Making Software Easy-Build Friendly: Streamlining HPC Software Deployment
Reproducible, maintainable, and scalable software builds are fundamental to high-performance computing (HPC) environments. However, the traditional approach of manually compiling and configuring scientific software is often time-consuming, error-prone, and difficult to replicate across systems. This presentation explores how making software EasyBuild-friendly significantly simplifies the process of deploying complex software stacks on HPC systems. EasyBuild is an open-source framework that automates the building, installation, and management of scientific software on HPCs. By creating EasyBuild “easyconfigs” and optionally writing custom “easyblocks,” users can define builds declaratively—reducing the need for manual intervention and ensuring consistent results across installations. The session will cover key principles for adapting or packaging software to be EasyBuild-friendly, including proper handling of dependencies, versioning, environment modules, and integration with software stack management tools like Lmod. Through practical examples, we will demonstrate how to convert manual build processes into automated EasyBuild recipes. We will highlight the benefits for both research software engineers and end-users: faster deployment, simplified software stack maintenance, and improved reproducibility. Attendees will also learn how to contribute back to the EasyBuild ecosystem, making their software accessible to a broader HPC audience. By adopting EasyBuild in research environments, institutions can reduce redundant work, standardize their software environments, and free up valuable researcher time. This talk aims to empower RSEs to leverage EasyBuild not just as users, but as contributors shaping the future of sustainable HPC software deployment. A recording of this session is available on YouTube: https://youtu.be/ruyzrDYU2s0
Institutions
- Max Planck Society (DE)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2025-12-17
- DOI
- https://doi.org/10.5281/zenodo.17968172
- Primary Topic
- Scientific Computing and Data Management
- Type
- article
- Field-Weighted Citation Impact
- 0.00