This material can be found in the Intro to RCC Workshop
Create a Conda Environment¶
First ssh into midway2
$ ssh <your_id>@midway2.rcc.uchicago.eduYou’ll want to load a Python module
$ module avail pythonshould yield
---------------------------- /software/modulefiles2 ----------------------------
python/anaconda-2019.03 python/cpython-3.8.5
python/anaconda-2020.02(default) python/intel-2020.up1
python/cpython-3.7.0
------------------------------- /etc/modulefiles -------------------------------You can load the default python module (python/anaconda-2020.02) using:
$ module load pythonOnce you have loaded the module, you can create a conda environment for whatever project you’re working on
$ conda create –name=<my-env> python==<python-vers>
$ conda activate <my-env>Run a Notebook from a login node¶
WARNING: nothing computationally intensive should be run from a login node
$ module load python
$ # export the login node IP address
$ export IP_ADDR=`ifconfig eno1 | grep 'inet ' | awk '{print $2}'`
$ # choose a port number
$ export PORT_NUM=$(shuf -i8000-9000 -n1)
$ # launch jupyter
$ jupyter notebook --no-browser --ip=${IP_ADDR} --port $PORT_NUMYou can then open the link that the command creates