OSC paper earns honors at research computing conference
Authors focus on challenges of HPC software installation, testing
A paper authored by staff members of the Ohio Supercomputer Center (OSC) earned best paper honors in the Facilitation of Advanced Research Computing track at the Practice & Experience in Advanced Research Computing Conference (PEARC), being held July 28-Aug. 1 in Chicago.
The paper, titled “A Continuous Integration-Based Framework for Software Management,” describes a framework to automatically install and test software applications. OSC is a statewide resource supporting clients from academia, healthcare and industry. A wide range of these applications must be maintained to enable this broad user community to productively carry out its science. As new communities with new science needs start accessing OSC resources, the software requirements grow; the paper discusses OSC’s approach to managing this growth.
The paper’s authors describe a continuous integration-based framework for software management. In this framework, OSC staff rely upon automated software build tools and the ReFrame framework to implement automated testing, which can be triggered on commit to a local Gitlab repository. The paper documents the workflow and explains how staff deploy the software environment quickly and confidently using these tools.
The OSC paper was authored by Samuel Khuvis, Ph.D., scientific applications engineer; Zhi-Qiang You, Ph.D. scientific applications engineer; Heechang Na, Ph.D., scientific applications engineer; Scott Brozell, Ph.D., scientific applications senior engineer; Eric Franz, web & interface applications lead engineer; Trey Dockendorf, HPC systems engineer; Judith Gardiner, Ph.D., scientific applications senior engineer; and Karen Tomko, Ph.D., director of research software applications.
PEARC is an outgrowth of the Extreme Science and Engineering Discovery Environment (XSEDE) initiative of the National Science Foundation and serves as a forum for discussing challenges, opportunities and solutions among the broad range of participants in the research computing community. PEARC19 is focusing on machine learning and artificial intelligence, which are proving to be disruptive technologies in a diverse range of scientific fields from materials science to medicine.
According to the PEARC19 website, “The winners of the best paper and honorable mention paper awards were selected by independent juries from among the most highly-rated papers in the conference’s peer review process. The juries deemed the winning manuscripts to be comprehensively well rounded, reflective of deep understanding of the subject areas discussed and of very high quality overall. Multiple specific criteria were considered by the juries, depending on the track and topic of each paper, including potential influence on the state of the practice, potential impact on computation-based science, novelty, use of appropriate data and analysis methodologies, and clarity of presentation.”
The awards were announced last week and will be presented at a conference luncheon on Thursday.
Mr. Jamie Abel