sbms-server

School of Biomedical Sciences (HKU) server wiki repository

For official notices, creating user account, etc., please go to: http://biomed1.sbms.hku.hk

Computing guidance for labs

• For members in YH Huang lab, see yhhuang-lab.md page.

Login

For login via ssh, read SSH.md page.

Environment with Conda

Instead of installing anaconda yourself, we recommend using the shared anaconda platform (/opt/anaconda3).

There is a base environment. But for specific task, create your own conda environment, so that you can control all versions of the packages for your analysis, which is important for reproducibility. Follow the conda manual to manage your environments.

For each environment, you can also add it as a separated kernel on jupyter lab, see here:

conda activate myenv

pip install ipykernel

python -m ipykernel install --user --name myenv --display-name "Python (myenv)"

VS Code

Visual Studio Code is a handy tool to access to server, e.g., edit scripts, running jobs. See more official instructions:

• https://code.visualstudio.com/docs/remote/ssh
• https://code.visualstudio.com/docs/datascience/jupyter-notebooks

Jupyter lab on server

For connect to Jupyter lab on server, read Jupyterlab.md page.

R

For using R from conda, read R.md page.

Mounting Remote File System

You may consider mount the files on one server (e.g., the CPU server) to another server (e.g., the GPU server) or to your local machine.

The NFS is not available, but your can use SSHFS for user level mounting files. See basic documentation of SSHFS. Example on mount files on CPU server to GPU server, you can run the following line on GPU server:

sshfs YOURS@biomed1.sbms.hku.hk:/storage/yhhuang /home/YOURS/yhhuang -o idmap=user

# Unmount files
# fusermount -u /home/YOURS/yhhuang