SU2 3.2 parrallel bug in shape_optimization.py?

RLangtry · July 2, 2014, 22:27

Was wondering if there is a bug in the new shape_optimization.py script in SU2 3.2. Trying to run shape_optimization.py on the standard rans_optimization test cases and it runs in serial but fails in parallel on the first design run with the following partition error:

------------------------- Exit Success (SU2_PRT) ------------------------

cstr=mesh_NACA0012_inv.su2
There is no geometry file (GetnZone))!
cstr=mesh_NACA0012_inv.su2
There is no geometry file (GetnZone))!

parallel_computation.py works for ver. 3.2 so I don't believe it is my cluster/mpi setup. Also I have no problems with shape_optimization.py in SU2 3.1.

Anyone else encountered this?

shirazbj · July 2, 2014, 23:55

Quote:

Originally Posted by RLangtry

Was wondering if there is a bug in the new shape_optimization.py script in SU2 3.2. Trying to run shape_optimization.py on the standard rans_optimization test cases and it runs in serial but fails in parallel on the first design run with the following partition error:

------------------------- Exit Success (SU2_PRT) ------------------------

cstr=mesh_NACA0012_inv.su2
There is no geometry file (GetnZone))!
cstr=mesh_NACA0012_inv.su2
There is no geometry file (GetnZone))!

parallel_computation.py works for ver. 3.2 so I don't believe it is my cluster/mpi setup. Also I have no problems with shape_optimization.py in SU2 3.1.

Anyone else encountered this?

Did you get n mesh files after su2_prt? How many zones you specified?

RLangtry · July 3, 2014, 01:20

yes I ran:

shape_optimization.py -f inv_NACA0012_adv.cfg -n 16

and ended up with 16 mesh files in the /DSN_001/DIRECT/ directory. I can launch the solver manually with mpirun -n 16 /home/SU2/bin/SU2_CFD config_CFD.cfg which makes me think there is a bug in the shape_optimization.py script.

Could it be a race condition? Perhaps the partitions have not all been written before it starts the mpirun?

July 2, 2014, 22:27	SU2 3.2 parrallel bug in shape_optimization.py?	#1
RLangtry New Member Join Date: May 2014 Posts: 4 Rep Power: 12	Was wondering if there is a bug in the new shape_optimization.py script in SU2 3.2. Trying to run shape_optimization.py on the standard rans_optimization test cases and it runs in serial but fails in parallel on the first design run with the following partition error: ------------------------- Exit Success (SU2_PRT) ------------------------ cstr=mesh_NACA0012_inv.su2 There is no geometry file (GetnZone))! cstr=mesh_NACA0012_inv.su2 There is no geometry file (GetnZone))! parallel_computation.py works for ver. 3.2 so I don't believe it is my cluster/mpi setup. Also I have no problems with shape_optimization.py in SU2 3.1. Anyone else encountered this?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Official Release of SU2 V3.2	economon	SU2 News & Announcements	6	April 18, 2015 00:28
Pointwise-SU2 joint webinar (April 29th) and SU2 v3.1.0 new release	fpalacios	SU2 News & Announcements	1	April 30, 2014 03:40
Official release of SU2 V3.0 and SU2 Educational V1.0	fpalacios	SU2 News & Announcements	2	January 22, 2014 06:28
Welcome to the Stanford University Unstructured (SU2) forum!	economon	SU2	0	January 7, 2013 03:48
New SU2 Forum Opened at CFD Online	pete	Site News & Announcements	0	January 5, 2013 18:59

July 3, 2014, 01:20		#3
RLangtry New Member Join Date: May 2014 Posts: 4 Rep Power: 12	yes I ran: shape_optimization.py -f inv_NACA0012_adv.cfg -n 16 and ended up with 16 mesh files in the /DSN_001/DIRECT/ directory. I can launch the solver manually with mpirun -n 16 /home/SU2/bin/SU2_CFD config_CFD.cfg which makes me think there is a bug in the shape_optimization.py script. Could it be a race condition? Perhaps the partitions have not all been written before it starts the mpirun?