become 0 instead of the real values. Has anyone had this issue before could tell us the reason of not getting the results in parallel? Any idea, where to look in the source code to fix the issue about not getting the results in parallel. Best,

January 1, 2018, 18:07		#7
wyldckat Retired Super Moderator Bruno Santos Join Date: Mar 2009 Location: Lisbon, Portugal Posts: 10,975 Blog Entries: 45 Rep Power: 128	Quick answer: I'm very late to this discussion, but from the description, it seems that this was a bug. But only by knowing the OpenFOAM version and the function objects used, would it be possible to at least know if this is a known bug or not. __________________ OpenFOAM: FAQ \| Getting started Forum: How to get help, to post code/output and forum guide Read this before sending me PM

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
problem w/ running simpleFoam in parallel	mazdak	OpenFOAM Running, Solving & CFD	27	December 8, 2023 03:10
Foam::error::PrintStack	almir	OpenFOAM Running, Solving & CFD	91	December 21, 2022 04:50
[Salome] Salome import "Cannot find file "points" in directory..."	mismichael	OpenFOAM Meshing & Mesh Conversion	5	June 22, 2019 02:56
[mesh manipulation] Importing Multiple Meshes	thomasnwalshiii	OpenFOAM Meshing & Mesh Conversion	18	December 19, 2015 18:57
Trying to run a benchmark case with simpleFoam	spsb	OpenFOAM	3	February 24, 2012 09:07

October 17, 2016, 13:41		#2
Saideep Senior Member Saideep Join Date: Apr 2015 Location: INDIA Posts: 203 Rep Power: 12	Hi, I generally use "mpiexec -n 12 solverName -parallel > log&". I am not sure why do you need log.simpleFoam. But I never had a problem with my statement either run on parallel or on a single core(by removing the mpi related part). Saideep

October 21, 2016, 09:28		#5
Saideep Senior Member Saideep Join Date: Apr 2015 Location: INDIA Posts: 203 Rep Power: 12	Hi, It is a bit strange. Few things here. From what I understand is that, 1. your modified simpleFOAM is working when not using mpi. Can you check your results. It shall be spurious. I guess the problem comes from the modification part. 2. if results behave as you expect then the culprit is mpi, but I never encountered a problem with mpi on modifications. Saideep

October 21, 2016, 13:05		#6
Traction Member Join Date: Mar 2014 Posts: 39 Rep Power: 12	Hey, just to be sure: Do you use the specific mpi reduce operations for generating the output in your modification ?