Introduction¶
The contents of this documentation are mostly focused on our High-Performance Computing (HPC) Clusters at West Virginia University Research Computing (WVU-RC). However, WVU-RC offers a variety of technologies and services beyond HPC, the Introduction chapter summarizes those services and the policies regarding its usage. Purchasing nodes is also possible and described.
First-time users should go to Quick Start. This chapter is presented as a tutorial where the basic elements of using a computer cluster are presented and guide the user step by step from getting an account to submitting a job and take the results back to his/her own desktop computer.
Once you have followed the Quick Start we move into a chapter Basic Usage, this chapter assumes that user knows at least how to enter into the cluster and submit a basic job. The chapter provides basic information that was not covered in the Quick Start tutorial. We include more information about the command line interface, the various text-based editors, environment modules, and more details on the various elements that were skipped before.
The next chapter is Advanced Usage. Intended for users who feel comfortable with the Command Line Interface. Advanced users are supposed to know how to create scripts using Shell Scripting or any other interpreted language. Advanced users could be interested in installing their own software, using containers, or Conda environments.
The chapter Scientific Programming is intended for users with an interest in developing, testing and running their own programs on WVU’s clusters. We introduce the basics of using compilers, build systems, parallel programming, debugging, profiling, and optimization. Developers can be considered as advanced user who programs their own codes or integrates other codes in more than a simple linear workflow.
In the chapter Software Administration we make reference to the various tools that can be used to monitor the global health of the clusters. In general, users do not need to have that global view but knowing how the entire cluster works can give them insights about their own role in having an effective usage of the cluster.
The next chapter, Domain Specific Details collects sections for various packages that are relevant for a restricted number of researchers.
Chapter Clusters Specifications should serve as a reference for the current configuration of the clusters, in terms of hardware, software, modules, and queues.
The final chapter References collect tables and references about Unix commands, PBS options, and variables, it should be of good use when you know exactly what you are looking for.