Salome_Meca 2024 on WSL2 - C_A missing

mf

k_zurawski
It's would be the same.

Mario.

k_zurawski

mf Thank you for your help.

hberro

k_zurawski

Try binding /local00 to any local directory on your pc.

mkdir /tmp/local00
./salome-meca... -B /tmp/local00:/local00

Let me know if that helps

k_zurawski

hberro
Hello,
Singularity> ./salome_meca-lgpl-2024.1.0-1-20240327-scibian-11.sif -B /tmp/local00:/local00
bash: ./salome_meca-lgpl-2024.1.0-1-20240327-scibian-11.sif: Permission non accordée

w2@W2:~$ ./salome_meca-lgpl-2024.1.0-1-20240327-scibian-11.sif -B /tmp/local00:/local00
-bash: ./salome_meca-lgpl-2024.1.0-1-20240327-scibian-11.sif: Permission denied

what else should I do?

k_zurawski

Yesterday I installed Ubuntu 22.04 under Hyper-V, updated OS and installed Apptainer 1.3.4. The result is the same as in Debian under WSL. I had a gut feeling that there's a communication error between preprocessor and solver. I haven't seen it in WSL, but below screenshot might indicate that. Could anyone comment on this?

When I try to run as_run in shell I'm getting result as below. I'm really illiterate when it comes to dealing with Linux and these kinds of problems.

nicolas_p

You have a problem with as_run since you have no access to /local00/tmp

Try to see if it exists in your system and copy a file.

k_zurawski

nicolas_p
where should it be created? In / or in /home/u22/ ?

nicolas_p

yes directly in /.

k_zurawski

what can I do about it?

nicolas_p

here, this is not a problem with smeca but basic linux problem.

k_zurawski

nicolas_p
ok, so how to deal with it? I'd really appreciate if you could guide me through the whole process. I only know basic Linux commands and I'm far from being "proficient" with it. Please don't look at it as sort of ignorance. I really don't have much time. I'd rather spend it on Code_Aster, not on Linux itself.

jeanpierreaubry

nicolas_p

here, this is not a problem with smeca but basic linux problem.

is this good enough a reason not to help ?

StanislavPavlica

Hello Krzysztof,

try following steps:

I believe you should use singularity not apptainer to run salome_meca-lgpl-2024.1.0-1-20240327-scibian-11.sif.
I'm not using Windows so I'm not sure of this:

To quickly install the 4.0.0 package singularity use the following commands inside the WSL2 Ubuntu 22.04 window:

wget https://github.com/sylabs/singularity/releases/download/v4.0.0/singularity-ce_4.0.0-jammy_amd64.deb
sudo apt install ./singularity-ce_4.0.0-jammy_amd64.deb

File salome_meca-lgpl-2024.1.0-1-20240327-scibian-11.sif put in your home directory:
/home/u22/
Or you can make a link in your home directory.
Then run inside the WSL2 Ubuntu 22.04 window this command to install salome_meca (in /home/u22/ directory):
singularity run --app install salome_meca-lgpl-2024.1.0-1-20240327-scibian-11.sif
This command generate in your home directory a script with same name as a container but without extension .sif:
salome_meca-lgpl-2024.1.0-1-20240327-scibian-11
Then run this script to start salome_meca:
./salome_meca-lgpl-2024.1.0-1-20240327-scibian-11
Script prepare environment to properly run container including access to /local00/tmp

If you start container with command
singularity run salome_meca-lgpl-2024.1.0-1-20240327-scibian-11.sif
this environment is missing and salome _meca is not running properly.

Stano

k_zurawski

StanislavPavlica
I have Apptainer in virtual machine and Singularity 4.2.0 in Debian 12 on WSL. Both don't work. The last one screenshot is from Singularity and Debian.

k_zurawski

StanislavPavlica
I followed your step-by-step guide and AsterStudy sees Code_Aster on my VM. as_run also starts, but I haven't run any simulations yet.
BIG THANKS!!!

