Introduction to HPC Module system

Introduction to HPC Module System

We provide our users with a broad range of software and scientific libraries. On our primary clusters, we deploy software with spack and conda to foster reproducibility. Packages installed via spack have automatically generated module files available for each package, while our conda-deployed software is deployed in individual, centrally installed environments that can be loaded via module or activated traditionally.

Most installed packages are available as environment modules. You can find out about an installed software/module using the following command:-
 
module avail 
This command will list all the installed software and modules in your cluster.

Use:
module load <module/version>        to load a module
module unload <module/version>    when done.

Generally, use as few modules as possible at a time. Once you're done using a particular piece of software, unload it before you load another one to avoid incompatibilities.

Please read the "Software" chapter on our documentation for more details:-  https://docs.hpc.ucdavis.edu/software/ 

You can send a ticket to hpc-help@ucdavis.edu with any other software issues.

Conda

HPC@UCD clusters have a centrally installed Conda. Conda is our preferred and supported method for Python environments; we prefer not to centrally install any Python libraries. Users can also use our Conda install to install arbitrary software available in Conda repos in their home directories without needing to wait for systems administrators; many users may find this preferable, as our software install backlog is substantial.

To help with troubleshooting and compatibility, we do not provide support for Conda installations in user home directories; instead, users should load our central conda with module load Conda, which will automatically perform the initialization usually done by conda init.

$ module load conda
Loading conda/base/latest

This will activate the base environment automatically.

More specifically, the central install is a miniforge installation, which means that most conda commands can be replaced with mamba commands for a performance increase.

Here are the commands to create a new environment and install the environment:-

module load conda
conda create -n [name of the environment] -c [name of the channels] [list of packages to install]
conda activate [name of environment]

Note that you can leave off the -c bioconda, because the system-wide conda configuration already has the bioconda channel configured as well as conda-forge

You can also replace conda with mamba in the create phase, it will solve the environment much more quickly. When the environment is active, you can run the following command to see that it lives in your home directory:

echo $CONDA_PREFIX

Please see the page https://docs.hpc.ucdavis.edu/software/conda/ on our documentation for further information.

Python
 

Python is also installed and is available in the cluster Conda; you can load it using:

module load conda
conda activate
Python

R

R is a programming language and software environment for statistical computing and graphics. R offers a wide variety of statistics-related libraries and provides a favorable environment for statistical computing and design. In addition, the R programming language gets used by many quantitative analysts as a programming tool since it's useful for data importing and cleaning.

To see different versions of R installed on the cluster, run the command:

$ module avail -l R/

You can obtain R in your environment by loading the R module for the version you want, i.e.:

$ module load R/4.4.2

Once the module is loaded, you can run R like usual:

$ R

To check available libraries and packages within R, use the following command:

library(library-name) OR installed.packages()

You can install additional R packages with the normal method:

install.packages("package-name")

See more about R on our documentation:- https://docs.hpc.ucdavis.edu/software/rstudio/ 

RStudio Server

RStudio Server enables you to provide a browser-based interface to a version of R running on a remote Linux server, bringing the power and productivity of the RStudio IDE to server-based deployments of R.

RStudio Server is available through Open OnDemand on select clusters. See https://docs.hpc.ucdavis.edu/software/ondemand/ for more information.
 

How do I send Software Installation Request?