-
Notifications
You must be signed in to change notification settings - Fork 7
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
mpi4py and hdf5 need to be fix for ASE runs #117
Labels
Comments
I thought I post the actual problem below for clarity ...
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
@jrmaddison @dngoldberg
I created a working environment on shin & bow, that runs the inversion for the ASE tomls. The problem is with one of these two libraries.
mpi4py and or hdf5
From the start, so from the
environment.yml
file, we need to fixh5py=3.6.0
andmpi4py=3.1.1
. I can't pin the problem to a specific library and both have open issues now related to openmpi, memory, etc.see: https://github.com/mpi4py/mpi4py/issues
and: https://github.com/HDFGroup/hdf5/issues
We just need to wait and see as for us these dependencies are broken ... for the time being if you want to create an environment on shin @dngoldberg you need to use the following
environment.yml
.You can just copy that to a new yml file and do:
via conda:
conda env create --file new_environment.yml
via mamba:
mamba env create --file new_environment.yml
After that Install fenics_ice and tlm_adjoint manually via pip. See instructions here.
There is no point of changing this on the fenics_ice repository instructions or
install.sh
as everything works except for very large areas thus very large arrays e.g. Amundsen basin. This issue is to do with mpi4py and hdf5 having problems handling those large arrays. An output of the error is detailed in this other issue #116 .The text was updated successfully, but these errors were encountered: