code_saturne User's Forum

Hello, No, we currently only use MPI-IO for EnSight format output and checkpoint/restart files. MED and CGNS still use serialized I/O, so serial HDF5 is fine. The code should look in /usr by itself, but BLAS might be only linked if forced. BLAS are only used for some unit performance comparison test...

Hello, Due to the multiple builds of HDF5 (with and without MPI), the files may not be directly in /usr/include and /usr/lib; but under a subdirectory. You might need to check the file paths by listing files associated with the packages under Synaptic to adapt them to Ubuntu, but on some of our Debi...

Hello,

At least for HDF5 (required by MED), the med.h file is not installed. Did you install the associated "-dev" package series as described in the installation manual ?

Regards,

Yvan

Hello,

Could you post the details as listed in this Forum section's recommendations ?

Also, avoid running "configure" in the source directory (using a separate build directory is strongly recommended).

Regards,

Yvan

Hello, With X86_64/ Infiniband clusters, we usually have reasonably good scaling at least down to about 30000 cells per MPI rank for small meshes, and 50000 cells per rank for very large meshes on hundreds of nodes, so a mesh of almost one million cells should provide good scaling at least up to 20 ...

Hello, I have not re-tested recently, but this specific isue was reported to the Debian package maintainer a few months ago (in the fall), and fixed (at least in Debian testing) within a few days. It seems that testing has moved to 4.2 recently (which is a good thing, though we should hurry to relea...

Hello,

My suggestion with a prescribed velocity assumed you had some idea of a expected, or realistic velocity from which you could start, then iterate.

Otherwise, I'll remind Erwan to check these posts. He knows the different compressible BC types much better than I do...

Regards,

Yvan

Hello, Yes, this is probably due to OAR not being supported by Code_Saturne . For supported batch systems, the environment variables related to known batch systems are handled in bin/cs_exec_environment.py (which should be relatively easy to extend to a different batch system once you have its docum...

Hello, The next to last line translates to: "Number of reactions intervening for global species/stoichiometry in number of moles global species" (sorry, this is not a correct sentence, but the one in French is not clearer)... I'll ask a colleague who works in fire modelling if he can provi...

Hello, The pressure solver seems to diverge, so your boundary conditions or initialzation are probably too "steep". Did you run the same mesh for an incompressible computation first ? This might help adjust the time step and other boundary conditions ? I'll remind Erwan, who knows the comp...