I need to replicate the same on WSL and Debian and see if this also solves the issue. If not, I'll try WSL with Ubuntu 22.04 as in case above. Pity I can't have 2 separate workstations, one with Windows and one with Ubuntu.

k_zurawski

Well, I can't run the simulation because the same thing happens as in mf case. When I click on 3 dots to IMPR_RESU Salome_Meca crashes.
I checked WSL+Debian and SM doesn't start. I'm getting additional errors. I'll try later with Ubuntu 22.04 on WSL. I don't want to use VM, VM is only a temporary solution to check if is actually works.

k_zurawski

On Debian 12 on WSL and Singularity 4.2.0 only as_run works. Which is half of the success.

Again, thank you both. I really appreciate your help.

hberro

I got this issue when putting together a ready to use WSL2 distrib based on the latest 2024 sif file.

I only had to create a /local00 folder as root and then chown it by the regular user.

In the tutorial below, everything should work out of the box :

https://forum.code-aster.org/public/d/28014-salome-meca-20241-on-wsl2-ready-to-use-distribution-call-for-beta-testers

jeanpierreaubry

hberro

thanks this helpful hint
once this is done here is the .mess file of a known problem

Traceback (most recent call last):
File "./avera.comm.changed.py", line 10, in <module>
from code_aster.Commands import *
File "/home/dedalus/sandbox2024/opt/salome_meca/V2024.1.0_scibian_univ/tools/Code_aster_stable-1670/lib/aster/code_aster/Commands/init.py", line 33, in <module>
from ..Supervis import CO
File "/home/dedalus/sandbox2024/opt/salome_meca/V2024.1.0_scibian_univ/tools/Code_aster_stable-1670/lib/aster/code_aster/Supervis/init.py", line 26, in <module>
from .CommandSyntax import CommandSyntax
File "/home/dedalus/sandbox2024/opt/salome_meca/V2024.1.0_scibian_univ/tools/Code_aster_stable-1670/lib/aster/code_aster/Supervis/CommandSyntax.py", line 62, in <module>
from ..Objects import DataStructure
File "/home/dedalus/sandbox2024/opt/salome_meca/V2024.1.0_scibian_univ/tools/Code_aster_stable-1670/lib/aster/code_aster/Objects/init.py", line 26, in <module>
from libaster import *
ImportError: libmed.so.11: cannot open shared object file: No such file or directory

EXECUTION_CODE_ASTER_EXIT_15953=1

--- DIAGNOSTIC JOB : <F>_ABNORMAL_ABORT

line 10 is LIRE_MAILLAGE as could be expected

hberro

jeanpierreaubry

Jean Pierre, I think that you need to use the salome meca launcher with " -- shell -- " before running as_run so that the dynamic libs are loaded properly!

StanislavPavlica

Hello Krzysztof,

please include always full output from shell. I was able to help you because in first post you have
u22@W2:~$ apptainer run salome_meca-lgpl-2024.1.0-1-20240327-scibian-11.sif...

How are you starting as_run?

To open a shell inside container with salome_meca environment use command:
./salome_meca-lgpl-2024.1.0-1-20240327-scibian-11 -- shell
Then you can start as_run.

watch out, there is a space between -- and shell. Without space, there is no salome_meca environment.

For more options look at output from command:
./salome_meca-lgpl-2024.1.0-1-20240327-scibian-11 -- help

and (this is different commmand):
./salome_meca-lgpl-2024.1.0-1-20240327-scibian-11 --help

To start as_run in one command for file ~/example/example.export you can use:

~/salome_meca-lgpl-2024.1.0-1-20240327-scibian-11 -- shell -- as_run ~/example/example.export

(~ is same like /home/u22)

Stano

k_zurawski

StanislavPavlica
I apologise for that, I'll keep that in mind. I know that sometimes my posts are messed up. I'm trying to do few things at the same time. Usually with poor effect.

I'm starting as_run by
w2@W2:~$ ./salome_meca-lgpl-2024.1.0-1-20240327-scibian-11 shell
and it works fine, as I posted above. I run only one small study, but this confirms, that it works (I think). I don't know if it's important, but the result is the same without "--" before "shell"