Environment¶

This page describes the environments in which Fab can be run.

Requirements¶

The minimum dependencies for Fab to build a project are:

Python >= 3.7
Compiler
Linker (or archiver, if building libraries)

If your project includes C code:

libclang
python-clang

Running Fab at The Met Office¶

VDI¶

Without clang¶

If you don’t need to build C, you can use this modules command:

$ module use $EXTRA_SOFTWARE
$ module load fab/0.11.0/python/3.7.0

With clang¶

This is currently a little awkward but will hopefully become simpler soon. We use a combination of module commands and conda to get Fab running at The Met Office. If you don’t need to compile C code, you may not need conda. We create the conda environment as follows:

$ conda env create -f envs/conda/dev_env.yml
$ conda activate sci-fab

Then install fab.

We do this in a before any module commands. Then we set up our environment in a new terminal as follows.

For use with gfortran:

$ module use $LFRIC_MODULES
$ module load environment/lfric/gnu
$ conda activate sci-fab
$ PYTHONPATH=~/.conda/envs/sci-fab/lib/python3.7/site-packages:$PYTHONPATH

For use with ifort:

$ module use $LFRIC_MODULES
$ module load environment/lfric/ifort
$ conda activate sci-fab
$ PYTHONPATH=~/.conda/envs/sci-fab/lib/python3.7/site-packages:$PYTHONPATH

The PYTHONPATH line gives us access to a newer version of fparser in the conda environment. Otherwise we get the older one from the modules commands.

PyCharm¶

If you run pycharm-community from the command line, after the above commands, PyCharm will be able to run Fab, the tests, etc. You can set the project interpreter to be that in the conda environment.

Rose¶

Various configs for building projects using Rose on SPICE can be found in run_configs.

Singularity¶

You can run Fab in a singularity container as follows:

singularity run oras://metoffice-docker-local.jfrog.io/picasso/metomi/fab/MyImage:latest

If you need to use git from within the container, you’ll need to set a couple of environment variables first:

$ export SINGULARITY_BIND="/etc/pki/ca-trust/extracted/pem:/pem"
$ export SINGULARITYENV_GIT_SSL_CAPATH="/pem"
$ singularity shell oras://metoffice-docker-local.jfrog.io/picasso/metomi/fab/MyImage:latest

Authentication¶

You’ll need to authenticate if it’s your first time:

$ singularity remote login -u firstname.surname@metoffice.gov.uk docker://metoffice-docker-local.jfrog.io

You will be asked for your access token.

Outside The Met Office¶

Docker¶

The dockerfile in envs/docker can be used to create a container in which to run Fab. This work-in-progress solution was tested on Windows, running Ubuntu in WSL.

Build the image:

$ docker build -t fab envs/docker

Run the image, replacing <path_to_fab> with the path on your host machine and <user> with using your username:

$ docker run --env PYTHONPATH=/fab -v <path_to_fab>/source:/fab -v /home/<user>:/home/<user> -it fab bash

Other¶

You may need to ask your system administrator to install the above requirements.

Using Python venv¶

Create an environment using Python’s builtin venv

$ python -m venv <env name>
$ cd <env name>
$ . bin/activate

Then install fab

$ pip install sci-fab

You’ll have to make sure the non-Python requirements are installed.