CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > SU2

Parallel Computation Setup Issues

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   January 23, 2023, 12:13
Default Parallel Computation Setup Issues
  #1
New Member
 
Philip Agenmonmen
Join Date: Oct 2022
Posts: 10
Rep Power: 4
PhilipA is on a distinguished road
Hello all,

I’m currently struggling to get parallel computation to work through the installation guides.

I am on a 2017 MacBook Pro, MacOS Monterey Version 12.6.2, Dual-Core Intel Core i5 processor. I attached an image of the number of processors section of my Hardware Overview .

Doing a simulation with 1 core works fine but running a simulation with more than 1 core leads to the following error:

SU2.EvaluationFailure: Path = /Users/PhilipAgenmonmen/Desktop/Year-4/Dissertation/Gmsh-SU2/NACA-0021/SU2/,
Command = mpirun -n 2 /Users/PhilipAgenmonmen/Desktop/Year-4/Dissertation/Gmsh-SU2/NACA-0021/SU2/bin/SU2_CFD config_CFD.cfg
SU2 process returned error '1'
--------------------------------------------------------------------------
There are not enough slots available in the system to satisfy the 2
slots that were requested by the application:

/Users/PhilipAgenmonmen/Desktop/Year-4/Dissertation/Gmsh-SU2/NACA-0021/SU2/bin/SU2_CFD

Either request fewer slots for your application, or make more slots
available for use.

A "slot" is the Open MPI term for an allocatable unit where we can
launch a process. The number of slots available are defined by the
environment in which Open MPI processes are run:

1. Hostfile, via "slots=N" clauses (N defaults to number of
processor cores if not provided)
2. The --host command line parameter, via a ":N" suffix on the
hostname (N defaults to 1 if not provided)
3. Resource manager (e.g., SLURM, PBS/Torque, LSF, etc.)
4. If none of a hostfile, the --host command line parameter, or an
RM is present, Open MPI defaults to the number of processor cores

In all the above cases, if you want Open MPI to default to the number
of hardware threads instead of the number of processor cores, use the
--use-hwthread-cpus option.

Alternatively, you can use the --oversubscribe option to ignore the
number of available slots when deciding the number of processes to
launch.
--------------------------------------------------------------------------



Some things of note
1. I downloaded MPICH for the MPI version of SU2. However, I believe I now have both Open MPI & MPICH (when I do mpirun --version it says mpirun (Open MPI) 4.1.4). I haven't been able to find the location of Open MPI as I was thinking I may need to delete it.

2. At one point I edited the SU2_MPI_COMMAND in SU2's interface.py file to use 4 cores (oversubscribed or via use-hwthread-cpus).
- However, this led to a recursive call of MPI and the file did not run

Potentially my Mac has 2 cores when hyper-threaded and is already running in parallel?

Any help would be much appreciated,

Thanks,

Philip
Attached Images
File Type: png Screenshot 2023-01-23 at 17.03.02.png (74.3 KB, 10 views)
PhilipA is offline   Reply With Quote

Old   January 23, 2023, 18:00
Default
  #2
Senior Member
 
bigfoot
Join Date: Dec 2011
Location: Netherlands
Posts: 782
Rep Power: 21
bigfootedrockmidget is on a distinguished road
you can check with a command like 'lscpu' how many cpu's you can actually use (or System Report->Hardware ), just make sure the system reports what you expect (I guess 2).
There are some simple example programs to make sure that openmpi (or mpich) works correctly, e.g. here:

https://docs.open-mpi.org/en/v5.0.x/validate.html#

https://mpitutorial.com/tutorials/mpi-hello-world/

For mpich I actually use these compilation options:
Code:
./meson.py build -Dwith-mpi=enabled -Dcustom-mpi=true -Dextra-deps=mpich
and I see messages like this:
Quote:
C compiler for the host machine: /usr/bin/mpicc.mpich (gcc 11.3.0 "gcc (Ubuntu 11.3.0-1ubuntu1~22.04) 11.3.0")
C linker for the host machine: /usr/bin/mpicc.mpich ld.bfd 2.38
C++ compiler for the host machine: /usr/bin/mpicxx.mpich (gcc 11.3.0 "g++ (Ubuntu 11.3.0-1ubuntu1~22.04) 11.3.0")
C++ linker for the host machine: /usr/bin/mpicxx.mpich ld.bfd 2.38

...
...

Run-time dependency mpich found: YES4.0




Can you verify your SU2 output?

bigfootedrockmidget is offline   Reply With Quote

Old   January 24, 2023, 12:16
Default
  #3
New Member
 
Philip Agenmonmen
Join Date: Oct 2022
Posts: 10
Rep Power: 4
PhilipA is on a distinguished road
Hey. Thanks for the information.

Here is my core & thread count:
machdep.cpu.core_count: 2
machdep.cpu.thread_count: 4

I'm currently working on running test files with mpi and will update you on how that goes.

One thing I wanted to mention is that I am using a pre-compiled SU2 MPI version and I don't have that meson.py file.

Just to check, should I look to build a new SU2 from source?
PhilipA is offline   Reply With Quote

Old   January 24, 2023, 15:47
Default
  #4
Senior Member
 
bigfoot
Join Date: Dec 2011
Location: Netherlands
Posts: 782
Rep Power: 21
bigfootedrockmidget is on a distinguished road
ok, I missed that you were using the precompiled SU2, my guess is that SU2 sees openmpi instead of mpich. It would be best if you can safely remove openmpi and install mpich only.
bigfootedrockmidget is offline   Reply With Quote

Old   January 26, 2023, 13:30
Default
  #5
New Member
 
Philip Agenmonmen
Join Date: Oct 2022
Posts: 10
Rep Power: 4
PhilipA is on a distinguished road
That seems to have fixed things! Thank you very much. I can now run 2 cores!

I'll need to double-check but I don't think I can run 4 cores. This may be since the extra 2 are not physical.

Uninstall
I used point no. 3 in this post to uninstall openmpi for anyone else’s info.
https://github.com/open-mpi/ompi/issues/600
PhilipA is offline   Reply With Quote

Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Scaling of parallel computation? Solver/thread count combinations? tdof OpenFOAM 8 January 13, 2022 11:39
Any Procedure to setup parallel computation?? fluidmechanics SU2 Installation 10 July 21, 2017 04:31
Fluent parallel computation using Xeon duo-processor with CFD occupation below 50% zhujj09 FLUENT 1 October 19, 2016 15:02
FE-3.1 Parallel issues with BCs (gMax, reduce...) Phicau OpenFOAM Bugs 2 August 1, 2014 06:22
SU2 Tutorial 2 Parallel Computation CrashLaker SU2 7 April 5, 2014 16:14


All times are GMT -4. The time now is 23:26.