RE: Vegetation with MPI - Forum - XBeach
bug message reminder
RE: Vegetation with MPI
I wonder if the current xbeach version can handle properly vegetation simulation with multiple processors.
I am testing one of example for vegetation (Modeling wave attenuation by vegetation (2D)) which was downloaded from the Xbeach site (https://github.com/openearth/xbeach-docs/tree/master/docs/tutorials/vegetation_field2d/files/Model).
When I tested this example with single processor, the simulation was completed without any problems.
However, the model showed some problems with multiple processors.
Looking at the XBlog.txt, writing log stopped while vegetation parameters were read (See attached file)
I suspected the vegetation part in the model.
So, I deactivated the vegetation process and resubmitted the simulation with multiple processes. In this case, the model finished successfully.
I wonder if you have experienced the same problems.
I am using xbeach_5153_win32_mpi_netcdf.
We recently found a bug in the vegetation module (only with MPI) and are currently trying to fix this.
Sorry for the delay. Hopefully we can get you a new fixed version soon!
Thanks for your response.
Please let us know if you fix that.
Just a brief update on this: Unfortunately this bug seems to be a bit more complicated than we thought. Due to this, in combination with some busy agendas, I expect it may take some more time for this to get resolved.
We are now aiming to have it resolved before the next release (probably towards the end of this year), but until that time I'd recommend to use the vegetation module without MPI for now.
I'm trying to run xbeach in parallel but I am also getting the problems described above. Is there any update on the bug in the vegetation module?
This issue should be fixed, could you please specify what version of XBeach you are using? If you are using the latest trunk/release, could you please upload your model input files so I can have a look?