open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi All
I have just installed and compiled the v8p0r0 version and am running some examples to start.
I'm running the m2_wave examples in telemac2d which is successful in scalar mode but when I specify > zero processor (parallel mode) I get an argument error in run_cas.py.
Can anyone help with solving this please.


... checking the executable
Traceback (most recent call last):
File "/gpfs/home/e088/opentelemac_80/v8p0r0/scripts/python3/runcode.py", line 288, in <module>
main(None)
File "/gpfs/home/e088/opentelemac_80/v8p0r0/scripts/python3/runcode.py", line 271, in main
run_study(cas_file, code_name, options)
File "/gpfs/home/e088/opentelemac_80/v8p0r0/scripts/python3/execution/run_cas.py", line 142, in run_study
run_hpc_cas(my_study, options)
File "/gpfs/home/e088/opentelemac_80/v8p0r0/scripts/python3/execution/run_cas.py", line 115, in run_hpc_cas
my_study.compile_exe(options.bypass)
TypeError: compile_exe() takes 1 positional argument but 2 were given

Could you post the command you used ?

I used:
../v8p0r0/scripts/python3/runcode.py -c submpi telemac2d -s t2d_m2wave.cas

where submpi is the configuration 3 in HPC clusters and changed the cas to PARALLEL PROCESSORS=1

Hi,

Can you try by adding --ncsize=1 to your command.
That should solve the issue.
I think in my modifications i am not reading the number of parallel processor from the steering file anymore.
The python3 part is still evolving.

I've tried with:

../v8p0r0/scripts/python3/runcode.py -c submpi --ncsize=1 telemac2d -s t2d_m2wave.cas and I get the same error.

I just tried and I do not have the error.
What version of python3 are you using ?
Could you post you configuration file as well ?

Yugi

I'm using python 3.5.5 which is loaded with python anaconda 4.2.
I also have an earlier version of anaconda available (2.4.0) with python 3.5.

Attached is the config file.

Thanks
John

You can try adding "options: hpc" in the submpi part.

Hi Yugi
I am still having difficulty getting the parallel runs to work.
but I still have problems getting the full compilation of the system - the APi compilation fails. I have three configs, one for scaler interactive jobs, and two for parallel through the queue - one of them without the merge (as per supplied configurations 1,3 and 4)

The compilation errors through python traceback are attached in file -
for 1: an KeyError in 'libs_all'

for 2 and 3: Missing keyword pyd_fcompiler in configuration file.

Could you possibly suggest the fixes for these please?

John

Fr the first one: had libs_all: (under general in your .cfg file)

As for 2 and 3 you are missing keywords mandatory for compiling the api in python3.
Below is what i use for our configuration: