Tutorial 2 - Parallel run

nicholas_yue · March 20, 2014, 21:47

Hi,

I have successfully run tutorial 2 (using the test case data I downloaded)

http://adl-public.stanford.edu/docs/...iscid+ONERA+M6

in serial mode and looked at the results in Paraview.

When I continue the same tutorial to learn about the parallel execution, I found that the Python script parallel_computation.py makes assumption that the mesh is named "mesh.su2" rather than the mesh name specified in the configuration file.

To make it easier to continue, I elected to modify the name of the mesh file and the inv_ONERA6.cfg file instead of modifying the python script.

The next problem I encountered is more complicated. The generated file config_DDC.cfg did not contain the LOWER_SIDE definition expected by SU2_DDC

Is there any other tutorial/example (in 3D) which I can look at that is known to work via a parallel approach ?

Cheers

CrashLaker · March 24, 2014, 20:33

Imma give it a try and post here what I got.

hlk · March 25, 2014, 12:53

If the config file is being called correctly, the mesh used should be the one specified in the config file and not the default, so I suspect that there may have been a typo when typing the config file name. (this is an error that I have made myself multiple times) That would also explain why SU2_DDC is having trouble - if the main config file is not being read in, the default settings loaded by SU2_DDC will not match the mesh file.

To run the case in parallel on two processors, after making sure that you have configured SU2 to run in parallel, type the following command in your run directory:

>> parallel_computation.py -f inv_ONERAM6.cfg -p 2

March 20, 2014, 21:47	Tutorial 2 - Parallel run	#1
nicholas_yue New Member Nicholas Yue Join Date: Mar 2012 Posts: 5 Rep Power: 14	Hi, I have successfully run tutorial 2 (using the test case data I downloaded) http://adl-public.stanford.edu/docs/...iscid+ONERA+M6 in serial mode and looked at the results in Paraview. When I continue the same tutorial to learn about the parallel execution, I found that the Python script parallel_computation.py makes assumption that the mesh is named "mesh.su2" rather than the mesh name specified in the configuration file. To make it easier to continue, I elected to modify the name of the mesh file and the inv_ONERA6.cfg file instead of modifying the python script. The next problem I encountered is more complicated. The generated file config_DDC.cfg did not contain the LOWER_SIDE definition expected by SU2_DDC Is there any other tutorial/example (in 3D) which I can look at that is known to work via a parallel approach ? Cheers

March 24, 2014, 20:33		#2
CrashLaker Member Carlos Alexandre Tomigawa Aguni Join Date: Mar 2014 Posts: 40 Rep Power: 12	Imma give it a try and post here what I got. Last edited by CrashLaker; March 26, 2014 at 10:27.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Dam Break Tutorial - 2.3.11 Running in parallel	vitors	OpenFOAM Running, Solving & CFD	3	August 2, 2019 19:22
Parallel run: boundary condition missing on processorX/0	lovecraft22	OpenFOAM Running, Solving & CFD	6	May 29, 2012 16:06
problem in parallel run in Workbench	Mina_Shahi	CFX	0	September 8, 2011 11:27
Parallel run in fluent for multiphase flow	apurv	FLUENT	2	August 3, 2011 19:44
First Parallel Run - need some help	Gian Maria	OpenFOAM	3	June 17, 2011 12:08

March 25, 2014, 12:53		#3
hlk Senior Member Heather Kline Join Date: Jun 2013 Posts: 309 Rep Power: 13	If the config file is being called correctly, the mesh used should be the one specified in the config file and not the default, so I suspect that there may have been a typo when typing the config file name. (this is an error that I have made myself multiple times) That would also explain why SU2_DDC is having trouble - if the main config file is not being read in, the default settings loaded by SU2_DDC will not match the mesh file. To run the case in parallel on two processors, after making sure that you have configured SU2 to run in parallel, type the following command in your run directory: >> parallel_computation.py -f inv_ONERAM6.cfg -p 2